Bridge extreme stress dynamic prediction based on improved Gaussian mixed particle filter new algorithm
The extreme stress data is taken as a time series,an improved Gaussian mixed particle filter(IGMPF)dynamic prediction new approach of bridge extreme stresses is proposed.Firstly,the dynamic nonlinear model,which provides state equation and monitored equation for the particle filter,is built with the monitored bridge extreme stress data;then,the EM algorithm is introduced to estimate the probability density function(PDF)of the target state and embedded in the Gaussian mixed particle filter(GMPF);further,with the IGMPF prediction approach,structural stresses are dynamically predicted based on the monitored extreme stress data;finally,the monitored stress data of an actual bridge is provided to illustrate the feasibility and application of the proposed models and methods.The result shows that the proposed algorithm has good prediction accuracy,can apply to real engineering.
NETL
NSTL
万方数据
