Intelligent Optimization Design Method

The wide use of optimization technique in engineering tasks requires designer not only to master his own domain knowledge, but also to be equipped with optimization background theory and design methodology. Experts have gained a lot of experience though years of work. Using scientific ways to extract rules from this experience and denote it with computer language, based on this, software was built which can perform optimization automatically. This automatic optimization method is easy to use among engineers and designers, and therefore, reaching the goal of popularizing optimization technique in engineering and improving the quality of design.This paper combines the production expert system technique in Al with the traditional numerical optimization technique, study the knowledge processing mechanism in expert system and experience knowledge driven deductive optimization, design a framework of expert system upon the numerical optimization, and provide a software with some certain degree of AL, facilitating the wide adoption of optimization technique in engineering design practice.With standard C++ language, a framework of expert system was built which seamlessly combined numerical optimization and knowledge based deductive optimization technique. This software can extract rules from the experience that domain experts gained though years of work, store it into its knowledge database hence upgrading its problem solving capabilities. This prototype software can use its heuristic rules contained in its knowledge database to guide the parameter optimization process, refine the outcome faster and more efficient. Finally, there is a real engineering optimization problem for this software, which proved its practicability.