Дисертації з теми "+519.86"

Стаття зосереджена на проблемах управління логістичними процесами в умовах зростаючої конкуренції. Узагальнено перелік сфер діяльності підприємства, де можна виявити резерви за рахунок скорочення логістичних витрат. Проведено дослідження «вузьких місць» у взаємодії логістичних компаній з клієнтами та запропоновані варіанти їх усунення.
The article is concentrated on the problems of logistics processes management in the conditions of growing competition. The list of the enterprise activity spheres where it is possible to reveal reserves via reduction of logistic expenses is summarized. There is the study of "bottlenecks" in the interaction of logistics companies with customers and proposed options for their elimination.

Дисертація на здобуття наукового ступеня кандидата економічних наук за спеціальністю 08.00.04 – економіка та управління підприємствами (за видами економічної діяльності). – Національний технічний університет "Харківський політехнічний інститут", Харків, 2019. Дисертаційна робота присвячена актуальним проблемам теоретичних, методичних та практичних аспектів управління енергоефективністю промислового підприємства. В дисертації досліджено теоретичне підґрунтя, розглянуто різноманіття формулювань понять "енергоефективність", "енергозбереження", "енергетичний менеджмент", "управління енергоефективністю" та визначено використання їх у роботі. Узагальнено та систематизовано основні елементи та розвинуто теоретичний базис управління енергоефективністю підприємства. Запропоновано оцінювати енергоефективність підприємства за показником енергоємності, який являє собою відношення двох економічних величин: щорічних витрат на енергетичні ресурси до вартості виробленої продукції. Враховано, що управління енергоефективністю промислового підприємства базується на загальних принципах теорії управління, згідно якої система управління складається з двох підсистем - керуючої, яка управляє, і керованої, яка піддається управлінню і відчуває на собі дію (вплив) керуючої підсистеми. Акцентована увага на тому, що особливо важливим чинником впливу є завдання і умови сформульовані вищими ієрархічними рівнями управління енергоефективності. Встановлено, що завдання підвищення енергоефективності промислових підприємств української економіки вимагає створення відповідних умов на всіх рівнях управління. Такими рівнями є: державний, галузевий, регіональний та рівень підприємства. Встановлено, що концептуально рівні системи управління енергоефективністю можуть бути представлені у вигляді ієрархічної вертикалі із зворотними зав'язками. Проведений аналіз нормативно-правової бази у сфері енергоефективності та енергозбереження засвідчив, що на кожному етапі управління енергоефективністю для забезпечення процесу реалізації державної політики визначені конкретні завдання, при цьому характер завдань для кожного з рівнів має свою специфіку. Встановлено, що після першої світової енергетичної кризи промислово розвинуті країни приділили багато уваги вирішенню проблеми підвищення енергоефективності їхніх економік і набули значний досвід в цьому напрямку, що було відображено в міжнародному стандарті ISO 50001:2011 "Системи енергетичного менеджменту", що надалі був доповнено рядом покращуючих положень в стандарті ISO 50001:2018 "Системи енергетичного менеджменту. Вимоги та настанови щодо застосовування", в якому реалізовано управління за циклом постійного поліпшення PDCA. Цей стандарт повинен бути імплементован у нормативно-правові акти при реалізації "Енергетичної стратегії України до 2035 р.". На основі проведеного аналізу Законів України щодо енергозбереження, міжнародних та національних стандартів, методики енергоаудиту встановлена доцільність внесення доповнень з врахування мультиплікативного та синергетичного ефектів економії енергетичних ресурсів при модернізації енергоустаткування. Результати аналізу дозволили встановити взаємозв'язок рівня впровадження країнами стандарту ISO 50001 з показником енергоємності, а саме зниження його при поширеному впровадженні систем енергетичного менеджменту у промисловості, шо дозволило скоротити обсяги споживання енергії на одиницю продукції за період 2000 – 2017 рр. майже на 20%. Підвищення енергоефективності ВВП для України є особливо важливим напрямком розвитку економіки. Для підвищення енергоефективності необхідні значні інвестиції і для країни особливе значення має пошук внутрішніх інвестиційних джерел, для чого бажано мати інструментарій розрахунку прогнозного значення економії коштів на підприємстві від енергозбереження. Встановлено, що основними важелями покращення енергоефективності провідних країн були інвестиції та поліпшення управління. Значне підвищення енергоефективності може бути досягнуте за рахунок покращення управління енергоефективністю промислових підприємств; підвищення показника енергоефективності на промислових підприємствах, поліпшенням управління та впровадженням стандарту ISO 50001. Запропоновано та обґрунтовано доцільність використання в управлінні енергоефективністю промислового підприємства методу нейронних мереж для визначення показника енергоємності, який може бути використаний для оперативного надання інформації з прогнозної економії коштів за рахунок енергозбереження при багаторазових уточненнях в процесі визначення кінцевих значень показників виробництва на наступний рік, при формуванні енергетичної політики, створенні щорічних планів підвищення енергоефективності, планування інших показників виробництва, від яких залежить енергоефективність промислового підприємства. Розроблено методичний підхід визначення показника енергоємності підприємства за допомогою штучних нейронних мереж, із залученням методів експертних оцінок, апріорного ранжування та кореляційно-регресійного аналізу. Створена економіко - математична модель на базі багатошарового прямоспрямованого персептрону, синтезованого методом генетичного алгоритму. Виконано перевірка працездатності методичного підходу, а також його апробація за статистичними даними провідних харківських підприємств. В процесі апробації та аналізу її результатів зроблені висновки про необхідність: систематичного накопичення щорічних статистичних даних, як самого показника, так і чинників, що на нього впливають; аналізу розрахунків і встановлення коректності врахування даних; виявлення типових розбіжностей у трактовці того чи іншого чинника і введення відповідних корегувань; врахування особливостей підприємства; проведення за допомогою ШНМ досліджень з вибору раціональних значень виробничих показників – вхідних величин нейронної мережі, що впливають на показник енергоємності; дослідження по підвищенню точності і прогнозування за рахунок змін у структурі та виду нейронної мережі та інш. Такий широкий набір функцій і необхідність систематичного їх здійснення призвели до висновку про доцільність створення у структурі управління промислового підприємства підструктури, яка б включала фахівців здатних виконувати ці функції - групу інтелектуального енергоменеджмента (ІЕМ). Практичним результатом дисертаційної роботи, щодо врахування мультиплікативного та синергетичного ефектів є: по-перше, висновок про доцільність врахування цих ефектів при проведені внутрішнього енергоаудиту, для чого необхідно внести відповідні доповнення в його методику його проведення; по-друге, висновок про доцільність внесення у законодавчі та нормативно-правові акти доповнень, які стимулюють підприємства на пріоритетну модернізацію енергообладнання з мультиплікативним та синергетичним ефектом (як таку, що надає значну економію паливно-енергетичних ресурсів на державному рівні). Надані рекомендації з використання результатів роботи на різних ієрархічних рівнях управління енергоефективністю та у інших видах менеджменту промислового підприємства.
Thesis for granting the Degree of Candidate of Economic Sciences in specialty 08.00.04 - economy and management of the enterprises (by types of economic activity). - National Technical University "Kharkiv Polytechnic Institute", Kharkiv, 2019. The thesis is devoted to actual problems of theoretical, methodical and practical aspects of management of energy efficiency of an industrial enterprise. In the dissertation the theoretical basis is investigated, a variety of formulations of concepts "energy efficiency", "energy saving", "energy management", "energy efficiency management" are considered and their use in work is determined. It is proposed to evaluate the energy efficiency of the enterprise by the indicator of energy intensity, which is the ratio of two economic values: the annual cost of energy resources to the cost of production. It is taken into account that the management of energy efficiency of an industrial enterprise is based on the general principles of the theory of control, according to which the control system consists of two subsystems - the managing one, and the controlled one, which is exposed to the control and has the effect (influence) of the control subsystem. Emphasis is placed on the fact that tasks and conditions formulated by higher hierarchical levels of energy efficiency management are a particularly important factor of influence. It is established that the task of improving the energy efficiency of industrial enterprises of the Ukrainian economy requires the creation of appropriate conditions at all levels of government. Such levels are: state, sectoral, regional and enterprise level. It has been established that conceptually levels of energy efficiency management systems can be represented as a hierarchical feedback loop. The conducted analysis of the regulatory framework in the field of energy efficiency and energy saving showed that at each stage of energy efficiency management specific tasks were defined for ensuring the process of implementation of state policy, with the nature of tasks for each of the levels having its own specificity. It was found that after the first world energy crisis, industrialized countries paid much attention to solving the problem of improving the energy efficiency of their economies and gained considerable experience in this direction, which was reflected in the international standard ISO 50001: 2011 "Energy Management Systems", which was further supplemented by a number of improving of the provisions in ISO 50001: 2018 "Energy management systems. Requirements and Guidelines for Implementation, "which implements PDCA Continuous Improvement Management. This standard should be implemented in the regulatory acts when implementing the "Energy Strategy of Ukraine to 2035". On the basis of the conducted analysis of the Laws of Ukraine on energy saving, international and national standards, the methodology of energy audit, it is established the expediency of introducing amendments to take into account the multiplicative and synergistic effects of saving energy resources in the modernization of energy equipment. The results of the analysis made it possible to correlate the level of ISO 50001 implementation by countries with the indicator of energy intensity, namely reducing it with the widespread introduction of energy management systems in industry, which allowed to reduce energy consumption per unit of production for the period 2000 - 2017 by almost 20%. Increasing GDP energy efficiency for Ukraine is a particularly important area of economic development. Significant investments are needed to improve energy efficiency, and the search for domestic investment sources is of particular importance for the country, for which it is desirable to have a toolkit for calculating the projected value of energy savings for the enterprise. It was found that the main levers for improving the energy efficiency of the leading countries were investment and improved governance. A significant increase in energy efficiency can be achieved by improving the management of energy efficiency of industrial enterprises; improving the energy efficiency of industrial enterprises, improving management and implementing the ISO 50001 standard. It is proposed and substantiated the feasibility of using the neural network method in the energy efficiency management of an industrial enterprise to determine the energy intensity index, which can be used for prompt provision of information on forecasted cost savings due to energy savings with multiple refinements in the process of determining the final values of the production form for the next year policies, setting up annual energy efficiency plans, planning other indicators production, on which the energy efficiency of an industrial enterprise depends. The methodical approach of determination of the index of energy intensity of the enterprise with the help of artificial neural networks (ANN), with the use of methods of expert estimations, a priori ranking and correlation-regression analysis is developed. An economic and mathematical model was created bas ed on a multilayer directional perceptron synthesized by the genetic algorithm method. The efficiency of the methodological approach was checked, as well as its testing according to the statistics of the leading Kharkiv enterprises. In the process of testing and analyzing its results, the conclusions were drawn about the need for: systematic accumulation of annual statistics, both the indicator itself and the factors that influence it; analysis of calculations and establishment of correctness of data accounting; identifying typical differences in the interpretation of a factor and making appropriate adjustments; taking into account the features of the enterprise; carrying out with the help of ANN research on the choice of rational values of production indicators - input values of the neural network that affect the indicator of energy intensity; studies on improving accuracy and forecasting due to changes in the structure and type of neural network, etc. Such a wide range of functions and the need for their systematic implementation led to the conclusion that it is advisable to create in the management structure of an industrial enterprise a structure that would include specialists capable of performing these functions - the Intelligent Energy Management Group (IEM). The practical result of the dissertation work on taking into account the multiplicative and synergistic effects is: first, the conclusion about the expediency of taking into account these effects when conducting an internal energy audit, for which it is necessary to make appropriate additions to its methodology of its conduct; secondly, the conclusion about the expediency of introducing into the legislative and regulatory acts the amendments that stimulate the enterprises to prioritize the modernization of power equipment with multiplicative and synergistic effect (as such, which provides significant savings of fuel and energy resources at the state level). Recommendations are given on the use of the results of work at different hierarchical levels of energy efficiency management and in other types of industrial enterprise management.

