基于IMM的光电经纬仪机动目标跟踪优化算法

An Optimization Algorithm for Improving Markedly the Tracking of a Maneuvering Target

提出一种简化交互多模型算法(IMM)与去偏转换测量卡尔曼滤波算法(CMKF-D)相结合的机动目标跟踪优化算法。该算法通过机动检测判别函数D(k)与门限T的关系,自适应调整CMKF-D的部分参数,解决了常规交互多模算法需要大量先验知识的问题,同时克服了扩展卡尔曼滤波(EKF)对非线性模型线性化所引入的误差。仿真实验验证了该算法的有效性。利用该算法可显著改善基于非线性测量方程下光电经纬仪对机动目标的跟踪性能,其位置跟踪误差小于1.5m,速度误差小于1.5 m/s,加速度误差小于0.7 m/s2。

Existing algorithms using Interaction Multiple Model（IMM） and Extended Kalman Filter（EKF） for tracking maneuvering targets suffer,in our opinion,from two shortcomings：（1） traditional IMM algorithm requires a large amount of a priori knowledge,（2） error is unavoidable when a linear model is substituted for a non-linear model through EKF.Our efficient algorithm can suppress to a considerable extent these two shortcomings.In the full paper,we explain our algorithm in some detail;in this abstract,we just add some pertinent remarks to listing the three topics of explanation.The first topic is： Debiased Converted Measurements Kalman Filter（CMKF-D）.The second topic is： simplified IMM algorithm. In these two topics,we discuss how our algorithm adaptively adjusts part of the CMKF-D parameters according to the relations between maneuvering discriminant function D（k） and threshold τ,summarized in Eqs.（15） and（16） in the full paper.At the end of the paper,we simulate our algorithm using a photoelectrical theodolite to track a maneuvering target.The simulation results,given in 6 figures in the full paper,show preliminarily that our algorithm can significantly improve the tracking of a maneuvering target.The position-tracking error is less than 1.5m,the velocity-tracking error is less than 1.5m/s and the acceleration-tracking error is less than 0.7 m/s^2.