A Novel Interacting T-S Fuzzy Multiple Model by Using UKF for Maneuvering Target Tracking

Dynamic models of maneuvering targets in nonlinear systems are usually difficult to be modeled, and inaccurate dynamic model will lead to the poor performance of tracking algorithm. For these problems, in this paper, a novel interacting Takagi-Sugeno (T-S) fuzzy multiple model maneuvering target tracking algorithm by using UKF for parameter identification is proposed (ITS-UKF). The ITS-UKF algorithm uses multiple semantic fuzzy sets to represent the target feature information, and construct a general T-S fuzzy semantic multiple model framework. In the T-S fuzzy semantic multiple model framework, the intersection degree between fuzzy sets is used to estimate the transition probabilities between different fuzzy rules; fuzzy C-regression model clustering (FCRM) is used to adaptively identify the premise parameters. Moreover, the UKF is also used to identify the consequent parameters to improve the performance for nonlinear system. Simulation results show that the performance of ITS-UKF is superior to IMM-EKF (interacting multiple model extended Kalman filter), IMM-UKF (interacting multiple model unscented Kalman filter), particularly, when the target is maneuvered or the model of the target is inaccurate, the ITS-UKF algorithm has better performance.