Robust Design with Sequential Uniform Algorithm for Optimization by Means of PMOO

Regulation of optimum parameters in sequential uniform design of subsequent optimization is developed by means of probabilistic multi-objective optimization (PMOO) in term of total preferable probability. A series of temportary candidate “optimum statuses” which were produced in the subsequent optimization of sequential uniform algorithm is used to form a “special point set”; the objective responses of “special point set” are evaluated once more with PMOO, the total preferable probability is comparatively evaluated to determine the final optimum status and parameters of the entire sequential uniform design process. Comparatively, the final optimum status is with the highest total preferable probability. Besides, under condition of “target value being the best”, both discrepancy of average value \(\overline{Y}\) of a response from its target value Y0, \(\varepsilon \equiv \left| {\overline{Y} - Y_{0} } \right|\) , and averaged deviation γ of actual response value Y from the target value Y0 are taken as the dual individual sub-objectives to conduct the simultaneous optimization. Two examples are given to illuminate the procedure.