Research On Immunity Particle Swarm Optimization Particle Filter Algorithm And Its Hardware Implementation

In nonlinear filtering ,particle filter algorithm has became more and more popular after algorithms of Extended Kalman Filter and Unscented Kalman Filter which is widely used in the fields of objective tracking ,navigation guidance and control, imagine processing and malfunction detecting. Particle filter is a kind of mixture of Monte Carlo estimation theory and technology of Bayes filtering which ia making use of a swarm of particles transported in possible areas to estimate the function of prior density. In essential, it fits for any non-Gauss and nonlinear situations. But particle filter algorithm has its own drawbacks such as particle degeneration, impoverishment and slow velocity in calculating.The inherence drawbacks of particle filter algorithm are taken into consideration in this paper with improved particle filter algorithm and hardware accomplishment as paper research objectives. The main research work is as following:1 ,Taking particle degeneration and sample impoverishment into account, proposes immunity particle swarm optimization particle filter algorithm (IMPSOPF) which can deal with both particle degeneration problem and sample impoverishment. Simulation in the fields of state estimating and objective tracking show its effectiveness.2, Taking particle filter algorithm slow velocity of calculating into account, on the base of method of MATLAB aiding ISE designing; we decompose the particle filter algorithm into some modules and then accomplish function of algorithm in hardware of XILINX V5 FPGA serial. Simulation results show algorithm accomplished in FPGA is both arising velocity and holding on the accuracy which would be the base of accomplishing the new particle filter algorithm in the fields of objective tracking, navigation guidance and so on.