Дисертація на здобуття наукового ступеня кандидата економічних наук за спеціальністю 08.00.04 – економіка та управління підприємствами (за видами економічної діяльності). – Національний технічний університет "Харківський політехнічний інститут", Харків, 2019. Дисертаційна робота присвячена актуальним проблемам теоретичних, методичних та практичних аспектів управління енергоефективністю промислового підприємства. В дисертації досліджено теоретичне підґрунтя, розглянуто різноманіття формулювань понять "енергоефективність", "енергозбереження", "управління енергоефективністю". Проаналізовані Закони України щодо енергозбереження, міжнародні та національні стандарти, методика енергоаудиту та встановлена доцільність внесення відповідних доповнень з врахування мультиплікативного та синергетичного ефектів економії енергетичних ресурсів при модернізації енергоустаткування. Запропоновано та обґрунтовано доцільність використання в управлінні енергоефективністю промислового підприємства методу штучних нейронних мереж (ШНМ). Розроблено методичний підхід визначення показника енергоємності підприємства за допомогою ШНМ, із залученням методів експертних оцінок, апріорного ранжування та кореляційно-регресійного аналізу. Створена математична модель на базі багатошарового прямоспрямованого персептрону, синтезованого методом генетичного алгоритму. Виконано перевірка працездатності методичного підходу, а також його апробація за статистичними даними провідних харківських підприємств. Надані рекомендації з використання результатів роботи на різних ієрархічних рівнях управління енергоефективністю та у інших видах менеджменту промислового підприємства.
Thesis for granting the Degree of Candidate of Economic Sciences in specialty 08.00.04 – economy and management of the enterprises (by types of economic activity). – National Technical University "Kharkiv Polytechnic Institute", Kharkiv, 2019. The thesis is devoted to actual problems of theoretical, methodical and practical aspects of management of energy efficiency of an industrial enterprise. In the dissertation the theoretical basis is investigated, the variety of formulations of the concepts of "energy efficiency", "energy saving", "energy efficiency management" and the use of them in work are considered. The Laws of Ukraine on energy conservation, the international and national standards, the methodology of energy audit are analyzed, and the expediency of introducing the corresponding additions to take into account the multiplicative and synergetic effects of energy resources saving during the modernization of the power equipment is established. It is suggested and justified the feasibility of using artificial neural networks (ANN) in the management of energy efficiency of an industrial enterprise. The method of determination of the energy intensity of an enterprise with the help of ANN is developed, with the involvement of expert estimation methods, a priori ranking and correlation-regression analysis. A mathematical model based on a multilayered straightforward perceptron synthesized by the genetic algorithm method was created. The efficiency of the methodological approach was checked, as well as its testing according to the statistics of the leading Kharkiv enterprises. In the process of testing and analyzing its results, the conclusions were drawn about the need for: systematic accumulation of annual statistics, both the indicator itself and the factors that influence it; analysis of calculations and establishment of correctness of data accounting; identifying typical differences in the interpretation of a factor and making appropriate adjustments; taking into account the features of the enterprise; carrying out with the help of ANN research on the choice of rational values of production indicators - input values of the neural network that aff ect the indicator of energy intensity; studies on improving accuracy and forecasting due to changes in the structure and type of neural network, etc. Such a wide range of functions and the need for their systematic implementation led to the conclusion that it is advisable to create in the management structure of an industrial enterprise a structure that would include specialists capable of performing these functions - the Intelligent Energy Management Group (IEM). The practical result of the dissertation work on taking into account the multiplicative and synergistic effects is: first, the conclusion about the expediency of taking into account these effects when conducting an internal energy audit, for which it is necessary to make appropriate additions to its methodology of its conduct; secondly, the conclusion about the expediency of introducing into the legislative and regulatory acts the amendments that stimulate the enterprises to prioritize the modernization of power equipment with multiplicative and synergistic effect (as such, which provides significant savings of fuel and energy resources at the state level). Recommendations are given on the use of the results of work at different hierarchical levels of energy efficiency management and in other types of industrial enterprise management.

Розроблена математична (імітаційна) модель ідеального ЕКГ-сигналу у вигляді реалізованого програмного коду з унікальним інтерфейсом. Даний програмний продукт (з розробленим інтерфейсом) можна використовувати для побудови схем порівняння в медичних експертних і автоматичних діагностичних системах.

