Passive sonar multiple target tracking with different resampling algorithms
Erhan Özer, Ali Onur Akar, Alı Köksal Hocaoǧlu · 2018
In addition to many new application areas that have entered our lives, the need for identification of nonlinear systems has been born. Especially in the estimation of dynamic systems which has Non-Gaussian Measurement Noise; Kalman, grid-based etc. algorithms do not provide an optimal solution. Particle filter can estimate dynamics of nonlinear systems more accurately and realistically without assuming measurement noise distribution. In this study, particle filter was used in the application of multiple target tracking in passive sonar with different resampling methods. In simulations, the performance differences of the Minimum Variance (MSV) method were analyzed based on other classical resampling methods for the problem of passive sonar multi-target tracking.