Abstract: An auxiliary incremental particle filter method is proposed, while the algorithm process is established. The method can integrate increment form into the auxiliary particle filtering. It can solve the problem of unsetting up the accurate measurement likelihood function due to the unknown system errors in the practical engineering. On the other hand, the advantage of the auxiliary particle filtering which can select the important density function effectively by using the latest observation information is remained. This method decreases the numbers of resample obviously, preserves the diversity of the particle effectively and improves the accuracy of nonlinear filtering. Simulation results show that the method of auxiliary incremental particle filter can reduce the errors of nonlinear filtering effectively and increase the filtering accuracy by 50% compared to classical methods.

Key words: incremental particle filter, auxiliary particle filter, auxiliary incremental particle filter, filtering accuracy

摘要： 提出辅助增量粒子滤波方法并给出其算法过程。该方法将增量形式融入辅助变量粒子滤波中，解决由于工程实际中量测可能存在未知系统误差导致无法精确建立量测似然函数的问题，另一方面，其又能保持辅助变量粒子滤波方法的优势，在选取重要性密度函数上有效利用最新观测的信息。该方法能减少重采样次数，较好保持粒子的多样性，使得非线性滤波的精度得以提高。仿真实验结果表明，辅助增量粒子滤波方法能有效减少非线性滤波问题的误差，相对经典滤波方法的滤波精度提高了50%。

关键词: 增量粒子滤波, 辅助变量粒子滤波, 辅助增量粒子滤波, 滤波精度