论文部分内容阅读
针对不同阶数的Kalman滤波器具有不同的跟踪能力与跟踪效率之间存在的矛盾,设计了一种模糊自适应变维跟踪算法(FAVD)。该算法使用两级滤波器,根据目标机动性的变化,适当地调整滤波器的阶数,使跟踪结果快速收敛,很好地解决了矛盾。同时通过模糊推理机制,在线调节高阶滤波器的参数,使适用范围大大增强,提高自适应能力,从而使该算法可以采用较少的模型覆盖较多的目标运动模式,达到很好的跟踪滤波效果,计算量也会大大减小。通过对计算机仿真结果分析表明,提出的算法具有可靠、计算简便、快速等特点,模型滤波精度较高,并可实现实时跟踪预测,具有一定的理论价值和实用价值。
Aiming at the contradiction between the different tracking ability and tracking efficiency of Kalman filter with different orders, a fuzzy adaptive variable dimensional tracking algorithm (FAVD) is designed. The algorithm uses a two-stage filter, according to the change of the target mobility, properly adjust the order of the filter, so that the tracking results quickly converge, a good solution to the contradiction. At the same time, through the fuzzy inference mechanism, the parameters of the high-order filter can be adjusted on-line so that the scope of application is greatly enhanced and the self-adaptability is improved. Therefore, the algorithm can cover more target motion modes with fewer models and achieve good tracking filtering Effect, the amount of calculation will be greatly reduced. The results of computer simulation show that the proposed algorithm is reliable, easy to calculate, fast and so on. The model has high filtering accuracy and real-time tracking and prediction, which has a certain theoretical value and practical value.