Application of genetically-generated fuzzy knowledge bases in manufacturing

The need of an expert to build the knowledge base (KB) of fuzzy decision support systems (FDSS) is a strong limitation to the expansion of their use in industry. However, this paper presents a genetic algorithm (GA) capable of automatically building KBs from a set of sampled data. The GA produces a KB allowing an optimal approximation of a set of sampled data from a low amount of input information. The GA is validated with a theoretical surface and applied to a tool wear monitoring problem.