An intelligent CAD system for cold-forging parts in automobile industry

Abstract

In order to improve the productivity of automobile industry, cold-forging technology has developed rapidly, so it is very important to advance the cold-forging process and die design, which has been heavily dependent on the experience of engineers. The great development of AI technology has made it possible to build up intelligent systems to resolve problems requiring qualitative and ambiguous human experience. In this paper, some new work and results are presented on the development of an intelligent CAD system for cold-forging parts in the automobile industry. The knowledge of the system is classified into three groups: facts, procedures and heuristics, then the structure of knowledge base is illustrated. To describe the geometry features of cold-forging parts, the coding of product geometry is analyzed and studied, and the coding method is shown in a figure. To predict the elastic deformation and die life, a newly developed AI technology-artificial neural network (ANN), along with fuzzy mathematics, is introduced to establish a model of fuzzy neural networks. Then learning algorithms are illuminated, and the effectiveness of this model is verified by feeding the network with test data not included in the training data.