Development of genetic algorithm-based fuzzy rules design for metal cutting data selection

Abstract Fuzzy rules optimization is always a problem for a complex fuzzy model. For a simple 2-inputs-1-output fuzzy model, the designer has to select the most optimum set of fuzzy rules from more than 10 000 combinations. The authors have developed fuzzy models for machinability data selection (Int. J. Flexible Autom. Integrated Manuf. 5 (1 and 2) (1997) 79). There are more than 2×10 29 possible sets of rules for each model. The situation would be more complicated if there were a further increase in the number of inputs and/or outputs. The fuzzy rules (Turning Handbook of High-Efficiency Metal Cutting, General Electric Co., Detroit) were selected based on trial and error and/or intuition. Genetic optimization has been suggested in this paper to further optimize the fuzzy rules. The development of a Fuzzy Genetic Optimization algorithm is presented and discussed. An object-oriented library to handle fuzzy rules optimization with genetic optimization has been developed. The effect of constraint rules is also presented and discussed. Comparisons between the results from the optimized models and literature are made.