基于卡尔曼滤波的动目标预测

Research on prediction of moving targets with Kalman filtering method

对非机动目标可利用标准卡尔曼滤波算法对其运动状态进行预测,因为其运动数学模型精确可知.而当运动目标处于机动时,准确描述目标运动状态的数学模型难以建立,标准卡尔曼滤波算法难以进行对其状态预测.因此,文中采用将目标加速度作为虚拟噪声的自适应卡尔曼滤波算法,进行动目标运动状态的预估.仿真结果表明了该算法有效、可行,具有一定应用参考价值.

For non-mobile targets, their movements can be predicted by normal Kalman filtering, since their mathematic models are known. However, when the targets are in motion, their accurate mathematic models are difficult to be established, so the normal Kalman filtering is also difficult to predict a mobile target in this case. In this paper, an adaptive Kalman filtering method was used to estimate the acceleration and predict the motion of mobile targets, in which the acceleration of targets is taken as a virtual noise. Simulation results showed that the proposed method is very useful to predict the mobile targets.