We give criteria for the stability of a very general queueing model under different levels of control. A complete classification of stability (or positive recurrence), transience and null-recurrence is presented for the two queue model. The stability and instability results are extended for models with N > 3 queues. We look at a broad class of models which can have the following features: Customers arrive at one, several or all of the queues from the outside with exponential inter arrival times. We often have the case where a arrival stream can be routed so that under different routing schemes each queue can have external arrivals, i.e. we assume we have some control over the routing of the arrivals. We also consider models where the arrival streams are fixed. We view the service in a more abstract way, in that we allow a number к of different service configurations. Under every such service configuration service is provided to some or all of the queues, length of service time can change from one service configuration to another and we can change from one configuration to another according two some control policy. The service times are assumed to be exponentially distributed. The queueing models we consider are networks where, after completion at one queue, a customer might be fed back into another queue where it will be served another time often under with a different service time. These feedback probabilities change with the service configurations. Our interest is in different types of control policies which allow us to change the routing of arrivals and configurations of the service from time to time so that the controlled queue length process (which in most cases is Markov) is stable. The semi-martingale or Lyapunov function methods we use give necessary and sufficient conditions for the stability classification. We will look at some two queue models with different inter arrival and service times where the queueing process is still Markov.

Chapter 1 gives a brief introduction to modelling epidemics in both a deterministic and stochastic framework. Chapter 2 gives a brief introduction to the single-group Susceptible-Infected-Susceptible (SIS) model and we mention and establish all the necessary background theory we apply to the model with the aim of extending these theories and techniques to our own multi-group SIS model in order to obtain results concerning the short and long-term behaviour of the epidemic. In Chapter 3 we define a k-group SIS model which will allow us to examine the effects of heterogeneity in three different categories - the infectivity of infectious individuals, their mixing behaviour and an individual's susceptibility to the disease. In doing this, we describe the dynamics of the model including how to represent it in a deterministic framework, the quasi-stationary distribution and the time to extinction. We apply a branching process approximation to the model, viable for the early stages of a disease, using well established theory. In Chapter 4 we look specifically at the early stages of the epidemic based on the branching process approximation and produce numerical results on how the probability of disease emergence behaves as the basic reproduction number Ra increases. We contrast 2-group heterogeneous models against a homogeneous model, for epidemics assuming either an exponential or constant infectious period. We then analyse these results with an iterative and inductive proof showing that the emergence probability for a heterogeneous model will always be less than that for a homogeneous model, not just in the limit but at all stages of iterative convergence. Next we provide a proof which shows that for the non-separable general model this ordering exists for any given infectious period. We then go on to look at comparing two heterogeneous models to one another under various sets of parameters and use majorization theory as a tool for doing so. We use orderings referred to as ordinary majorization, p-majorization and pq-majorization to show that there is an inferred ordering of emergence probabilities when comparing multi-group he- terogeneous models to one another. In Chapter 5 we study the long-term behaviour of the stochastic multi-group SIS model. We begin by formulating conditions for the general model under which feasible equilibria exist and conditions where either the disease-free or endemic equilibria are stable. For a 2-group version, we calculate numerically the determi- nistic equilibrium values, stochastic means and quasi-stationary distributions for a range of Ra values. We use an Ornstein-Uhlenbeck process to approximate the quasi-stationary distribution and assess the accuracy of the approximation. We then calculate the expected time to extinction and use a coefficient of variation approximation as a proxy for this and discuss the suitability of such an approxi- mation to the exact results. These analyses are all carried out for models which exhibit heterogeneity in infectivity, mixing or susceptibility.

This thesis introduces a novel model for characterising network delays and a method derived from it for generating representative synthetic network delays. The model of network delays is based on combining multiple Weibull probability distributions to accurately fit the delay histogram observed in the delay traces. The idea of using the Weibull distribution as a basis to build the delay histogram is based on earlier studies on queueing theory under self-similar input traffic. However, such theoretical results have not been validated in real end-to-end scenarios. In this work, a method for finding the optimal model parameters will be introduced, tested and validated with measurements collected under real network activity. Additionally, two extensions of this algorithm shall be introduced: a real-time modification for tracking network delays adaptively; and an algorithm for generating synthetic but statistically equivalent network delays. Finally, network research topics will be introduced as possible applications and further directions of research. These include: real-time network management, service differentiation, QoS routing and delay-based congestion-control.

There is a growing recognition for the multiple criteria decision making (MCDM) based multi-objective evolutionary algorithms (MOEAs) for tackling many-objective optimisation problems (MaOPs). In that, the aim is to utilise the decision makers’ (DMs’) preferences to guide the search towards a few solutions, as against the whole Pareto-optimal front (POF). This thesis is based on the premise that the practical utility of the MCDM based MOEAs may be impaired due to the lack of–objectivity (a rational basis); repeatability (identical preferences for identical options); consistency (alike preferences across multiple interaction stages); and coherence (alike preferences by multiple DMs) in the DMs’ preferences. To counter these limitations, this thesis aimed at developing offline and online decision support by capturing the preference-structure of the objective functions inherent in the problem model itself. This aim has been realised through the following objectives: • Identification of a criterion for the decision support: in that, preservation of the correlation– structure of the MOEA solutions is found to be a robust criterion in the case of MaOPs. • Development of a machine learning based offline objective reduction framework: it com¬prises of linear and nonlinear objective reduction algorithms, which facilitate the decision support through revelation of: (i) the redundant objectives (if any), (ii) preference-ranking of the essential objectives, (iii) the smallest objective sets corresponding to pre-specified errors, and (iv) the objective sets of pre-specified sizes that correspond to minimum error. • Development of an online objective reduction framework: it addresses a major pitfall associated with the offline framework, that–an essential objective if erroneously eliminated as redundant, has no scope of being reconsidered in the subsequent analysis. This pitfall is countered through a probabilistic retention of all the objectives, and this serves as a self-correcting mechanism that enhances the overall accuracy. • Timing the decision support: it is acknowledged that the revelations by the decision support may vary depending on when the offline or online framework is applied during an MOEA run. This uncertainty on the timing of the decision support is countered through the proposition of an entropy based dissimilarity measure. The efficacy of the proposed frameworks is investigated against a broad range of test problems (scaled up to 50 objectives) and some real-world MaOPs; and, the accuracy of the corresponding decision support is compared against that of an alternative approach based on preserving the dominance relations. The results illustrate that the proposed frameworks and the corresponding decision support bear significant utility for those MaOPs, where not all the objectives are essential, or equally important for describing the true POF. The considered real-world problems also bear evidence of the fact that MaOPs with redundant and disparately important objectives may commonly exist in practice.

In today‟s highly competitive business environment it is vital to have smart and robust decision making framework for companies to be competitive or even to stay in the business. Profit margin increases is no longer a result of producing and bring more products to the market. Instead it is also a result of reducing cost, in particular tooling cost. In order to succeed with this, industry need to look in to innovative intelligent systems to enhance their process development so that maximum utilisation of tools can be achieved. Tooling is part of a process hence having an optimal process design is one ideal strategy for best utilising of tools. In design optimisation however presence of uncertainty in design variables and in the mathematical model (used for representing the real life process) is inevitable. For reliable design solution to be found this process complexity need to be addressed. The aim of this research is to understand work roll system optimisation and thermal issues within rolling system design, understand uncertainties and sources of uncertainties and develop Genetic Algorithm (GA) based solution frameworks so that a conscious decision, that prolong tool life can be made. The thesis has proposed a framework for generating approximate models from numeric finite element (FE) data. Using the proposed framework a number of quantitative work roll system thermal analysis and optimisation models were generated and used in subsequent optimisation process. In the absence of a suitable multi-pass model that exhibits the features of a multi-stage process; confident decision making will not be possible. Hence the research has developed a quantitative multi-pass model to simulate the work roll system thermal analysis and optimisation problem that represents the relationships as well as inherited features between passes. The research has developed a Genetic Algorithm based optimisation framework that deals with the constraint quantitative problem as well as the uncertainty, in the design space and fitness function. The research also proposed a post GA result analysis methodology for identifying the final best optimal design solution for the research many objective high dimensional work roll system problems in presence of uncertainty The performance of the proposed frameworks was studied and analysed through case studies. The research also identifies future research directions in the subject area.

