Target Capturing Control In Wireless Sensor And Actor Networks

With rapid development of electronic communication, computing and automaticcontrol technologies, Wireless Sensor Networks (WSN), which is integrated with thecapacities of information sensing, processing and transmission, begins to emerge. WSNperforms the wireless networking through sensor node (SN), and then realizes theinformation collection and transparency for physical environment. However, in manyapplications, such as forest fire danger detection and earthquake disasters rescue，it notonly needs SN to sense the environment conditions, but also needs SN to transmitemergency information. Then, event can be timely managed by controlling actor nodes.According to this demand, Wireless Sensor and Actor Networks (WSAN), which hasdeveloped quickly in recent years, may provide essential support for monitoring theapplications in surveillance and controlling fields.Attracted by its tremendous potential, many researchers from the academic andindustry communities have already devoted huge efforts to the research of WSAN. As oneof the fundamental applications in WSAN, target capturing control has been a hot andchallenging research topic. Most of existing works aim at dealing with the coordinationrelationship between sensor-sensor and sensor-actor coordinations, and then investigatethe reliability of data transport. However, apart from the reliability of data transport, thecoordination between actors is also important to the target capturing process. In fact, thesuccess of the whole coordination for WSAN directly hinges on the target capturingcontrol by actors. Based on this, this thesis studies the researches in wireless and actornetwork and multi-agent systems (MAS), with the goal of investigating the targetcapturing control in WSAN. The main research works are summarized as follows:(1) For the event report transmission and multi-target capturing problems, a virtualagent-based WSAN model is proposed, and then a Ballooning protocol is designed whichis used to guarantee the event report can be transmitted to actor nodes within a boundedtime delay. Based on the above designs, a multi-target choosing strategy is provided, and aLeader-following based adaptive capturing algorithm is proposed such that the ability of implementing multi-target tasks for WSAN can be improved.(2) For the static and dynamic target capturing problems, a sliding mode variablestructure-based target capturing strategy is proposed. To save the communication energy inactor nodes during the coordination process, a Daisy-Chain formation algorithm isdesigned. Then, all actor nodes can successfully capture the different types of target whilemaintaining less communication cost.(3) For the target capturing problem in static obstacle environment, a recedinghorizon control-based target capturing strategy is proposed. Information on obstacles isincorporated online as a part of the cost function. Then, a receding control algorithm isproposed, such that actor nodes can adjust their positions according the sensed information.The stability is guaranteed by adding a terminal-state penalty to the cost function.(4) For the target capturing problem in dynamic obstacle environment, two kinds ofpotential function are presented by according to the state whether an actor is within thearea of its neighbors’ influence. Correspondingly, the control algorithm is provided, whoseaim is to make the actor nodes track a moving target, while avoiding collision with amoving obstacle and other actor nodes.(5) For the WSAN target capturing task in the master-slave teleoperation system, adynamic quantization strategy is provided, where the output signals of master actor nodeand slave actor nodes are quantized before being transmitted. Then, a novel master-slavetarget capturing protocol is designed, such that the master actor can reach to the formationcenter of slave actor nodes, while slave actor nodes can keep a desired formation controlaccording to the position of master actor node.