论文部分内容阅读
本文提出了一种新的机动目标跟踪滤波的模型方法。它将目标的机动加速度作为一状态变量引入模型而直接进行估计,并通过卡尔曼滤波器的残差来检测目标机动与否。一旦检测出目标机动,马上重新启动卡尔曼滤波器以适应机动加速度的跳变。新的自适应滤波方法在这种情况下实现了最佳滤波。计算机仿真结果表明,在计算量远少于Moose方法计算量的情况下,本文方法的滤波精度与Moose方法的滤波精度相当。
This paper presents a new maneuvering target tracking filter model method. It introduces the target maneuvering acceleration as a state variable into the model and directly estimates the target maneuver by using Kalman filter residuals. Once the target maneuver is detected, the Kalman filter is restarted immediately to accommodate the maneuvering jump. The new adaptive filtering method in this case to achieve the best filter. Computer simulation results show that the filtering accuracy of this method is equivalent to that of the Moose method when the computational load is far less than that of the Moose method.