Research On Moving Target Tracking Algorithm Based On Wireless Sensor Technology

Wireless sensor network is an integrated measurement and control network what is acollection of electronic, computer technology. What is more, it involves the discipline ofautomatic control, computer technology, communication, artificial intelligence and so on,what is highly practical. Target tracking technology based on wireless sensor network hasbeen a hot research topic in recent years that has very important applications in military andcivilian. Tracking technology based on wireless sensor network can not only be used inmissile systems, air defense, coastal defense and combat surveillance in the military, but alsobe used for medical testing, smart home, urban transport, research on animal migration incivilian. Thus, the research on target tracking algorithm based on wireless sensor network hasimportant academic value and practical significance.This article introduces the algorithms used frequently in the moving target trackingtechnology based on wireless sensor network and analyzes the extended kalman filteralgorithm, the unscented kalman filter and the standard particle filter algorithm. Then, thearticle establishes a nonlinear model of the movement. After simulation, we understandparticle filter has many prominent advantages in tracking of moving targets. However, it hasthe defects of the particle sample dilution in essence. For the defect, two optimizationalgorithms are proposed: the first one is a hybrid algorithm of the trilateration method,weights sorting and particle filter, which introduces trilateration method and weight valuessort thinking to optimize the process of particle filter resampling process. The new algorithmabolishes the assumptions of the initial moment, and uses trilateration method to locate thetarget. What is more, the selection of the weight values update the density function, and thefollow-up steps ensure that all particles join in the particles update to improve the diversity ofthe particles, and it solves the particle sample dilution phenomenon; The second one is ahybrid algorithm of the QPSO and particle filter, and the introduction of the quantumbehavior of particle swarm optimization (QPSO) algorithm is in order to find a global optimalparticle. The weight value of the particle closing to the true state will increase so as to solvethe poor particle problem of particle filter, and the number of particles required in the particlefilter can be reduced.The article had simulated in the MATLAB software platform for analysis of the standardparticle filter algorithm and improved particle filter algorithm in moving target tracking basedon wireless sensor network. The experiments show that the two improved particle filteringalgorithm can effectively improve tracking accuracy and the stability of the tracking process.