【摘要】 采用参数自适应扩展卡尔曼滤波,建立了隧道周边位移预测模型。基于预测残差特性,评价滤波的敛散性。对模型参数进行实时修正,抑制滤波发散,保证了滤波精度。通过滤波预测值与实测值对比,判断预测模型的适用性;比较标准型与参数自适应型的预测残差,以判断何种预测方法的更优。分析结果表明:参数自适应型能有效抑制滤波发散,较为适用,预测效果较好。更多还原

【Abstract】 Base on parameter adaptive extended Kalman filter(EKF),aprediction model of tunnel peripheral displacement is built.In order to prevent filter form divergency,a way to judge the convergency and divergency of the filter by residual error’speciality factor is proposed so as to adjust the model parameter,and ensure the precision.Judging The feasibility of the predict model is judged based on the comparison between prediction and observation.A comparison for predicting residual error between standard EKF and parameter adaptive EKF is presented to evaluate which method is better.Results indicate that:Parameter adaptive EKF of more feasibility can effectively prevent filter from the divergency,and predict well.更多还原