An Expert System Based on FBFN Using a GA to Predict Power Requirement in Cylindrical Traverse Grinding

Abstract

Power requirement is one of the most important process parameters in cylindrical traverse grinding. Due to the inherent complexity of the process, it may be difficult to derive the exact mathematical expression of the input-output variables relationships. An expert system is developed in this paper, based on the fuzzy basis function network (FBFN) to predict power requirement in grinding process. An approach for automatic design of RB and the weight factors for different rules is developed using a GA training algorithm based on error reduction measures. To increase the accuracy of the FBFN, the membership function distributions of both input and output variable are tuned simultaneously. Simulation and experiment studies are performed to demonstrate advantages of the proposed modeling framework with the training algorithm in modeling grinding processes.