对地面机动目标的跟踪问题是目标跟踪领域的难点,多传感器数据融合技术以及变结构交互多模型算法是解决地面目标跟踪问题的有效途径。针对目标运动特点,将地面目标运动特性和运动状态作为先验信息,设计了一种地面目标跟踪算法,给出了算法的详细步骤,算法主要采用了加权观测融合卡尔曼滤波器及采用模型组切换方法的变结构多模型算法实现。仿真及误差分析结果表明,该算法优于传统交互多模型算法。为地面目标跟踪算法的设计提供了参考。

The problem of ground target tracking is one of the difficulties in target tracking fields. Multi-sensor data fusion technology and variable structure interacting multiple model algorithms are effective ways to solve this problem. Aiming at the moving character of targets, an algorithm for ground target tracking which takes the moving character and state as the given information was designed. The algorithm includes the measure fusion Kalman filter and model group switching algorithm for variable structure interacting multiple model algorithms, the detailed process was given out. The simulation results show that this algorithm is better than the IMM algorithm. This has provided reference for design of algorithms for ground target tracking.