基于自适应UKF微型航姿系统噪声在线估计  

Online noise estimation of mini-AHRS based on adaptive UKF algorithm

作　　者：刘宇[1] 刘琼[1] 周帆[1] 李云梅[1] 向高林[1] 

机构地区：[1]重庆邮电大学光电信息感测与传输技术重庆市重点实验室,重庆400065

出　　处：《重庆邮电大学学报（自然科学版）》2016年第3期285-290,311,共7页Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)

基　　金：国家自然科学基金(51175535);MEMS振动传感与微姿态组合测井技术国际联合研究中心科技平台与基地建设(cstc2014gjhz0038)~~

摘　　要：针对先验噪声与系统真实噪声不符引起标准无迹卡尔曼(unscented Kalman filter,UKF)性能退化的情况,提出一种应用于非线性时变状态和参数联合估计的自适应UKF(adaptive unscented Kalman filter,AUKF)算法。根据新的协方差矩阵与相应估计值之间存在的误差,构建成本函数。采用梯度下降法进行在线预估,对噪声的协方差进行在线更新并反馈给标准的UKF。实验和仿真分析表明,与标准UKF相比,自适应UKF算法在精度上有较大的提高。对于时变噪声协方差不确定时,自适应UKF噪声在线估计的鲁棒性得到明显改善,验证了自适应UKF噪声在线估计模型的准确性和可行性。Considering that the prior noise of a normal unscented Kalman filter does not agree with its daptive unscented Kalman filter algorithm applied to nonlinear joint estimation of both time-varying stateproposed. Firstly , a cost function is built based on the error betveen the covariance matrsponding estimations. Then the gradient descent method for online forecast is used. Finally , the noise covariance is onlineupdated , the updated covariance feedback to the standard UKF. Experimental and simulation analysis indicates that adaptive UKF provides higher estimation precision than the nomal UKF algorithm. For time-varying noise covadaptive UKF online noise estimation robustness is improved significantly , ad the accuracy and feasibility of onlinUKF noise estimation model is verified.

关 键 词：无迹卡尔曼 自适应UKF 联合估计 成本函数 梯度下降算法 鲁棒性 

分 类 号：TN212[电子电信—物理电子学]