Research On Task Allocation And PHD Filtering Algorithm For Multi Target Tracking Based On WSN

Because of the particularity of wireless sensor network(WSN),the research of multi target tracking based on it has been a hot topic in the field of target tracking.Especially,it is very important to study the task allocation and tracking algorithm based on WSN multi target tracking:In order to solve the problem of multi object tracking when the target is close to or to meet,the task assignment conflict of competition is proposed.In this chapter,we propose a cooperative task assignment algorithm based on Q learning.Algorithm in multi target encounter or similar moment,breaking the traditional N target must construct N cluster idea is selected to simultaneously detect multiple targets to node consists of a single cluster,responsible for multiple target tracking to meet state,at the same time using Q learning method,which combined the cluster the best opportunity the use of both the comprehensive performance index of energy consumption and the accuracy of the determination of cluster heads and cluster members.Finally,according to the features of object labels separate target information,task allocation of multi target tracking.The simulation results show that the proposed algorithm can optimize the task allocation of multi target tracking in the near or the encounter situation,and has the advantages of reducing the energy consumption of the system.Aiming at the unknown clutter intensity in multiple target tracking,a tracking algorithm based on entropy penalized EM(Expectation Maximization)of unknown clutter estimation is proposed.The clutter intensity is modeled by finite mixture model.And the entropy penalized factor is applied on mixing weight and the missing parameter,with the appropriate dynamic accommodation coefficient attached,which accelerates the extinction of low weight mixing components and also decreases the times of iterations.And the algorithm is not sensitive to initial parameters.Simulation results show that the algorithm has the advantages of high precision and stable tracking,which improves the performance of probability hypothesis density filter in multiple target tracking.For the expansion of target tracking in unknown clutter intensity in the research and development of target estimation algorithm is proposed for clutter probability hypothesis density.In this algorithm,the finite mixture model is used to estimate the probability density of clutter.And it uses the Dirichlet distribution to estimate the mixed weights.Simulation results show that the algorithm can effectively perform the extended target tracking in clutter environment.A target tracking simulation platform based on WSN is constructed,including the six-element array ultrasonic sensor network and the monitoring software platform.Asynchronous sensor data acquisition and wireless network data transmission are realized by the former part.The target tracking trajectory can be calculated with the acquisition data and the performance of the target tracking algorithm under different parameters can be studied by the latter part.The test results show that the platform can achieve the goal of target tracking.