一种限定记忆的自适应扩展Kalman滤波器.pdfVIP

第 36 卷第 4 期 Vol.36 No.4 井冈山大学学报( 自然科学版) 2015 年 7 月 Jul. 2015 Journal of Jinggangshan University (Natural Science) 49 井冈山大学学报( 自然科学版) 49 文章编号：1674-8085(2015)04-0049-06 一种限定记忆的自适应扩展Kalman 滤波器 朱文超 ( 中国电子科技集团第三十八研究所，安徽，合肥 230041) 摘 要：为解决扩展卡尔曼滤波器（EKF ）鲁棒性差，且无法实时精确跟踪系统突变状态的问题，研究一种基于 限定记忆滤波的自适应 EKF 算法。算法将 EKF 与限定记忆滤波器相融合，减小旧量测数据对滤波效果的影响， 提高估计精度；引入自适应因子与渐消因子，通过实时调节新旧滤波增益阵以及预测状态值，精确地跟踪系统突 变状态。仿真实例表明，强跟踪算法与经典 EKF 算法相比，自适应 EKF 算法鲁棒性好，滤波精度高，能够有效 地跟踪系统突变状态。 关键词：扩展卡尔曼滤波；限定记忆滤波；旧量测数据；自适应算法；系统突变状态 中图分类号：TP391 文献标识码：A DOI:10.3969/j.issn.1674-8085.2015.04.009 ADAPTIVE EXTENDED KALMAN FILTERING BASED ON LIMIT MEMORY ALGORITHM ZHU Wen-chao (The 38th research institute of China electronics technology group corporation, Hefei, Anhui 230041, China) Abstract: In order to solve the problem that Extended Kalman filtering (EKF) cannot tracking mutant system state accuracy and improve the robustness of the filtering. A new adaptive Kalman filtering (AEKF) based on limited memory has been proposed. This algorithm combines EKF with Limit memory filtering that minimizes the influence of the old measurements. Forgetting factor and weakening factor can track mutant state accuracy by the technology of dynamically adjusting the weight of state prediction and Kalman gains in the filter estimation. The results of simulation experiments demonstrate that in comparison with Strong Tracking Filtering and Extended Kalman Filtering, AEKF provides higher estimated accuracy and better .robustness to track mutant state. Key words: extended Kalman filtering; limit memory