Research On The Intelligent Design Of Injection Mold Scheme Based On Multi-Agent

The reasonability of injection mold scheme design directly affects molding quality and production efficiency of plastic parts, as well as manufacturing cost and production cycle of injection mold. And according to different structural features of plastic parts, how to obtain the rational injection mold structure which meets the need of molding quality in the shortest time is one of the important contents about advanced mold technology research. And the development of intelligent CAD technology provides effective ways and methods to achieve high-quality injection mould design. Therefore, research on intelligent design techniques and methods of injection mold is required by the development of modern industrial products.On the basis of deep analysis for process characteristics, knowledge types and details of injection mold design, a hybrid method combining multi-agent, extenics and inventive problem solving (TRIZ) was proposed to study the intelligent implementation of injection mold scheme design. And the overall framework of intelligent design system for injection mold was built by utilizing JAVA programming language and SQL database techniques based on JADE. Through the in-depth analysis of functional components in injection mold, as well as the communication mechanism and architecture of agents, the intrinsic relationship between functional components and the corresponding design agents was established. Meanwhile, a general structural model of design agents was presented. And on that basis, task allocation and cooperation mechanism between different agents within injection mold design system which was based on the improved contract net protocol, was built to realize reasonable and orderly distribution of the whole mold design task. Then, the knowledge representation approach was studied to describe different types of injection mould design knowledge and expert experience through integrating the improved extension matter element model and method of production knowledge representation. Besides knowledge base for injection mould design was built. Inside design agents, the reasoning mechanism based on extenics and TRIZ was investigated. And the extensional rhombus thinking method was applied together with the correlation between design features and engineering parameters of TRIZ contradiction matrix to realize effective reasoning and quick resolution of the conflicts for the intelligent design process. Moreover, an optimal design program of injection mold functional component was obtained.Finally, the rectangular box shell plastic part was taken for an example and the intelligent reasoning process of injection mold scheme design was given. The results show that a correct and proper injection mold scheme can be gained from the injection mold intelligent design system built in this paper by by combining multi-agent technology and extenics and TRIZ.