Multiobjective Optimization Design of Gear Reducer.pdfVIP

Multi-objective Optimization Design of Gear Reducer Based on Adaptive Genetic Algorithms Rui Li, Tian Chang , Jianwei Wang, Xiaopeng Wei, Jinming Wang Advanced Design Technology Center Dalian University, Liaoning,China Abstract. An adaptive Genetic Algorithm (GA) is introduced to solve the multi-objective optimized design of the reducer. Firstly, according to the structure, strength, etc. in a reducer, a multi-objective optimized model of the helical gear reducer is established. And then an adaptive GA based on a fuzzy controller is introduced, aiming at the characteristics of multi-objective, multi-parameter, multi-constraint conditions. Finally, a numerical example is illustrated to show the advantages of this approach and the effectiveness of an adaptive genetic algorithm used in optimized design of a reducer. Keywords: Fuzzy Logic, Genetic Algorithm, Optimization Design, Reducer. PACS: TH132.46 INTRODUCTION Reducer is an important mechanical part which is widely used in metallurgy, mine, architecture, material handling, etc. But it is very difficult to obtain an optimal scheme using the traditional method. Hence Thompson [1] used negotiation to make optimized design aimed for the minimum volume, and Gao[2] adopted evolutionary program. Li and Symmons[2] made optimized design of a helical gear reducer using the minimum centre distance as an objective function. These researches have achieved some promising results. But because single objective function is often used in design, the optimized result is only local optimum in concrete implementation. It is recognized that GAs[4] are very suited for multi-objective optimization problems since they can find a lot of unifor