The desire to optimise problems is as prevalent in today's society as it has ever been. The demand for increases in speed and efficiency is relentless and has resulted in the need for mathematical models to bear greater resemblance to real-life situations. This focus on increased realism has paved the way for new dynamic variants to classic optimisation problems. This thesis begins by considering the Dynamic Vehicle Routing Problem. The basic premise of this routing problem is as follows a percentage of customers are known a priori, for which routes are constructed, further customers then arrive during the course of the working day and need to be incorporated into an evolving schedule. Literature has proposed a timeslot approach, whereby one partitions the working day into a series of smaller problems, that one is then required to solve in succession. This technique is used to produce a variety of metaheuristics based implementations, most noticeably Ant Colony Optimisation and Tabu Search. Consideration is then given to the Dynamic Vehicle Routing Problem with Time Windows. This problem is similar to the Dynamic Vehicle Routing Problem, but requires each customer to be serviced within a predefined period of the day. A metaheuristic approach adapted from the most successful algorithm implemented on the Dynamic Vehicle Routing Problem is presented. Finally consideration is given to a time-based decomposition technique for the Vehicle Routing Problems with Time Windows (Large-Scale instances). This work makes use of the dynamic solution technique developed in the preceding work, and is used in conjunction with an Ant Colony Optimisation algorithm and a descent algorithm.

In this thesis, the Erlang queueing model Af/i/l, where customers arrive at random mean rate A and service times have an Erlang distribution with parameter k and iro service rate u, has been considered from different perspectives. Firstly, an analytic metl of obtaining the time-dependent probabilities, pn,,(() for the M/Ek/l system have t> proposed in terms of a new generalisation of the modified Bessel function when initk there are no customers in the system. Results have been also generalised to the case wl initially there are a customers in the system. Secondly, a new generalisation of the modified Bessei function and its generating function have been presented with its main properties and relations to other special functii (generalised Wright function and Mittag-Leffler function) haw been noted. Thirdly, the mean waiting tune in the queue, H',(f), has been evaluated, using Lucha results. The double-exponential approximation of computing Yq(t) has been proposed different values of p. which gives results within about % of the 'exact1 values obtained fr numerical solution of the differential-difference equations. The advantage of this approximation is that it provides additional information, via its functional form of the characterisl of the transient solution. Fourthly, the inversion of the Laplace transform with the application to the queues 1 been studied and verified for A//A//1 and M/Ek/l models of computing Wq{t}. Finally, an application of the A//fi/l queue has been provided in the example of hour traffic flow for the Severn Bridge. One of the main reasons for studying queue models from a theoretical point of view is to develop ways of modelling real-life system. The analytic results have been confirmed with the simulation.

The Critical Care Unit (CCU) is the sector of the hospital where, as the name suggests, critically ill patients receive treatment. The main aim of this research is to identify and apply suitable Operational Research techniques to model patient flow in the CCU at the University Hospital of Wales, Cardiff. The Operational Research techniques employed in this thesis include queueing theory and simulation. These methods have been utilised previously in the field of healthcare with much success. The thesis begins by considering two aspects of queueing theory, namely batch service queueing theory and batch arrival queueing theory. The latter of these is utilised to model patient flow within the CCU. Although queueing theory may be used as a good approximation to activities in the Unit, it does not incorporate all aspects of real-life. Thus discrete-event simulation is suggested as an alternative approach. Two types of statistical analysis, CART and Regression, are applied to both length of stay and mortality variables. The results from these statistical tests are compiled and investigated in more depth. Finally, a discrete event simulation model is built in Visual Basic for Applications, for Microsoft Excel. This simulation model incorporates many of the complexities of a CCU, such as patient priority and cancellation of scheduled patients if all beds on the Unit are occupied. The model is then used to test various "what-if type" scenarios, including the possibility of funding additional beds, the concept of ring-fencing of beds for different levels of care, and the likely effect of reducing the impact of bed-blocking.

This research focuses on nature-inspired optimisation algorithms, in particular, the Particle Swarm Optimisation (PSO) Algorithm and the Bees Algorithm. The PSO Algorithm is a population-based stochastic optimisation technique first invented in 1995. It was inspired by the social behaviour of birds flocking or a school of fish. The Bees Algorithm is a population-based search algorithm initially proposed in 2005. It mimics the food foraging behaviour of swarms of honey bees. The thesis presents three algorithms. The first algorithm called the PSO-Bees Algorithm is a cross between the PSO Algorithm and the Bees Algorithm. The PSO-Bees Algorithm enhanced the PSO Algorithm with techniques derived from the Bees Algorithm. The second algorithm called the improved Bees Algorithm is a version of the Bees Algorithm that incorporates techniques derived from the PSO Algorithm. The third algorithm called the SNTO-Bees Algorithm enhanced the Bees Algorithm using techniques derived from the Sequential Number-Theoretic Optimisation (SNTO) Algorithm. To demonstrate the capability of the proposed algorithms, they were applied to different optimisation problems. The PSO-Bees Algorithm is used to train neural networks for two problems, Control Chart Pattern Recognition and Wood Defect Classification. The results obtained and those from tests on well known benchmark functions provide an indication of the performance of the algorithm relative to that of other swarm-based stochastic optimisation algorithms. The improved Bees Algorithm was applied to mechanical design optimisation problems (design of welded beams and coil springs) and the mathematical benchmark problems used previously to test the PSO-Bees Algorithm. The algorithm incorporates cooperation and communication between different neighbourhoods. The results obtained show that the proposed cooperation and communication strategies adopted enhanced the performance and convergence of the algorithm. The SNTO-Bees Algorithm was applied to a set of mechanical design optimisation problems (design of welded beams, coil springs and pressure vessel) and mathematical benchmark functions used previously to test the PSO-Bees Algorithm and the improved Bees Algorithm. In addition, the algorithm was tested with a number of deceptive multi modal benchmark functions. The results obtained help to validate the SNTO-Bees Algorithm as an effective global optimiser capable of handling problems that are deceptive in nature with high dimensions.

Quantum annealing is the quantum equivalent of the well known classical simulated annealing algorithm for combinatorial optimization problems. Despite the appeal of the approach, quantum annealing algorithms competitive with the state of the art for specific problems hardly exist in the literature. Graph colouring is a difficult problem of practical significance that can be formulated as combinatorial optimization. By introducing a symmetry-breaking problem representation, and finding fast incremental techniques to calculate energy changes, a competitive graph colouring algorithm based on quantum annealing is derived. This algorithm is further enhanced by tuning simplification techniques; replica spacing techniques to increase robustness; and a messaging protocol, which enables quantum annealing to efficiently take advantage of multiprocessor environments. Additionally, observations of some patterns in the tuning for random graphs led to a more effective algorithm able to find new upper bounds for several widely-used benchmark graphs, some of which had resisted improvement in the last two decades.

Solving real-world combinatorial problems is involved in many industry fields to minimise operational cost or to maximise profit, or both. Along with continuous growth in computing power, many asset management decision-making processes that were originally solved by hand now tend to be based on big data analysis. Larger scale problem can be solved and more detailed operation instructions can be delivered. In this thesis, we investigate models and algorithms to solve large scale Geographically Distributed asset Maintenance Problems (GDMP). Our study of the problem was motivated by our business partner, Gaist solutions Ltd., to optimise scheduling of maintenance actions for a drainage system in an urban area. The models and solution methods proposed in the thesis can be applied to many similar issues arising in other industry fields. The thesis contains three parts. We firstly built a risk driven model considering vehicle routing problems and the asset degradation information. A hyperheuristic method embedded with customised low-level heuristics is employed to solve our real-world drainage maintenance problem in Blackpool. Computational results show that our hyperheuristic approach can, within reasonable CPU time, produce much higher quality solutions than the scheduling strategy currently implemented by Blackpool council. We then attempt to develop more efficient solution approaches to tackle our GDMP. We study various hyperheuristics and propose efficient local search strategies in part II. We present computational results on standard periodic vehicle routing problem instances and our GDMP instances. Based on manifold experimental evidences, we summarise the principles of designing heuristic based solution approaches to solve combinatorial problems. Last bu not least, we investigate a related decision making problem from highway maintenance, that is again of interest to Gaist solutions Ltd. We aim to make a strategical decision to choose a cost effective method of delivering the road inspection at a national scale. We build the analysis based on the Chinese Postman Problem and theoretically proof the modelling feasibility in real-world road inspection situations. We also propose a novel graph reduction process to allow effective computation over very large data sets.

