Particle Filter Based on Cuckoo Search for Non-linear State Estimation

the aim of this paper is to propose an algorithm for particle filter which will overcome its problem of particle impoverishment. Our approach embed cuckoo search via levy flight algorithm into standard particle filter for Non-linear and Non-Gaussian state estimation. The use of cuckoo search via levy flight optimization overcomes the problem of particle impoverishment which is generated during resampling. To validate the efficacy of the proposed algorithm, its performance is compared with the particle filter and PSO Particle Filter (PSO-PF). Simulation results for generic one dimensional problem and two dimensional classic bearing only tracking problem show that our novel Cuckoo-PF outperforms other algorithms when RMSE, robustness and sample impoverishment are considered as metric for performance measurement.