Research On Adaptive Algorithms For Highly Maneuvering Target Tracking

Maneuvering target tracking has always been a key and difficult issue in the military field.Particularly,in recent years,the world powers led by the United States and Russia are making every effort to promote the development of hypersonic flight weaponization and gradually form an arms race.Once the formation of combat capability will affect the pattern of world military power,China’s national defense will also face a serious threat.Therefore,the research on high speed and high maneuvering target tracking algorithm is particularly urgent and important.High-speed and high-mobility targets make tracking filter modeling difficult because of its high speed,maneuvering speed and unpredictability.This paper introduces several typical maneuvering target tracking models and filtering algorithms,and compares the tracking performance of different tracking algorithms.The factors affecting the tracking filtering performance are analyzed,which leads to the necessity of adaptive tracking filtering algorithm research.Based on the "current" statistical model,this paper proposes an improved single model adaptive tracking algorithm,which improves the tracking performance degradation caused by the poor adaptability of fixed model parameters.Based on the interactive multi-model,an adaptive grid interaction multi-model(AGIMM)tracking algorithm is introduced,which has better adaptive ability than the fixed model interactive multi-model algorithm.Aiming at the problem that AGIMM algorithm relies on central grid model and slow convergence of model set,a variable structure interactive multi-model based on maneuver discrimination is proposed.This algorithm overcomes the shortcomings of AGIMM algorithm and improves the tracking performance of maneuvering targets.Finally,the motion trajectory of HTV2 aircraft is optimized by using the space hypersonic vehicle as the application background.Several adaptive tracking algorithms are applied to the simulation tracking of the trajectory,which further validates the effectiveness of the proposed tracking algorithm.