Seriation is a combinatorial optimisation problem that aims to sequence a set of objects such that a natural ordering is created. A large variety of applications exist ranging from archaeology to bioinformatics and text mining. Initially, a thorough and useful quantitative analysis compares different seriation algorithms using the positional proximity coefficient (PPC). This analysis helps the practitioner to understand how similar two algorithms are for a given set of datasets. The first contribution is consensus seriation. This method uses the principles of other consensus based methods to combine different seriation solutions according to the PPC. As it creates a solution that no individual algorithm can create, the usefulness comes in the form of combining different structural elements from each original algorithms. In particular, it is possible to create a solution that combines the local characteristics of one algorithm together with the global characteristics of another. Experimental results show that compared to consensus ranking based methods, using the Hamming, Spearman and Kendall coefficients, the consensus seriation using the PPC gives generally superior results according to the independent accumulated relative generalised anti-Robinson events measure. The second contribution is a metaheuristic for creating good approximation solutions very large seriation problems. This adapted harmony search algorithm makes use of modified crossover operators taken from genetic algorithm literature to optimise the least-squares criterion commonly used in seriation. As for all combinatorial optimisation problems, there is a need for metaheuristics that can produce better solutions quicker. Results show that that algorithm consistently outperforms existing metaheuristic algorithms such as genetic algorithm, particle swarm optimisation, simulated annealing and tabu search as well as the genetic algorithm using the modified crossover operators, with the main advantage of creating a much superior result in a very short iteration frame. These two major contributions offer practitioners and academics with new tools to tackle seriation related problems and a suggested direction for future work is concluded.

In this thesis we study three new extensions of scheduling models with both practical and theoretical relevance, namely synchronization, pliability and resiliency. Synchronization has previously been studied for flow shop scheduling and we now apply the concept to open shop models for the first time. Here, as opposed to the traditional models, operations that are processed together all have to be started at the same time. Operations that are completed are not removed from the machines until the longest operation in their group is finished. Pliability is a new approach to model flexibility in flow shops and open shops. In scheduling with pliability, parts of the processing load of the jobs can be re-distributed between the machines in order to achieve better schedules. This is applicable, for example, if the machines represent cross-trained workers. Resiliency is a new measure for the quality of a given solution if the input data are uncertain. A resilient solution remains better than some given bound, even if the original input data are changed. The more we can perturb the input data without the solution losing too much quality, the more resilient the solution is. We also consider the assignment problem, as it is the traditional combinatorial optimization problem underlying many scheduling problems. Particularly, we study a version of the assignment problem with a special cost structure derived from the synchronous open shop model and obtain new structural and complexity results. Furthermore we study resiliency for the assignment problem. The main focus of this thesis is the study of structural properties, algorithm development and complexity. For synchronous open shop we show that for a fixed number of machines the makespan can be minimized in polynomial time. All other traditional scheduling objectives are at least as hard to optimize as in the traditional open shop model. Starting out research in pliability we focus on the most general case of the model as well as two relevant special cases. We deliver a fairly complete complexity study for all three versions of the model. Finally, for resiliency, we investigate two different questions: `how to compute the resiliency of a given solution?' and `how to find a most resilient solution?'. We focus on the assignment problem and single machine scheduling to minimize the total sum of completion times and present a number of positive results for both questions. The main goal is to make a case that the concept deserves further study.

In this thesis we present a new stochastic optimisation model arising from supplyside management of power networks. We provide the exact optimal solution under assumption that the environment is Markovian. For the semi-Markovian environment we establish existence of an optimal policy in an important subclass of policies. Finally, we solve the problem for a number of particular examples of environment.

This thesis considers two di?cult combinatorial optimization problems for request routing and client assignment in content distribution networks. The aim is to introduce lower and upper bounds to estimate optimal solutions. Existing solution methods and techniques for similar problems have been reviewed. The first problem consists of minimizing the total network cost for request routing with no origin server by considering the delay function. The second problem is cache allocation problem. Lagrangian relaxation and two perspective cuts are used to linearize first problem and to find valid lower bounds. Two different linearization methods are implemented for cache allocation problem. Iterated Variable Neighborhood Descent and Tabu Search are two solution methods which are suggested to find best upper bounds. Different local search operators are introduced to improve objective function values as follows: swap, remove-insert, insert from origin server to a proxy, insert from one proxy to another proxy, swap between origin server and a proxy, swap between two proxies and cyclic exchange. All computational results are presented on randomly generated instances.

A novel extension of the classical vehicle routing and scheduling problem is proposed that integrates aspects of machine scheduling into vehicle routing. Associated to each customer is a release date that defines the earliest time that the order is available to leave the depot for delivery, and a due date that indicates the time by which the order should ideally be delivered to the customer. The objective is to minimise a convex combination of the operational costs and customer service level, measured as total distance travelled and total weighted tardiness, respectively. A formal definition of the problem is presented, and a variety of benchmark instances are generated to analyse the problem experimentally, and evaluate the performance of any solution approaches developed. Both experimental and theoretical contributions are made in this thesis, and these arise from the development of mixed integer linear programming(MIP) formulations, efficient heuristics, and a Dantzig-Wolfe decomposition and associated column generation algorithm. The MIP formulations extend commodity flow formulations of the capacitated vehicle routing problem, and are generally related by aggregation or disaggregation of the variables. Although a set of constraints is presented that is only valid form-commodity flow formulations. A path-relinking algorithm (PRA) is proposed that exploits the efficiency and aggressive improvement of neighbourhood search, but relies on a new path-relinking procedure and advanced population management strategies to navigate the search space effectively. To provide a comparator algorithm to the PRA, we embed the neighbourhood search into a standard iterated local search algorithm. The Dantzig-Wolfe decomposition of the problem yields a traditional set-partitioning formulation, where the pricing problem (PP) is an elementary shortest path problem with resource constraints and weighted tardiness. Two dynamic programming (DP) formulations of the PP are presented, modelling the weighted tardiness of customers in a path as a pointwise function of the release dates, or decomposing the states over the release dates. The CG algorithm relies on a multi-phase pricing algorithm that utilises DP heuristics, and a decremental state-space relaxation algorithm that solves an ng-route relaxation at each iteration. Extensive computational experiments on the benchmark instances show that the newly defined features have a significant and varied impact on the problem. As a result, finding tight lower bounds and eventually optimal solutions is highly complex, but tight upper bounds can be found efficiently using advanced heuristics.

The vehicle routing problem (VRP) has received particular attention, in the field of transportation and logistics. Producing good solutions for the problem is of interest both commercially and theoretically. Reliable solutions to real life applications require an approach based on realistic assumptions that resemble real-world conditions. In that respects, this thesis studies vehicle routing problems on real road networks addressing aspects of the problem that need to be modelled on the original road network graph and aims to provide appropriate modelling techniques for solving them. As a preliminary step, chapter 2 studies the travelling salesman problem (TSP) on real road networks, referred to as the Steiner TSP (STSP) and proposes alternative integer programming formulations for the problem and some other related routing problems. The performances of formulations is examined both theoretically and computationally. Chapter 3 highlights the fact that travel speeds on road networks are correlated and uses a real traffic dataset to explore the structure of this correlation. In conclusion, it is shown that there is still significant spatial correlations between speeds on roads that are up to twenty links apart, in our congested road network. Chapter 4 extends chapter 2 and incorporates the findings of chapter 3 into a modelling framework for VRP. The STSP with correlated costs is defined as a potentially useful variant of VRP that considers the costs in the STSP to be stochastic random variables with correlation. The problem is then formulated as a single-objective problem with eight different integer programming formulations presented. It is then shown how to account for three different correlation structures in each of the formulations. Chapter 5 considers the VRPs with time windows and shows how most of the exact algorithms proposed for them, might not be applicable if the problem is defined on the original road network graph due to the underlying assumption of these algorithms that the cheapest path between a pair of customers is the same as the quickest path. This assumption is not always true on a real road network. Instead some alternative pricing routines are proposed that can solve the problem directly on the original graph.

