桥梁极值应力的改进高斯混合粒子滤波器动态预测

发布时间：2018-08-12 07:48

【摘要】：为合理地动态预测在役桥梁的极值应力信息,应用桥梁健康监测(BHM)系统的长期日常监测极值应力数据,建立非线性动态模型,引入扩展卡尔曼滤波器(EKF)与高斯混合粒子滤波器(GMPF)相结合的改进高斯混合粒子滤波器(IGMPF)预测算法,对监测极值应力的一步向前预测分布参数及其状态变量的后验分布参数进行预测分析,并进行了实例验证.IGMPF不仅可以得到实测极值应力状态的合理重要性函数,还可以解决传统预测方法的短期性和精度不高的问题,为实际BHM系统的动力响应预测提供了理论基础.
[Abstract]:In order to reasonably and dynamically predict the extreme stress information of existing bridges, a nonlinear dynamic model is established by using the long-term daily monitoring extreme stress data of the bridge health monitoring (BHM) system. An improved Gao Si hybrid particle filter (IGMPF) prediction algorithm based on extended Kalman filter (EKF) and Gao Si hybrid particle filter (GMPF) is introduced. The one-step forward prediction distribution parameters of monitoring extreme stress and the posterior distribution parameters of state variables are predicted and analyzed. An example is given to verify that .IGMPF can not only obtain the reasonable importance function of the measured extreme stress state. It can also solve the problems of short term and low precision of traditional prediction methods and provide a theoretical basis for dynamic response prediction of actual BHM systems.
【作者单位】：兰州大学西部灾害与环境力学教育部重点实验室;兰州大学土木工程与力学学院;哈尔滨工业大学结构工程灾变与控制教育部重点实验室;哈尔滨工业大学土木工程学院;
【基金】：国家自然科学基金(51608243) 甘肃省自然科学基金(1606RJYA246) 中央高校基本科研业务费专项资金(lzujbky-2015-300,lzujbky-2015-301)
【分类号】：U446;U441.5

【相似文献】