Radar Maneuvering Target Tracking Based On Fuzzy Inference

With the continuous development of modern navigation,the mobility of the target is getting higher and higher,which seriously affects the safety of water traffic.Fuzzy inference has a thinking similar to that of human brain and can be used to predict and estimate the maneuvering state of targets in order to improve the tracking performance of maneuvering targets.Therefore,this paper studies the maneuvering target tracking algorithm based on fuzzy inference.Firstly,discussed the basic principle of maneuvering target tracking and the basic method of maneuvering target tracking based on fuzzy inference.The common models of target maneuvering are analyzed,the algorithms of interactive multi-model and fuzzy interactive multi-model are discussed,and the characteristics of several common motion models are simulated and demonstrated,which establishes the foundation for the subsequent research of maneuvering target tracking algorithms.Secondly,this paper proposes an improved target tracking algorithm of the"current" statistical model based on fuzzy inference(IFI-CS).In practical application,in order to adapt to the high maneuvering change of the target and make the tracking of the maneuvering target more fit to the reality,the "current" statistical model filtering algorithm is discussed.Through analysis,it is found that the acceleration limit value cannot be changed adaptively in the tracking of the maneuvering target,and then the current statistical model filtering algorithm based on fuzzy inference is discussed.Using fuzzy inference system to adjust the limit value of acceleration adaptively,But when the target mutates,The tracking error is large,Therefore,this paper proposes an improved "current" statistical model based on fuzzy inference(IFI-CS),The fuzzy membership function withoutmutation is used to enhance the tracking of weak maneuvering targets,A strong tracking filter is introduced to enhance the tracking of targets with sudden changes in motion state.Finally,the experimental simulation proves that the improved algorithm in this paper has improved the tracking performance of maneuvering targets compared with the"current" statistical model filtering algorithm(CS)and the "current" statistical model filtering algorithm based on fuzzy inference(FI-CS)Finally,this paper proposes an improved fuzzy interactive multi-model algorithm(IF-IMM).Due to the time variability and complexity of target maneuver,It is difficult to achieve accurate tracking using only a single model,This paper discusses the Interactive Multiple Model Algorithm(IMM),Because IMM chooses a large number of model sets,which leads to the disadvantages of increased computation and inaccurate tracking,Furthermore,the fuzzy interactive multiple model algorithm(F-IMM)is discussed,On this basis,An improved fuzzy interactive multiple model algorithm(IF-IMM)is proposed,The model covariance matrix is taken as the input of fuzzy inference system,The membership function of the ascending half trapezoid is used to leave the useful models in the model set.Eliminating useless models can reduce the probability of competing among model sets to update models.Finally,through simulation experiment,interactive multiple model algorithm(IMM),fuzzy interactive multiple model algorithm(IF-IMM)and the improved algorithm in this paper are compared and analyzed to verify the effectiveness and superiority of the algorithm in this paper.