基于单神经元自适应PSD的α-β目标跟踪滤波算法

α-βTracking Filter Algorithm Based on Single Neuron Adaptive PSD

α-β跟踪滤波器具有结构简单、计算方便等优点,在工程中得到了广泛应用,但同时由于参数固定而无法适用于机动场景中的目标跟踪。针对该问题,文章将单神经元自适应PSD(Proportion Sum Differential)算法与α-β滤波相结合。改进后的α-β滤波算法利用单神经元PSD算法的自适应参数调整能力,根据系统跟踪误差实时调整滤波器参数,因而能够跟踪机动目标。仿真和实测数据处理结果表明,改进的α-β滤波算法能够有效应对目标机动,形成稳定航迹,跟踪性能优于Kalman滤波。

α-βtracking filter is widely used due to its simple structure and little calculation amount.However,it cannot be used to track maneuvering targets easily because of its fixed parameters.The single neuron adaptive proportion sum differential(PSD)algorithm andα-βtracking filter are combined in the improved filter.The single neuron PSD’s turning ability is used to adjust filter’s parameters adaptively.The adjustment is based on tracking errors.The improved filter can be used to track maneuvering targets.Simulation and experimental data processing results show that the newα-βtracking filter proposed can deal with maneuvering targets,form stable tracks.Its performance is better than the traditional Kalman filter.