Research On Offline Model Updating Hybrid Simulation Method Of Component Constitutive Model

In the model updating hybrid test,the update of the num erical sub-structure model parameters usually selects three levels of component constitutive model,crosssectional constitutive and material constitutive as the update objects,which enhance the accuracy of hybrid test.In the study of the seismic perform ance of complex structures,the component constitutive model update can only update the numerical substructures with the same component parameters,which limits its practical application.Based on the method of component constitutive model updating,this thesis proposes an updating method by contrasting empirical formula to complete the updating of similar component models with different parameters in complex frame structures.This method takes the identification result of the physical substructure test data parameter as the identification value,and takes the model parameter calculated by the empirical formula as the initial value,obtains a set of correction coefficients for correcting the model parameters by comparing the identification value and the initial value.Furthermore,the empirical formula calculation results of similar related components with different component parameters in the numerical sub-structure are revised to improve the accuracy of the hybrid test.The main contents of this paper are as follows:(1)Explain the principle of iterative parameter identification algorithm differential evolution algorithm and the basic process of the algorithm parameter identification.Based on the different cross mutation strategies of the DE algorithm,the parameters of the single degree of freedom Bouc-Wen model are identified.The parameter identification results show that the DE algorithm has better efficiency and accuracy under certain strategies.Compared with the recursive parameter identification algorithm UKF method,an example is given to analyze the accuracy of parameter identification.The results show that both of the two algorithms have good accuracy.Finally,adaptive operator to improve the original DE algorithm is introduced to avoid premature convergence,and this method provide a basis for subsequent offline model update parameter identification.(2)Explain the basic principle of updating the constitutive model of the component constitutive model contrasting empirical formula and explain the selection of the relevant empirical formula.Based on the improved IMK model empirical formula,the improved adaptive DE algorithm is used to identify the parameters of the quasi-static test restoring force model of reinforced concrete components,and the feasibility of the update method against the empirical formula is verified through the prediction of similar component model parameters.Using Open Sees finite element software,complete the hybrid test of the offline component constitutive model update of the elastic reinforced concrete frame based on the contrast empirical formula,and study the accuracy of this type of hybrid test.This method can better improve the simulation accuracy of similar components in the numerical substructure,and the accuracy of the hybrid test is therefore improved.(3)Based on a 5-story-3-bay elastic-plastic RC frame structure,compared with traditional component update,contrast component update,and Open Sees numerical simulation based on empirical formulas,the offline model update hybrid test accuracy under several different working conditions.Use the physical substructure te st data to realize the reuse of offline model update data to complete the numerical simulation test of a 3-story-2-bay RC frame.Compare the correction efficiency of three different scale frameworks under different control component update strategies,summ arize and analyze the influencing factors of the control empirical formula component constitutive model update method.