Target Tracking Algorithm Based On Event-triggered Strategy For Sensor Networks

Wireless sensor networks(WSNs)integrating data acquisition,processing and transmission functions have the characteristics of large-scale,low-cost,self-organizing,strong reliability and high integration whose application has also gradually expanded from the original military field to many civilian fields including smart agriculture,smart city,environmental monitoring,traffic navigation,marine resource exploration and so on.Target positioning and tracking,as a key technology of WSNs,combines many disciplines such as sensing,network communication,information processing,probability statistics etc,and has broad application prospects in the field of intelligent monitoring.In order to meet the requirements of low power consumption and distribution of the system,the distributed energy-saving mechanism and target tracking algorithm of the sensor networks are comprehensively investigated for the following three different situations:When a single sensor node is observable and the topology is fixed,a fully distributed eventtriggered sensor node measuring/sleeping strategy to optimize performance metrics consisting of tracking error and energy consumption,and consensus Kalman filter algorithm based on the minimum trace principle are proposed for sensor networks with limited node measurement and communication capabilities,as well as the necessary and sufficient conditions for guaranteeing the convergence of the algorithm are given.Compared with the time-triggered mechanism,the superiority of the mechanism in measuring energy saving is proved theoretically and experimentally.A distributed energy-saving target tracking algorithm is proposed for sensor networks with limited sensing/communication when the single node is unobservable and the network topology changes.Considering the node measurement energy consumption,an event-triggered sensor node measurement strategy is proposed to achieve a better trade-off between estimation error and energy consumption of each node.A distributed information filter based on the consensus weight design of the minimum trace principle is designed.It is verified theoretically and experimentally that the proposed algorithm has better system performance index than the traditional time-triggered algorithm.It is also proved that when each sensor is unobservable and the network's sensing topology is time-varying,as long as there exists at least one collaboratively observable sensor in the network at each sampling instant,then the estimation error of the system will be bounded in the mean square sense under any intimal estimation.Furthermore,considering measurement energy consumption and communication energy consumption,an energy-saving target tracking algorithm based on event-triggered measurement and communication strategy is designed.It is theoretically proved that as long as there is at least one collaboratively observable node in the network at each sampling instant,the algorithm can guarantee that the system estimation error is bounded in the mean square sense.When there are multiple targets to be tracked in the monitoring area,the k-Means clustering algorithm is adopted to solve the data association problem,and then the distributed multi-target tracking algorithm based on event-triggered measurement and communication strategy is applied to the multi-target tracking problem.It is proved theoretically that the algorithm is better than the time-triggered algorithm under the condition of convergence of the clustering algorithm,and the system estimation error converges in the mean square sense under the condition that there are collaboratively observable nodes in the network.Finally,a simulation example is given to verify the effectiveness and feasibility of the algorithm: in addition to ensuring the tracking accuracy of each target,the energy consumption of the network is effectively reduced.