The Application of Support Vector Regression in Particle Filtering

In this paper, we propose a resampling method for particle filtering (PF) based on the support vector regression (SVR). The SVR is introduced to fit the posteriori probability density function of the current state in the procedure of the filtering. Particles was resetting in the high-probability region, and the weights of these particles are calculated according the fit function. Simulation shows that the proposed resampling method is effective for non-linear filter problem.