Дисертації з теми "Systems intelligence"

Interactions among agents, either direct or indirect through the environment in which they act, are fundamental for swarm intelligence to exist; however, there is a class of interactions, referred to as "interference", that actually blocks or hinders the agents' goal-seeking behavior. For example, competition for space may reduce the mobility of robots in a swarm robotics system, or misleading information may spread through the system in a particle swarm optimization algorithm. One of the most visible effects of interference in a swarm intelligence system is the reduction of its efficiency. In other words, interference increases the time required by the system to reach a desired state. Thus, interference is a fundamental problem which negatively affects the viability of the swarm intelligence approach for solving important, practical problems.

We propose a framework called "incremental social learning" (ISL) as a solution to the aforementioned problem. It consists of two elements: (i) a growing population of agents, and (ii) a social learning mechanism. Initially, a system under the control of ISL consists of a small population of agents. These agents interact with one another and with their environment for some time before new agents are added to the system according to a predefined schedule. When a new agent is about to be added, it learns socially from a subset of the agents that have been part of the system for some time, and that, as a consequence, may have gathered useful information. The implementation of the social learning mechanism is application-dependent, but the goal is to transfer knowledge from a set of experienced agents that are already in the environment to the newly added agent. The process continues until one of the following criteria is met: (i) the maximum number of agents is reached, (ii) the assigned task is finished, or (iii) the system performs as desired. Starting with a small number of agents reduces interference because it reduces the number of interactions within the system, and thus, fast progress toward the desired state may be achieved. By learning socially, newly added agents acquire knowledge about their environment without incurring the costs of acquiring that knowledge individually. As a result, ISL can make a swarm intelligence system reach a desired state more rapidly.

We have successfully applied ISL to two very different swarm intelligence systems. We applied ISL to particle swarm optimization algorithms. The results of this study demonstrate that ISL substantially improves the performance of these kinds of algorithms. In fact, two of the resulting algorithms are competitive with state-of-the-art algorithms in the field. The second system to which we applied ISL exploits a collective decision-making mechanism based on an opinion formation model. This mechanism is also one of the original contributions presented in this dissertation. A swarm robotics system under the control of the proposed mechanism allows robots to choose from a set of two actions the action that is fastest to execute. In this case, when only a small proportion of the swarm is able to concurrently execute the alternative actions, ISL substantially improves the system's performance.
Doctorat en Sciences de l'ingénieur
info:eu-repo/semantics/nonPublished

This thesis aims to evaluate four artificial neural network architectures, each of which implements the sensory-motor mapping in an embodied, situated, and autonomous agent set up to reach a goal area in one out of six systematically varied T-maze environments. In order to reach the goal the agent has to turn either to the left or to the right in each junction in the environment, depending on the placement of previously encountered light sources. The evaluation is broken down into (i) measuring the reliability of the agents' capacity to repeatedly reach the goal area, (ii) analyzing how the agents work, and (iii) comparing the results to related work on the problem.

Each T-maze constitutes an instance of a broad class of problems known as delayed response tasks, which are characterized by a significant (and typically varying) delay between a stimulus and the corresponding appropriate response. This thesis expands this notion to include, besides simple tasks, repeated and multiple delayed response tasks. In repeated tasks, the agent faces several stimulus-delay-response sequences after each other. In multiple tasks, the agent faces several stimuli before the delay and the corresponding appropriate responses. Even if simple at an abstract level, these tasks raise some of the fundamental issues within cognitive science and artificial intelligence such as whether or not an internal objective world model is necessary and/or suitable to achieve the appropriate behavior. For such reasons, these problems also constitute an interesting base for evaluating alternative ideas within these fields.

The work leads to several interesting insights. Firstly, purely reactive controllers (as represented by a feed-forward network) may be sufficient, in interaction with the environment, to solve both simple and repeated delayed response tasks. Secondly, an extended sequential cascaded network that selectively replaces its own sensory-motor mapping achieves significantly better performance than the other networks. This indicates that selective replacement of the sensory-motor mapping may be more powerful than both modulation (as represented by a simple recurrent network) and replacement in each step (as represented by a standard sequential cascaded network). Thirdly, this thesis demonstrates that even reactive controllers may contribute to behavior, which, from an observer's point of view, may seem to require an internal rational capacity, i.e. the ability to represent and explore alternatives internally.

This thesis studies the issues of building and managing future Internet of Things (IoT) systems. IoT systems consist of distributed components with services for sensing, processing, and controlling through devices deployed in our living environment as part of the global cyber-physical ecosystem.

Systems with perpetually running IoT devices may use a lot of energy. One challenge is implementing good management policies for energy saving. In addition, a large scale of devices may be deployed in wide geographical areas through low bandwidth wireless communication networks. This brings the challenge of congfiuring a large number of duplicated applications with low latency in a scalable manner. Finally, intelligent IoT applications, such as occupancy prediction and activity recognition, depend on analyzing user and event patterns from historical data. In order to achieve real-time interaction between humans and things, reliable yet real-time analytic support should be included to leverage the interplay and complementary roles of edge and cloud computing.

In this dissertation, I address the above issues from the service oriented point of view. Service oriented architecture (SOA) provides the integration and management flexibility using the abstraction of services deployed on devices. We have designed the WuKong IoT middleware to facilitate connectivity, deployment, and run-time management of IoT applications.

For energy efficient mapping, this thesis presents an energy saving methodology for co- locating several services on the same physical device in order to reduce the computing and communication energy. In a multi-hop network, the service co-location problem is formulated as a quadratic programming problem. I propose a reduction method that reduces it to the integer programming problem. In a single hop network, the service co-location problem can be modeled as the Maximum Weighted Independent Set (MWIS) problem. I design algorithm to transform a service flow to a co-location graph. Then, known heuristic algorithms to find the maximum independent set, which is the basis for making service co-location decisions, are applied to the co-location graph.

For low latency scalable deployment, I propose a region-based hierarchical management structure. A congestion zone that covers multiple regions is identified. The problem of deploying a large number of copies of a flow-based program (FBP) in a congestion zone is modeled as a network traffic congestion problem. Then, the problem of mapping in a congestion zone is modeled as an Integer Quadratic Constrained Programming (IQCP) problem, which is proved to be a NP-hard problem. Given that, an approximation algorithm based on LP relaxation and an efficient service relocating heuristic algorithm are designed for reducing the computation complexity. For each congestion zone, the algorithm will perform global optimized mapping for multiple regions, and then request multiple deployment delegators for reprogramming individual devices.

Finally, with the growing adoption of IoT applications, dedicated and single-purpose devices are giving way to smart, adaptive devices with rich capabilities using a platform or API, collecting and analyzing data, and making their own decisions. To facilitate building intelligent applications in IoT, I have implemented the edge framework for supporting reliable streaming analytics on edge devices. In addition, a progression framework is built to achieve the self-management capability of applications in IoT. A progressive architecture and a programming paradigm for bridging the service oriented application with the power of big data on the cloud are designed in the framework. In this thesis, I present the detailed design of the progression framework, which incorporates the above features for building scalable management of IoT systems through a flexible middleware.