Research On Optimal Approaches To Two-Dimensional Cutting Stock Problems Based On Computational Intelligence

In the process of large-scale electromechanical equipment manufacturing,the degree of organization and management as well as measures of technique during the course of cutting stock,which do a great effect to produce cost and produce cycle. To different parts and material,it is a little difference in form to resolving cutting stock problem,but they are both very difficult to deal with In a long time,various intelligent methods of cutting stock problem based on different theory are hotspot project to be researched in domestic and overseas research. Their object is to find a way,which is effective and economical to resolve this kind of problem.The problem of optimal placement of stock sheets is that planning plane parts on the stock sheets in the most optimal combination and layout. Many scientific researchers have studied this question deeply and widely. It was proved that the problem of optimal placement of stock sheets is more difficult than the bin packing problem. The bin packing problem belongs to the most complicated calculated problem:NPC problem (Nondeterminstic Polynomial Complete). When there are a lot of parts to be cut,it need many stock sheets,and had to use the remaining sheets at first. In many cases,there are more than one sheet,and the acreage of these sheets are not the same. Furthermore,placement of stock sheets should not overlap on each other,and should do BOOL calculation,as well as switch their station and direction to these parts. All of these requirements make the problem more difficult to besolved.Since 1980s,there have been developed some modern optimal algorithms such as taboo search algorithm,simulated annealing algorithm,genetic algorithm and artificial neural networks algorithm. They have been applied to resolve NPC problems,but these algorithms can't find the best result unconditionally. When do the problem of optimal placement of stock sheets,it is not easy to get the satisfactory result just use one of these algorithms signally. According to the case,the cooperation optimal strategy has been brought forward in my paper. In fact,the cooperation optimal strategy is to cooperate genetic algorithm,local research algorithm and simulated annealing algorithm to resolve the problem.Made the producing process of welding row parts of electromechanical equipments as apply background,this paper design a kind of intelligence cutting stock sheets system based on cooperating various optimal algorithms. The system is made up of five sects:information of parts management,information of stock sheets management,cooperating optimal calculation,graph of placement build,cutting result output. The characteristic of the system is concision of interface,simple operation and high efficiency. The cooperation optimal strategy is the highlight of this system.