Optimization Method Based on Selection Strategy of Initial Point and Six Sigma Robust Design

Usually,there are three problems in the traditional optimization methods: (1) the local optima arising from choosing an inappropriate initial point; (2) low efficiency and accuracy in solving the complex optimization problems; (3) the robustness of optimal solution is not considered. To solve these problems, a new optimization method, the six sigma robust optimal design based on a strategy of initial point selection, is proposed. Firstly, latin hypercube design (LHD) is chosen to determine the optimal initial point and nonlinear Programming by Quadratic Lagrange (NLPQL) is used to search for the optimal solution. Then, the six sigma robustness of the optimal solutions is analyzed. Finally, robust optimizations of the responses which do not meet the six sigma reliability are performed and the global optimal solutions are obtained. The optimization of the machine tool spindle is used in this paper to validate the performance of this method. The results show that the optimal solution obtained by using the proposed method can ensure the high reliability, and also has a high convergence speed.