Abstract:  An improved Gaussian mixture particle filter algorithm is proposed for the highly non-linear bearing-only tracking system where the common tracking filters often fail to catch and keep tracking of the emitter. In the algorithm, based on the characteristics of SPKF and particle filter，the limited Gaussian mixture model is used to approximate the posterior density of states，system noise and measurement noise. The greedy EM is used to obtain the reduced order  model and overcome such disadvantages of the standard EM as the number of the mixture components is assumed a known a priori， the performance of the overall parameter estimation process depends on the given good initial settings， and the estimated parameter can be resulted from some local optimum points, thus lessening effects caused by sampling depletion. Simulation results show that the algorithm outperforms the one based on PF and the one based on EM-GMPF in tracking accuracy and stability. Therefore it is more suitable to the nonlinear state estimation.

Key words: Passive sensor, Greedy expectation maximization (EM) algorithm, Particle filter (PF), Gaussian mixture modeling, Model order reduction