NEURO-FUZZY CONTROL OF COMPLEX MANUFACTURING PROCESSES

A scheme for intelligent optimization and control of complex manufactu ring processes is presented. The underlying nonlinear process is model led by artificial neural networks and process control is performed by fuzzy logic. Fuzzy rules are automatically generated from the trained neural networks through a novel rule generation mechanism and fuzzy co ntrol is performed by Mamdani implication. Simulation results show tha t the proposed approach can provide a robust and accurate way of contr olling complex processes without comprehensive models or knowledge abo ut the process.