Performance Test of Interacting Multiple Model Particle Filter for Maneuvering Target Location

During the previous work of this paper, the change of a maneuvering target state and the nonlinear of maneuvering target system model are considered, and IMMPF (Interacting Multiple Model Particle Filter) is used to state estimation of the maneuvering target. Then some simulations have been done. The results showed that the idea is feasible, and the single GPS (Global Positioning System) location precision is improved with IMMPF. In addition, the location precision with IMPPF is better than that with the traditional IMMEKF (Interacting Multiple Model Extended Kalman Filter). Also, the method can work with other sensors data of the maneuvering target. In this paper, a real test of this method is proposed. A vehicle with GPS receiver runs around the sports filed several times. Constant velocity model, and coordinated turn model are used. The state estimation calculated by IMMPF with single GPS data is evaluated against DGPS (Differential GPS). The experiment results proved the feasibility in real application, and the location precision of maneuvering target with single GPS is improved by IMMPF.