摘要
非线性滞回模型具有强烈的非线性特征,瞬时输入与加载历史路径影响模型参数的在线识别精度。以非线性Bouc-Wen模型为例,采用容积卡尔曼滤波器与隐性卡尔曼滤波器分别识别模型参数,对比分析两种滤波器的识别精度和计算效率。结果表明,容积卡尔曼滤波器识别Bouc-Wen模型参数k、γ、n的相对误差分别比隐性卡尔曼滤波器降低了0.73%、1.95%和2.10%,计算时间相比隐性卡尔曼滤波器减少了0.95 s。在模型参数的识别精度和计算效率上容积卡尔曼滤波器优于隐性卡尔曼滤波器,适用于非线性模型的在线参数识别。
This paper is based on the insight that nonlinear hysteretic model has strong nonlinear characteristics,and transient input and loading history path influence the on-line identification accuracy of model parameters.The study involves identifying the model parameters by taking the nonlinear Bouc-Wen model as an example and using the cubature Kalman filter and the unscented Kalman filter;and comparing and analyzing the recognition accuracy and computational efficiency of the two filters.The results show that the cubature Kalman filter identifies the parameters k、γand n of the Bouc-Wen model,with lower relative errors of 0.73%,1.95%and 2.10%and uses a 0.95 less calculation time than the unscented Kalman filter.The cubature Kalman filter superior to the unscented Kalman filter in the accuracy and computational efficiency of model parameters identification works better for the on-line parameter identification of nonlinear models.
作者
王涛
李勐
孟丽岩
Wang Tao;Li Meng;Meng Liyan(School of Architecture & Engineering, Heilongjiang University of Science & Technology, Harbin 150022, China)
出处
《黑龙江科技大学学报》
CAS
2020年第5期551-555,共5页
Journal of Heilongjiang University of Science And Technology
基金
国家自然科学基金项目(51978213)。
