DETERMINATION OF MACHINING PROCESS FOR ROTATIONAL PARTS USING NEURAL NETWORKS

Computer-Aided Process Planning (CAPP) is the link between CAD and CAM system. CAPP interprets the design information and prescribes appropriate manufacturing processes consistent with the requirements set forth by the designer. Development of a machining process plane is a basic function in manufacturing process. It is a time consuming, and requires significant skills with great deal of experiential knowledge. Process selection is a difficult problem in CAPP since it requires productive CAPP system containing a huge amount of knowledge-facts which limits CAPP capability and flexibility in real manufacturing systems. Al provided some tools for this problem such as artificial neural network (ANN). ANN has the capability of continuous learning and ability to learn arbitrary mappings between input and output spaces. In this paper, a neural network is used for machining process selection in CAPP for cylindrical axis-symmetrical parts. The part features and attributes are the input, and the output is the operation(s) required to produce. Each feature arranged in the same order of logical machining sequences.