Stochastic Optimization Method Based on HSS Technique and Expert Knowledge for a Metallurgical Blending Process

In order to make robust decisions under uncertainty, the stochastic optimization method is proposed to effectively realize the optimal control of alumina blending process. Firstly, the multi-objective stochastic optimization model is built to describe the quality requirements of multiple indices and to incorporate uncertainty into the optimization problem. Then, Hammersley sequence sampling(HSS)technique is used to generate the samples of the stochastic variables and the equivalent deterministic optimization model is obtained. Finally, the multi-objective hierarchical expert reasoning strategy based on the empirical knowledge accumulated in long-term production is proposed to realize the optimal control of blending process with uncertainty. The results of industrial application shows that the proposed method can effectively control the quality indices under parametric uncertainty. It provides a good optimization mode for other blending processes of nonferrous metal production with uncertainty.