摘要
针对量测噪声协方差不能准确获知或随时间不断变化这一问题,提出一种基于二型模糊逻辑系统的自适应卡尔曼滤波算法。利用二型模糊逻辑对量测噪声进行自适应调整,使残差的理论协方差和实际协方差相一致,影响滤波的增益值,修正状态的估计值。为验证该算法的有效性,将其与传统卡尔曼滤波、一型模糊卡尔曼滤波两种算法进行对比,比较结果表明,二型模糊卡尔曼滤波算法能有效提高状态估计的准确度。
To tackle the problem of uncertain or changeable measurement noise with time, an adaptive Kalman filter algorithm based on type-2 fuzzy logic system was proposed. The measurement noise was adjusted adaptively via type-2 fuzzy logic to make sure the theoretical value of the innovation sequence covariance matched with its actual value. The adjustment changed the value of the filter gain, and the state estimate was corrected. The type-2 fuzzy logic based Kalman filter was compared with the tradi- tional Kalman filter and type-1 fuzzy Kalman filter to illustrate its effectiveness. The simulation results show the proposed algo- rithm can improve the accuracy of state estimation effectively.
出处
《计算机工程与设计》
北大核心
2015年第12期3269-3272,3278,共5页
Computer Engineering and Design
基金
国家自然科学基金项目(61363041)
江西省教育厅科技基金项目(GJJ11090)
