| In the past two decades Computer-Aided Process Planning (CAPP) has been one of the dominant research topics in Computer-Integrated Manufacturing (CIM). The task of automating the process planning function to model the complexity of thinking process of an expert process planner is both complex and unstructured. This research describes the use of Artificial Intelligence to provide a general framework for the development of a CAPP system for non-traditional manufacturing processes, with wire electrical discharge machining as the application area.; The process planning problem was formulated in a hierarchical fashion. The important steps in the hierarchy were identified and discussed in detail in this research. Because of the complexity of the wire electrical discharge machine (WEDM) process, the CAPP problem was divided into two parts. In the first part using regression and artificial neural networks (ANN), complex relationships between the WEDM input and output process parameters were quantified. In the second part, using expert system (ES) techniques, a knowledge base (KB) was developed and process rules were established for the selection of machines, wires and sequencing of operations. Based on the process rules, algorithms were developed to compute the cost for each plan and determine the optimal process plan. Using Windows 95 as the platform, the WEDM CAPP system was developed in Visual Basic 5.0. The process plans generated were found to be compatible with the opinions of EDM experts. The results of the CAPP system are presented. |
