Model updating of engineering structures is usually facing the practical problems of high dimensional variables and strong nonlinearity,which seriously affects the accuracy and efficiency of model updating.In order to solve the problems,a combined method of advanced colliding bodies optimization(ACBO)and Gaussian-white-noise mutation(GM)is proposed.In this method,the ACBO is used to realize the transition from the traditional″one to one″ pattern to″optimal to many″ collision pattern,and the GM is used to ensure the diversity of population during the collision process.Hence,the efficiency of structural model updating is improved greatly.Based on a series of test functions,the optimization performances of the standard colliding bodies optimization(CBO)algorithm and the combined approach are analyzed and compared.The updating process based on ACBO-GM combined method is introduced,and the proposed method is applied to a steel frame and a cantilever beam.The updating performances such as accuracy and efficiency of the combined method and the genetic algorithm(GA)are compared.The results show that the ACBO-GM combined method is superior to the traditional CBO method in both accuracy and efficiency,which verifies the feasibility of the proposed method.Furthermore,the applications of ACBO-GM combined method to aforementioned two engineering structures indicate that the accuracy of this method is essentially the same as that of GA,but the efficiency is much higher than that of GA.
维普
万方数据