基于交互多模型的测量噪声累积误差修正方法研究

摘要

针对传感器的测量噪声水平会随时间变化，很难用单个的测量噪声水平模型进行描述；提出利用多个描述目标不同测量噪声水平模型次优的交互多模型方法。与单个模型的跟踪方法相比，多模型能够自适应修改不同模型的比例权重，从而选择局部最优的组合去逼近真实值，整个过程不需要设计复杂的逻辑跳变，具有较强的实用性。最后，通过仿真进一步验证了该方法的有效性和可行性。%Since sensor measurement noise level changes with time,it is difficult to describe with a single measurement noise level model. The suboptimal interaction multi⁃model method of different measurement noise level models for multiple de⁃scription objects is proposed in this paper. Compared with the single model tracking method,multi⁃model method can modify the proportion weight of different models adaptively,so that local optimal combination is selected to approximate the true value. It is unnecessary to design complex logic jumping in the whole process because it has strong practicability. The validity and feasibili⁃ty of the proposed method was verified further by simulation.