Abstract

This paper proposed a new model for integrating neural networks approach in the task of generating process planning for machining features. The main issue of process planning addressed in this paper is the optimization of machine tools selection for mechanical part containing simple and interacting features. First, this proposed method elaborated a knowledge database from the investigation with expert in manufacturing companies. Then, it finds the optimal machine tools selection by neural networks. Most importantly, the preliminary sequence is refined by including attributes of machining features. Two cooperated neural networks NN1 and NN2 are used for selection of machine tools according to machining features proposed; the first neural networks takes in input the attributes of machining features and produces the suitable classes of machine tools, the second neural networks used for optimization of machine tools selection is according to machining workshop capacity. Finally, a mechanical part is used as an example to illustrate the implementation of proposed method.

Key words: CAPP, CAD/CAM, machine-tools, machining features, neural networks.