Ant Colony Optimization (ACO) algorithms which belong to metaheuristic algorithms and swarm intelligence algorithms have been the focus of much attention in the quest to solve optimization problems. These algorithms are inspired by colonies of ants foraging for food from their nest and have been considered state-of-art methods for solving both discrete and continuous optimization problems. One of the most important phases of ACO algorithms is the construction phase during which an ant builds a partial solution and develops a state transition strategy. There have been a number of studies on the state transition strategy. However, most of the research studies look at how to improve pheromone updates rather than at how the ant itself makes a decision to move from a current position to the next position. The aim of this research is to develop a novel state transition strategy for Ant Colony Optimization algorithms that can improve the overall performance of the algorithms. The research has shown that the state transition strategy in ACO can be improved by introducing non-rational decision-making. The new proposed algorithm is called Cognitive Ant Colony Optimization and uses a new concept of decision-making taken from cognitive behaviour theory. In this proposed algorithm, the ACO has been endowed with non-rational behaviour in order to improve the overall optimization behaviour of ants during the process. This new behaviour will use a non-rational model named prospect theory (Kahneman & Tversky, 1979) to select the transition movements of the ants in the colony in order to improve the overall search capability and the convergence of the algorithm. The new Cognitive Ant Colony Optimization framework has been tested on the Travelling Salesman Problem (TSP), Water Distribution System and Continuous optimization problems. The results obtained show that our algorithm improved the performance of previous ACO techniques considerably.

The ant colony optimization (ACO) metaheuristic is inspired by the foraging behaviour of real ant colonies. Similarly with other metaheuristics, ACO suffers from stagnation behaviour, where all ants construct the same solution from early stages. In result, the solution quality may be degraded because the population may get trapped on local optima. In this thesis, we propose a novel approach, called direct communication (DC) scheme, that helps ACO algorithms to escape from a local optimum if they get trapped. The experimental results on two routing problems showed that the DC scheme is effective. Usually, researchers are focused on problems in which they have static environment. In the last decade, there is a growing interest to apply nature-inspired metaheuristics in optimization problems with dynamic environments. Usually, dynamic optimization problems (DOPs) are addressed using evolutionary algorithms. In this thesis, we apply several novel ACO algorithms in two routing DOPs. The proposed ACO algorithms are integrated with immigrants schemes in which immigrant ants are generated, either randomly or with the use of knowledge from previous environment(s), and replace other ants in the current population. The experimental results showed that each proposed algorithm performs better in different dynamic cases, and that they have better performance than other peer ACO algorithms in general. The existing benchmark generators for DOPs are developed for binary-encoded combinatorial problems. Since routing problems are usually permutation-encoded combinatorial problems, the dynamic environments used in the experiments are generated using a novel benchmark generator that converts a static problem instance to a dynamic one. The specific dynamic benchmark generator changes the fitness landscape of the problem, which causes the optimum to change in every environmental change. Furthermore in this thesis, another benchmark generator is proposed which moves the population to another location in the fitness landscape, instead of modifying it. In this way, the optimum is known and one can see how close to the optimum an algorithm performs during the environmental changes.

We use control theory to explore management of populations affected by disturbances and uncertainty. We consider five related topics. Chapter 2 uses linear programming to find optimal translocation strategies between wild and captive populations. To allow comparison of the solutions we classify the optimal strategy depending on which stage classes are kept in captivity. We find depending on species, that different stages are targeted when the resource available is limited. In Chapter 3 we use linear programming to create management strategies for an invading population affected by disturbance. For a sinusoidal disturbance, the final population with control is bounded between a transfer function approximation and a feedback control solution. Then we assume worst case disturbance, which creates a 2-player game. In this linear programming context then it is possible that minimax < maximin. Chapter 4 considers a 2-player linear-quadratic problem and introduces the use of disturbance attenuation into ecology. Disturbance attenuation shows how a disturbance is amplified or attenuated by the system. In Chapter 5 we consider an invading population, and we explore the effect that stochasticity has on the relationship between Allee effect and population inertia needed for successful invasion. We find that for small population densities, then demographic stochasticity dramatically reduces the likelihood of invasion and survival of the resident.

The methods for modelling uncertainty and assessing the risk of financial markets were placed under scrutiny after the 2008 crisis. To protect against the worst possible scenario, in a problem of asset allocation, robust optimisation is required. Still, within this framework, assumptions about the uncertainty set have to be made. In our work, we expand the possible options for describing uncertainty sets, through the use of copulas. Copulas are a useful tool for describing uncertainty because of the modelling flexibility that they provide. They are able to easily describe asymmetric dependence structures and tail risk. Both are vital for emulating the financial markets behaviour, during periods of extreme shocks and comovements. Also, copulas are associated with robust measures of dependence. We introduce copulas into the robust optimisation framework by following two different approaches. At first, we derive a Worst Case Conditional Value at Risk optimisation problem, in which the uncertainty set consists of a selection of copulas. We formulate the problem into a convex optimisation problem. The advantages of such a model are supported by numerical examples using real data. In the second approach, copulas are used as means for creating non-symmetric, convex uncertainty sets, in the form of domains. We present examples where these sets can be used in a robust optimisation problem.

Problems requiring combinatorial optimisation are routinely encountered in research and applied computing. Though polynomial-time algorithms are known for certain problems, for many practical problems, from mundane tasks in scheduling through to exotic tasks such as sequence alignment in bioinformatics, the only effective approach is to use heuristic methods. In contrast to complete strategies that locate globally optimal solutions through (in the worst case) the enumeration of all solutions to a problem, heuristics are based upon rules of thumb about specific problems, which guide the search down promising avenues. Work in the field of Operations Research has gone further, developing generic metaheuristics, abstract templates which may be adapted to tackle many different problems. Metaheuristic researchers have created a variety of algorithms, each with their own strengths and weaknesses, and development of metaheuristics now often tries to combine concepts from a number of existing strategies to balance the advantages of the originals, known as hybridisation. This interest in hybridisation has led to the creation of a number of frameworks in imperative languages to assist programmers in the rapid creation and experimentation upon the algorithms. However these existing frameworks have struggled to enable hybridisation of the two major classes of metaheuristic, point based and population based, while being large and complicated to use. This Thesis investigates a functional approach to hybridisation. Despite superficial analogies between hybridisation and function composition, there are substantial challenges: unlike global search methods that can be explained elegantly in terms of graph traversal, prior work on local search has struggled to articulate a common model, let alone one that can accommodate more esoteric optimisation techniques such as ant colony optimisation. At the same time, these implementations cannot ignore the fact that the development of these techniques is driven by large-scale problems, and computational efficiency cannot be ignored. Given this background, this Thesis makes three substantial contributions. It decomposes metaheuristic searchmethods into a set of finer-grained concepts and tools that can be reassembled to describe both standard search strategies and more specialised approaches. It resolves problems found in implementing these abstractions in the widely used language Haskell, developing a novel approach based on dataflow networks. The value of functional abstraction in the practice of metaheuristic development and tuning is demonstrated through case studies, including a substantial problem in bioinformatics.

Cooperative search provides a class of strategies to design more effective search methodologies by combining (meta-) heuristics for solving combinatorial optimisation problems. This area has been little explored in operational research. This thesis proposes a general agent-based distributed framework where each agent implements a (meta-) heuristic. An agent continuously adapts itself during the search process using a cooperation protocol based on reinforcement learning and pattern matching. Good patterns which make up improving solutions are identified and shared by the agents. A theoretical approach to the understanding of the potential of agent-based systems is also proposed. This agent-based system aims to raise the level of generality by providing a flexible framework to deal with a variety of different problem domains. The proposed framework so far has been tested on Permutation Flow-shop Scheduling, Travelling Salesman Problem and Nurse Rostering. These instances have yielded some promising results. As part of the nurse rostering work a novel approach to modelling fairer nurse rosters is proposed.

