Multi-target Tracking Control Of Sensor Network

Because of its tremendous application value,target tracking has attracted lots of researchers attention.Due to the complicated of external electromagnetic environment,the traditional single sensor filter algorithm can not adapt to the high precision working environment.In order to improve the estimation accuracy,it is important to study the sensor network that is composed of multiple sensors.Sensor network has lots of advantages such as the low cost,configuration easily and deployment flexibly.And now it has been applied in various fields such as target tracking,environmental monitoring,information collection and so on.The single target tracking problem has been widely researched,but the target that sensors track is not only a single target but more likely a multi-agent system,and perhaps there is interaction between targets.Because of those reasons,in this thesis,we will investigate mobile sensor network and event-triggered sensor network for multi-target tracking problem,respectively.Different from the single-target tracking problem,two-target tracking one has to consider the interaction between followers in different groups.We first investigate a new estimation and control strategy for two-target tracking mobile sensor networks.By this model all mobile sensors can split into two groups to track their own target and form a cohesive flock with their neighbors.Stability analysis is conducted based on cascading Lyapunov method and matrix theory.Furthermore,a sufficient condition for the convergence is given in the form of the boundary conditions of feedback gains.Real-time communication will lead to a certain degree of energy waste.In order to save system energy and reduce the requirements of communication equipment,event-trigger mechanism came into being.Event-triggered mechanism refers to sensor communication with its neighbors only when a certain condition(such as a certain threshold)is meeting.Then we combined the event-triggered strategy with Kalman-consensus filter for two-target tracking sensor networks.According to the mean-square analysis,a suboptimal Kalman gain matrix is derived and a suboptimal event-triggered distributed filter is obtained.Based on the proposed filter protocol,all sensors which only depend on its neighbors' information can track their corresponding targets.Utilizing Lyapunov method and matrix theory,some sufficient conditions are presented for ensuring the stability of the system.