基于Kriging模型的地质沉积模拟平台优化设计

In order to optimize the dynamic and static stiffness of the platform for geological sedimentation simulation, a method of multi-objective optimization driven by the maximum deformation and first natural frequency is proposed. By combining the ideas of the optimal design of space filling and Kriging interpolation to construct the initial agent model for optimal design, and by using experimental design method, the error is bigger regional encrypted local encryption processing optimization, the approximate optimal solution and iteration of ideas, both inside and outside layer updating is established both in local and global space with high approximation accuracy of Kriging model. On this basis, Pareto optimal solution of the objective function is obtained based on the multi-objective genetic algorithm. By the optimization, the maximum stress of the platform of geological sedimentation simulation is reduced by 11.71%, the maximum deformation is reduced by 22.27%, the mass is reduced by 1.6%, the first natural frequency is reduced by 0.67%.