关键词: 热处理, 炉次计划, 分类, 遗传算法, 模拟退火

Abstract: To improve the efficiency of casting heat treatment,a multi-objective integer programming model was presented by taking furnace combining constraints,furnace capacity utilization and delivery deadline into account,and the solving method by combining genetic algorithm with classification was also proposed.Based on furnace combining constraints,the proper candidate sets of charge plan were generated with classification.Each candidate set was evaluated by utilization of furnace and delivery,and the top five of candidate sets were selected as the final candidate sets for charge plan.The genetic algorithm was improved by adopting two different crossover operators,simulated annealing mechanism and reset operator.The optimal solutions for five candidate sets were obtained by the improved genetic algorithm,and the best of which was considered as the optimal charge plan.The effectiveness and applicability of the proposed model and solving algorithm were demonstrated by simulation and comparison with production practice.

Key words: heat treatment, charge plan, classification, genetic algorithms, simulated annealing