【摘要】 为了实时准确地获取爬架运行时的姿态信息，提出了一种基于改进Sage-Husa扩展卡尔曼滤波算法(improved Sage-Husa extended Kalman filter, ISHEKF)的爬架姿态估计方法。首先建立了爬架姿态估计模型，然后在扩展卡尔曼滤波(extended Kalman filter, EKF)对传感器进行融合滤波时，加入Sage-Husa扩展卡尔曼算法(SHEKF)对时变噪声进行调节，再以协方差匹配技术对Sage-Husa进行滤波发散判定，通过在调节因子中引入爬架实时运动速度，改进滤波发散判定依据，从而满足爬架运行时高精度的滤波要求。实验结果表明：在静态实验中，以横滚角为例，ISHEKF的最大误差较SHEKF减少了21.9%,较EKF的最大误差减少了70.8%;在动态实验中，ISHEKF表现出更好的稳定性和滤波精度，能够准确反映爬架运行的状态变化。更多还原

【Abstract】 In order to obtain real time and accurate attitude information of climbing frame, an attitude estimation method based on improved Sage-Husa extended Kalman filter(ISHEKF) algorithm was proposed. Firstly, the attitude estimation model of the climbing frame was established. Secondly, when the sensor was fusing with the extended Kalman filter(EKF), the Sage-Husa extended Kalman algorithm(SHEKF) was added to adjust the time-varying noise. Thirdly, the Sage-Husa filtering divergence was determined by the covariance matching technology. Finally, by introducing the real time motion speed of the climbing frame into the adjusting factor, the criterion of filtering divergence was improved, so as to meet the high precision filtering requirements of the running of the climbing frame. The experimental results show that in static experiments, taking the roll angle as an example, the maximum error of ISHEKF is reduced by 21.9% compared to SHEKF, and 70.8% compared to EKF. In dynamic experiments, ISHEKF exhibits better stability and filtering accuracy, which can accurately reflect the changes in the operation status of the climbing frame.更多还原