Online Dynamic Load Identification Method Based On Bayesian Method And Experimental Research

In engineering structures,dynamic loads cannot generally be measured directly.Therefore,the vibration response information of the structure is measured to push back to its real load.In this paper,Bayesian-based online dynamic load identification method and experimental research are carried out.This paper derives Bayesian Kalman filter load identification,applies it to online load identification,and uses Lab VIEW software to compile system programs.Kalman filtering is recursive and real-time,and can be used for online load identification.Firstly,the modal parameter identification method based on Bayes is deduced.The structural excitation is regarded as the environmental excitation,and the structure response is FFT transformed.The posterior probability density distribution and covariance of the modal parameters are obtained,and then the modal parameters of the structure are obtained.Simplified into a single-modal identification algorithm,the method is verified through simulation and experiment of a simply supported beam,and the identification result is used in load identification.Based on Bayesian theory,an online dynamic load identification method was derived in detail.Combined with modal coordinate conversion,an online load identification method based on Kalman filtering was proposed.Through numerical simulation and experimental verification,it is discussed that for different types of loads and different noise levels,the recognition effect is still high.For the situation where the modal parameters are unknown,the modal parameter estimates are obtained through Bayesian modal parameter reconstruction.The estimated values of the modal parameters are added to the state vector to form an augmented vector matrix,and the load identification method of extended Kalman filter is deduced,and the online recognition algorithm of extended Kalman filter load in modal space is completed.The identification effect is verified by numerical simulation and experiment.The dynamic load online identification system was designed,and the software development was completed by Lab VIEW Virtual Instrument Development Platform.Design the test: Excite the simply supported beam,read its acceleration response,identify the load and model parameters,and the test results verify the feasibility and availability of software.