Critical Care Units (CCUs) are one of the most complex and expensive of all medical resources and hospital managers are challenged to meet the demand for critical care services with adequate capacity. The pressure on critical care beds is continuously increasing as new medical equipment provides the opportunity to save more patients lives. It is therefore crucial that beds are managed well and used efficiently. This thesis describes two major projects, the first undertaken in conjunction with the CCU at the University Hospital of Wales in Cardiff (UHW); and the second with two CCUs from the Aneurin Bevan Health Board. In the first project data has been analysed to determine the flow of patients through the Unit. Admissions to CCUs were categorised under two headings: emergency, and elective. The length of stay in the CCU is heavily dependent on the admission category. In this thesis, both computer simulation and theoretical queueing models have been considered, which show how improvements in bed management may be achieved by considering these two categories of patients separately. The vast majority of previous literature in this field is concerned only with steady-state conditions, whereas in reality the processes are time-dependent. This thesis goes some way to addressing this deficiency. The second project relates to work undertaken with managers from the Royal Gwent Hospital in Newport and at the Nevill Hall Hospital in Abergavenny. Data from both hospitals have been analysed to define arrival and service processes. A state-dependent theoretical queueing model has been considered which has been used to investigate the significance of combining the two units. The model has been also utilised to advise on the number of beds the new combined unit should have in order to satisfy targets quoted by the hospital managers. In the final part of the thesis, consideration has been given to the impact of collaboration, or lack thereof, between hospitals using a game theoretical approach. The effect of patient diversion has been studied. To formally investigate the impact of patients transfers, a Markov chain model of the two CCUs has been set-up, each admitting two arrival streams: namely, their own patients and transfers from other hospital. Four different models were considered and for each model the effect of targets, demand and capacity were studied. The efficiency of a system which degrades due to selfish behaviour of its agents has been measured in terms of Price of Anarchy.

The work presented in this thesis concerns the problem of dynamic vehicle routing. The motivation for this is the increasing demands on transportation services to deliver fast, efficient and reliable service. Systems are now needed for dispatching transportation requests that arrive dynamically throughout the scheduling horizon. Therefore the focus of this research is the dynamic pickup and delivery problem with time windows, where requests are not completely known in advance but become available during the scheduling horizon. All requests have to be satisfied by a given fleet of vehicles and each request has a pickup and delivery location, along with a time window at which services can take place. To solve the DPDPTW, our algorithm is embedded in a rolling horizon framework, thus allowing the problem to be viewed as a series of static sub-problems. This research begins by considering the static variant of the problem. Both heuristic and metaheuristic methods are applied and an analysis is performed across a range of well-known instances. Results competitive with the state of the art are obtained. For the dynamic problem, investigations are performed to identify how requests arriving dynamically should be incorporated into the solution. Varying degrees of urgency and proportions of dynamic requests have been examined. Further investigations look at improving the solutions over time and identifying appropriate improvement heuristics. Again competitive results are achieved across a range of instances from the literature. This continually increasing area of research covers many real-life problems such as a health courier service. Here, the problem consists of the pickup and delivery of mail, specimens and equipment between hospitals, GP surgeries and health centres. Final research applies our findings to a real-life example of this problem, both for static schedules and a real-time 24/7 service.

This thesis is concerned with a real-world multi-shift drayage problem at a large international port with multiple docks being operated simultaneously. Several important issues in the drayage problem are identified and a set covering model is developed based on a novel route representation. The model adopts an implicit solution representation to reduce the problem size and aims to find a set of vehicle routes with minimum total cost to deliver all commodities within their time windows. As accurate travel time prediction is necessary to construct the vehicle routes, a short-haul travel time prediction model and an algorithm using real-life GPS data are studied. The output of the prediction model can be used as an input for the set covering model. The set covering model for the multi-shift full truckload transportation problem can be directly solved by a commercial solver for small problems, but results in prohibitive computation time for even moderate-sized problems. In order to solve medium- and large-sized instances, we proposed a 3-stage hybrid solution method and applied it to solve real-life instances at a large international port in China. It was shown that the method is able to find solutions that are very close to the lower bounds. In addition, we also proposed a more efficient hybrid branch-and-price approach. Results show the method performed well and is more suited for solving real-life, large-sized drayage operation problems.

The use of optimisation within financial markets is rapidly increasing. There is a growing demand for a class of new and improved methods to accurately price financial options. Semi-infinite optimisation (SIO) has become a vivid research area in mathematical optimisation during the recent two decades. This is due to the fact that there are many new theoretical advances as well as a broad variety of real-life problems where this mathematical model can be applied. This thesis considers particular applications of SIO to finding bounds on the prices of basket options. Original results have been derived for: • Finding a lower bound on European basket call option prices. • Calculating a lower bound on European basket call option prices, incorporating bid-ask prices within the model. • Analysing price bounds on various types of American basket options. • Deriving an upper bound on the price of a discretely sampled arithmetic average Asian basket option. • Finding an upper bound on the price of an Asian basket call option, incorporating bid-ask prices. • Calculating an upper bound on the price of an Altiplano Mountain Range option. The models and results obtained in this thesis can be used in financial markets by investors, investment banks and hedge funds amongst others.

As the prominence of the service sector is increasing in developed nations, new and exciting opportunities are arising for operational researchers to develop and apply models which offer managers solutions to improve the quality of their services. The development of time-dependent stochastic models to analyse complex service systems and generate effective personnel schedules are key to this process, enabling organisations to strike a balance between the provision of a good quality service whilst avoiding unnecessary personnel costs. Specifically within the healthcare sector, there is a need to promote efficient management of an Emergency Medical Service (EMS), where the probability of survival is directly related to the speed of assistance. Motivated by case studies investigating the operation of the Welsh Ambulance Service Trust (WAST), this thesis aims to investigate how operational research (OR) techniques can be developed to analyse priority service systems subject to demand that is of an urgent nature, cannot be backlogged, is heavily time-dependent and highly variable. A workforce capacity planning tool is ultimately developed that integrates a combination of forecasting, queueing theory, stochastic modelling and optimisation techniques into a single spreadsheet model in order to predict future demand upon WAST, set staffing levels, and optimise shift schedules and rosters. The unique linking together of the techniques in a planning tool which further captures time-dependency and two priority classes enables this research to outperform previous approaches, which have generally only considered a single class of customer, or generated staffing recommendations using approximation methods that are only reliable under limited conditions. The research presented in this thesis is novel in several ways. Primarily, the first section considers the potential of a nonparametric modelling technique known as Singular Spectrum Analysis (SSA) to improve the accuracy of demand forecasts. Secondly, the main body of work is dedicated to adapting numerical queueing theory techniques to accurately model the behaviour of time-dependent multi-server priority systems across shift boundaries and evaluate the likelihood of excessive waits for service for two customer classes. The final section addresses how shifts can be optimally scheduled using heuristic search techniques. The main conclusions are that in addition to offering a more flexible approach, the forecasts generated by SSA compare favourably to those obtained using traditional methods, and both approximate and numerical modelling techniques may be duly extended to set staffing levels in complex priority systems.

The thesis is concerned with novel Nature-Inspired heuristics for the so called NP-hard problems of optimisation. A particular algorithm which has been recently introduced and shown to be effective in continuous optimisation is the Plant Propagation Algorithm or PPA. Here, we intend to extend it to cope with combinatorial optimisation. In order to show that our extension is viable and effective, we consider three types of problems which are good representatives of the whole topic. These are the Travelling Salesman Problem or TSP, the Knapsack Problem or KP and the scheduling problem of Berth Allocation as arises in container ports or BAP. Because PPA is a population-based search heuristic, we devote a chapter to the important issue of generating good and yet computationally relatively light initial populations of solutions to kick start the search process. In the case of the TSP we revisit and extend the Strip Algorithm (SA). We introduce the 2-Part SA and show that it is better than the classical SA. We also introduce new variants such as the Adaptive SA and the Spiral SA which cope with clustered cities and instances with cities concentrated around the center of the unit square, respectively. In the case of KP we adapt the Roulette Wheel selection approach to generate solutions to start with PPA. And in the case of BAP, we introduce a number of simple heuristics which consider a schedule as a flat box with one side being the processing time and the other the position of vessels on the wharf. The heuristics try to generate schedules by avoiding overlap as much as possible. All approaches and algorithms are implemented and tested against well established algorithms. The results are recorded and discussed extensively. The thesis ends with a conclusion and ideas for further research.

