Research And Development On Simulation Optimization System Oriented To Manufacturing Resource Configuration

Virtual simulation technology has become an effective way to solve analysis, design, control and management of complex manufacturing system. However, as an experiment method, simulation technology could only evaluate and validate the manufacturing model without any intelligent decision, so it has prevented the further development of virtual simulation technique seriously. Simulation optimization technique which has been developed in resent years has bridged the gap. On one hand, it solves the problem that traditional optimization technique can not describe the model of complex system. On the other hand, it combines intelligent decision mechanism with simulation technology. But at present, most of researches in this field are only focused on simulation optimization algorithms, the studies about the features of specific application area and object are ignored, especially about the entire system of simulation optimization.Based on the predecessor's achievements, according to the separation principle of model and algorithm a new simulation optimization framework oriented to manufacturing system is put forward in this paper. Considering the characteristic of configuration of manufacturing line resources, a new genetic algorithm (NGA) is proposed on the basis of traditional GA. In this algorithm, niche population produce technology, whole and part crossover strategy and adaptive mutation operator are applied. In addition, three optimization strategies including border characteristic and so on are integrated with optimal algorithm. These above technologies improve the efficiency and quality of optimization greatly. Then combined with simulation optimization algorithm, a simulation optimization engine is created to solve the resource configuration optimization of manufacturing system. According to the above research, key technologies including optimization explain machine, optimization database, data interface and so on are solved, and based on QUEST 5.0 manufacturing system platform of Delmia Company, a simulation optimization system named SimOpt 1.1 for manufacturing resource configuration is developed. Finally, the system is validated by some examples and the comparing results of congener software are given in the paper.