Structural FEM Updating Based On Jaya Algorithm

Calculation efficiency is one of the main limitations of model modification in engineering application.Surrogate model can approximate the unknown relationship between design variables and optimization objectives with explicit function and calculate the input-output static and dynamic forces of the model according to this function relationship,which can greatly improve the calculation efficiency.In recent years,it is a new development trend in the field of model updating to transform the finite element model updating problem into optimization problem.Therefore,the optimization method based on surrogate model has attracted more and more attentions.Compared with other surrogate models,Kriging model can not only give the estimated values of unknown functions,but also get the error estimates of the estimated values.Jaya algorithm is a new intelligent optimization algorithm,which has a good application prospect.However,there are also some problems such as less initial population information and low convergence accuracy.Frequency response function can provide more structural information.Based on this,the main work of this thesis is as follows:Aiming at the problems of Jaya algorithm,an improved adaptive Jaya algorithm with Lévy flight(SL-Jaya algorithm)is proposed.Firstly,chaos mapping is used to make the initial population distribution more uniform.Secondly,the improved adaptive population strategy is used to improve the diversity of the population.Then,Lévy flight is introduced into the search strategy of the algorithm to make the algorithm jump out of the local optimum and enhance the global search ability.Finally,the Gauss-Cauchy hybrid mutation strategy is used to balance the global search and local search ability of the algorithm.The simulation results of12 different feature standard test functions in multiple dimensions show that the SL-Jaya algorithm has better convergence speed,optimization accuracy and robustness compared with the basic Jaya algorithm and the Artificial Bee Colony(ABC)algorithm.Kriging model combined with SL-Jaya algorithm is introduced into the model updating based on frequency response function.Firstly,the optimal excitation point is selected by modal participation criterion.Secondly,the parameters to be modified are determined.The sample points are selected by Latin hypercube sampling to construct Kriging model.Then,the minimum frequency response difference is taken as the objective function of model updating,and SL-Jaya algorithm is used to solve the parameter updating value.Finally,the truss structure and the frame structure of paper yarn composite bag bottom pasting machine are modeled by this method.Compared with the results by other algorithms.The results show that the structural dynamic response of the updated finite element model is consistent with the test response and the error of the updated parameters is less than 2.5%.SL-Jaya algorithm has better optimization results than the basic Jaya algorithm,Particle Swarm Optimization(PSO)and Cuckoo algorithm,which shows that the finite element model updating method based on the SL-Jaya algorithm combined with the Kriging model is suitable for the model updating of the truss structure and the frame structure of bottom pasting machine.The updating efficiency is higher and the calculation result is more accurate.