Target Tracking For Mobile Sensor Networks Based On Flocking Control

With the rapid development of micro-electro-mechanical technology, mobile sensor networks(MSN) are beginning to be widely applied in modern production and life. In dynamic dangerousenvironment, MSN can track a target and accomplish missions replacing human workers. It ispractically a significant research topic to design flocking algorithms of mobile sensor networks foravoiding collisions and protecting mobile sensors, as well as target tracking.The main contributions of this paper are described as follows:Firstly, on condition that relative location and velocity measurement of the target are availablefor every node in the MSN, a flocking algorithm is designed to regulate mobile sensors to track thetarget and avoid collisions. The terms for collision avoidance in the algorithm are applied in everysensor, which are derived from the field produced by a artificial potential function.Secondly, for the purpose of lowering the control cost a pinning control algorithm is designedto regulate the MSN to track the target. When location and velocity measurements of the target areavailable for the part of mobile sensor nodes, the designed control law coordinates the MSN to trackthe target relying on links among nodes.At last，it should be considered that velocity transducers are not used in some practical systems.It is valuable to investigate tracking algorithms with only location measurements. The solution is todesign a flocking algorithm without relative velocity measurements based on a passivity approachestimating the velocity information. The available location measurements and estimated the velocityvalues are used for the control laws of mobile sensor nodes regulating whole MSN to track thetarget.Quality of monitoring (QoM) of the network is an important performance index. Then, aevaluation function for quality of monitoring is designed based on the distance between the mobilesensors and the target.Sufficient conditions for the stability of these algorithms are obtained through Lyapunovstability approach and then some numerical simulations demonstrates the effectiveness of thealgorithms...