Global optimum design to problems with continuous and discrete design variables

In the research of optimization, the most significant issues involve the global optimum for multi-optima problem and the number of function evaluations, which is related to the cost of computation. In this study, to find the global optimum design efficiently, holographic neural network is introduced to be an activate function of response surface methodology to find the most probable optimal design. Since the accuracy of approximation function near the global optimal design is merely important, techniques to search the region of containing the global optimal design most in probability sense by using conditional random seeds, and techniques for finding more accurate approximation near the most probable global optimal design are exploited. The proposed approach called most probable optimal design (MPOD) method is employed to pick up one local optimum design which has the biggest probability in the design space. Design example of crash worthiness for the passenger injury with mixed continuous and discrete design variables are shown the validity of the method.