Actor Design For Wireless Sensor Networks Robot Pursuit Of Game Studies

With the development of communication technology, embedded technology, microelectronic systems and the sensor technology, the wireless sensor network which have the capacity of perceptual, computing and communications beginning to emerge, and its application received wide attention. At present, the object positioning and tracking in region become a research hotspot in this field, especially the robot PEG (Pursuit-Evasion Games) issue under the sensor networks controlled. This issue is how to guide one or a group of pursuer to capture one or a group of evader. As the classic game theory problem, PEG has broad application prospects, for example, it can provide solution for intercept missiles, two-plane fighting in military, it also can solve the problem of locomotives, ships encounter so as to avoid the occurrence of disasters in civilian, particularly in the transport sector.Under this background, this thesis discussed the strategy of single pursuer and single evader pursuit-evasion games under wireless sensor networks, adopted the Differential Games which is advanced by Dr. Isaacs to establishment model of optimal control law, and simulated it. In the selection of simulation platform, we used the methodology of Actor- oriented and heterogeneous in PtolemyⅡ. This design methodologies look upon wireless sensor networks system as a complete system, with an embedded systems perspective to divide wireless sensor networks system into some modules which have independent system behavior and uniformly modeling, covering the traditional short comings of single modeling for wireless sensor networks system. Therefore, it is significant to study the sensor networks system system-level design methodology, which will lay solid basis of heterogeneous environment system theory research. The main contributions as follows:Firstly, the structure of wireless sensor network system and sensor node for the PEG issues is built, and the method of Differential Games to solve PEG is researched, given the optimal control mathematical model; Secondly, the design methodology of system-level modeling and simulation in Ptolemy is researched; the concept of Actor, Heterogeneous, computing model and abstract semantics in Ptolemy are analysis; the model needed during simulating for wireless sensor network systems is analysis; Finally, established the model of two controlled robots in wireless sensor network for PEG under the guidance of the PtolemyⅡprinciple and gives the corresponding mechanisms and methods.Lastly, the obtained results are summarized as the conclusion of the thesis, and the topics for further research are also addressed.