Evolution Strategy Method For Structural Dynamic Optimization

Most engineering structures during their service may be inevitably affected by the earthquake, the wind, the waves, the explosion and the vibration or attack from other external environment. The dynamic damage and the dynamic fracture will be the major failure modes of the engineering structures. The dynamic optimization of the large and complex engineering structures is a highly non-linear global optimization problem. The traditional optimization algorithm often falls into the locally optimal solution in the solving process and its optimization results often depend on the initial optimization of the starting point. So the suitability of the traditional optimization algorithm is low. The intelligent optimization algorithms, such as the Genetic Algorithm (GA) and the Particle Swarm Optimization Algorithm (PSO), etc., focus globally but the computational cost is too high. Therefore, when searching the solution of the dynamic optimization, it is extremely urgent to find an optimization algorithm which has a good global performance and strong optimization ability.This paper mainly solves the problem of structural dynamic optimization and the research work carried out is basically as follows:first of all, taking the simple truss structure as example, the problem of the structural dynamic characteristics optimization is searched. The CMAES algorithm is applied into the design of the structural dynamic characteristics optimization and compared with the optimal results of the traditional optimization algorithm and PSO algorithm to verify the feasibility of the method. Then, on the basis of meeting the need of calculation accuracy, the GP-based CMAES method is applied into the design of structural dynamic characteristics optimization to verify the efficiency of the algorithm. Finally, taking the simple cross-section optimization of steel-framed structures, shape optimization of truss structures and cross-section optimization of reinforced concrete frame structure under the effects of earthquake as examples, the CMAES algorithm and GP-based CMAES algorithm are applied into the design of structural dynamic response optimization to show the feasibility of the method. And it is compared to the PSO algorithm which has good global performance to verify the computational efficiency of the algorithm.The results of the study show that based on CMAES algorithm, the design method of structural dynamic optimization, presented in the paper, is feasible and has the advantages of good global performance and high efficiency of optimization, which also has fine applicability to solve the problem of the simple engineering structural optimization. And compared to the traditional optimization algorithm and PSO algorithm, the efficiency of the proposed method is proved. At the same time, the application based on GP-based CMAES in the design of the structural dynamic optimization put forward in this paper is validated. Compared to the CMAES algorithm, this method can decrease the times of finite element analysis and the efficiency of algorithm has been further improved. The text method is simple and easy to implement, which provides a new way to rapidly and efficiently solve the problem of dynamic optimization of the large and complex engineering structures.