针对运动目标变维姿态滤波在姿态剧烈变化时滤波效果较差的问题，结合Jerk模型，提出一种改进的变维运动目标姿态滤波方法。该方法通过建立角加加速度滤波模型，结合变维姿态滤波中的角速度和角加速度模型，利用角度和角速度的新息残差，角加速度和角加加速度估计显著性检验的统计量，判定目标姿态的机动状态，调整跟踪模型，从而提高姿态跟踪精度。仿真表明，该方法在保持变维姿态滤波精度基础上，有效提高了姿态强机动情况下的跟踪精度。

In attitude tracking of target based on variable dimension,the filtering effect is not so good when the attitude changes highly.To solve this problem,an improved variable dimension attitude filter method was proposed based on Jerk model.The angular Jerk filtering model was established.Taking the angular velocity and angular acceleration models in variable dimension filter into account,the maneuvering state of target attitude was estimated based on innovation residual sequences of attitude,angular velocity,and statistics for significance test of angular acceleration and angular Jerk.And the tracking model was adjusted,so as to improve the attitude filtering precision.The simulation result showed that the method can effectively improve the attitude tracking precision as the attitude changes highly while keeping the precision of variable dimension attitude filtering.