Operational Research and Optimization are fundamental disciplines which, for decades, provided the real-world with tools for solving practical problems. Many such problems arise in container ports. Container terminals are important assets in modern economies. They constitute an important means of distributing goods made overseas to domestic markets in most countries. They are expensive to build and difficult to operate. We describe here some of the main operations which are faced daily by decision makers at those facilities. Decision makers often use Operational Research and Optimization tools to run these operations effectively. In this thesis, we focus on seaside operations which can be divided into three main problems: 1- the Berth Allocation Problem (BAP), 2- the Quay Crane Assignment Problem (QCAP), 3- the Quay Crane Scheduling Problem (QCSP). Each one of the above is a complex optimization problem in its own right. However, solving them individually without the consideration of the others may lead to overall suboptimal solutions. For this reason we will investigate the pairwise combinations of these problems and their total integration In addition, several important factors that affected on the final solution. The main contributions of this study are modelling and solving of the: 1- Robust berth allocation problem (RBAP): a new efficient mathematical model is formulated and a hybrid algorithm based on Branch-and-Cut and the Genetic Algorithm is used to find optimal or near optimal solutions for large scale instances in reasonable time. 2- Quay crane assignment and quay crane scheduling problem (QCASP): a new mathematical model is built to simultaneously solve QCASP and a heuristic based on the Genetic Algorithm is developed to find solutions to realistic instances in reasonable time. 3- Berth allocation, quay crane assignment and quay crane scheduling problem (BACASP): an aggregate model for all three seaside operations is proposed and to solve realistic instances of the problem, an adapted variant of the Genetic Algorithm is implemented.

This thesis discusses queueing systems in which decisions are made when customers arrive, either by individual customers themselves or by a central controller. Decisions are made concerning whether or not customers should be admitted to the system (admission control) and, if they are to be admitted, where they should go to receive service (routing control). An important objective is to compare the effects of "selfish" decision-making, in which customers make decisions aimed solely at optimising their own outcomes, with those of "socially optimal" control policies, which optimise the economic performance of the system as a whole. The problems considered are intended to be quite general in nature, and the resulting findings are therefore broad in scope. Initially, M/M/1 queueing systems are considered, and the results presented establish novel connections between two distinct areas of the literature. Subsequently, a more complicated problem is considered, involving routing control in a system which consists of heterogeneous, multiple-server facilities arranged in parallel. It is shown that the multiple-facility system can be formulated mathematically as a Markov Decision Process (MDP), and this enables a fundamental relationship to be proved between individually optimal and socially optimal policies which is of great theoretical and practical importance. Structural properties of socially optimal policies are analysed rigorously, and it is found that 'simple' characterisations of socially optimal policies are usually unattainable in systems with heterogeneous facilities. Finally, the feasibility of finding 'near-optimal' policies for large scale systems by using heuristics and simulation-based methods is considered.

The principal aim of this PhD program is the development of an optimisation and risk based methodology for reliability and robustness predictions of packaged electronic components. Reliability based design optimisation involves the integration of reduced order modelling, risk analysis and optimisation. The increasing cost of physical prototyping and extensive qualification testing for reliability assessment is making virtual qualification a very attractive alternative for the electronics industry. Given the availability of low cost processing technology and advanced numerical techniques such as finite element analysis, design engineers can now undertake detailed calculations of physical phenomena in electronic packages such as temperature, electromagnetics, and stress. Physics of failure analysis can also be performed using the results from these detailed calculations to predict modes of failure and estimated lifetime of an electronic component. At present the majority of calculations performed using finite element techniques assume that the input parameters are single valued without any variation. Obviously this is not the case as variation in design variables (such as dimensions of the package, operating conditions, etc) can have statistical distributions. The research undertaken in this PhD resulted in the development of software libraries and a toolset which can be used in parallel with finite element analysis to assess the impact of design variable variations on the package reliability and robustness. This resulted in the development of the ROMARA software which now contains a number of best in class reduced order modelling techniques, optimisation algorithms, and stochastic risk assessment procedures. The software has been developed using the C# language and demonstrated for a number of case studies. The case study detailed in this thesis is related to a power electronics IGBT structure and demonstrates the technology for predicting the reliability and robustness of a wirebond interconnect structure that is subjected to electro-thermo-mechanical loads. The design variables investigated in this study included wire-loop ratio, current in the wire, and thickness of the silicon die each represented as input variables with normal distribution. In terms of reliability the damage variable under investigation was the plastic strain at the wire/aluminium pad interface. Using ANSYS for predicting the physics in the package we have demonstrated the ability of the ROMARA code to optimise the design of wirebond, in terms of minimising the induced damage. Other real cases have been investigated using the developed ROMARA software and these are reported in the public domain and briefly detailed in this thesis.

Hospitals in the UK are increasingly having to cancel a large proportion of elective operations due to the unavailability of beds on hospital wards for post-operative recovery. The availability of post-operative beds is therefore critical to the scheduling of surgical procedures and the throughput of patients in a hospital. The focus of this research is to investigate, via data-driven modelling, systematic reasons for the unavailability of beds and to demonstrate how the Master Surgery Schedule (MSS) can be constructed using Operational Research techniques to minimise the number of cancellations of elective operations. Statistical analysis of data provided by the University Hospital of Wales, Cardiff was performed, providing information on patient demand and length of stay distributions. A two-stage modelling process was developed to construct and simulate an MSS that minimises the number of cancellations. The first stage involves a novel set partitioning based optimisation model that incorporates operating room and bed constraints. The second stage simulates the resulting optimal schedule to provide measures on how well the schedule would perform if implemented. The results from this two-stage model provide insights into when best to schedule surgical specialties and how best the beds are distributed between wards. Two optimisation under uncertainty techniques are then employed to incorporate the uncertainty associated with the bed requirements into the optimisation process. A robust optimisation (RO) approach that uses protection functions in each bed constraint is developed. Investigations into varying levels of protection are performed in order to gain insight into the so called `price of robustness'. Results show that MSSs that are constructed from protecting more of the uncertainty result in fewer cancellations and a smaller probability of requiring more beds than are available. The deterministic optimisation model is then extended to become a scenario-based optimisation model in which more scenarios of bed requirement are incorporated into a single optimisation model. Results show that as more scenarios are included, a more robust schedule is generated and fewer cancellations are expected. Results from the different approaches are compared to assess the benefits of using RO techniques. Future research directions following from this work are discussed, including the construction of the MSS based on sub-specialties and investigation of different working practices within the case study hospital.

A modified version of the Intelligent Water Drops algorithm (MIWD) was developed then used to determine the most stable configurations of Lennard-Jones (LJ), Binary Lennard-Jones (BLJ) and Morse Clusters. The algorithm is unbiased in that it uses no a priori cluster geometry information or cluster seeds. Results for LJ clusters show that the algorithm is effective and efficient in rediscovering all clusters up to size N = 104 with better success rates specially on difficult clusters compared to previous best methodologies reported in literature. Results on more difficult systems, such as the Binary Lennard Jones clusters up to size 50 (with 5 different atomic size ratios) and Morse clusters up to size 60 (with 2 interparticle range potentials), also showed the ability of MIWD to handle more complex systems. MIWD was then applied to predict the most stable structures of Janus clusters up to size 50 and on size 100 using a LJ potential model with a modulated angular term suited for two-patched Janus particles. Results show that MIWD is able to find well-structured geometries of Janus clusters. It is believed that this has been the first time that a nature-inspired stochastic algorithm and a variant of the IWD algorithm has been applied to the configurational optimization of Janus clusters.