Fuzzy model based in-process tool wear estimation

This paper discusses the issues of in-process tool wear state estimation which takes a key role in the realization of adaptive control with optimisation for machining processes. Since machining processes are highly nonlinear, and suffer from noise, disturbance and parameter uncertainty, it is impossible to achieve reliable and precise estimates of tool wear state by means of quantitative model based approaches. Here, a methodology for fuzzy model based in-process tool wear estimation is studied, and a systematic way to build the fuzzy inference model by means of genetic algorithms has been developed. The advantages of using artificial intelligence in fusing multi-sensor data to estimate tool wear state are further demonstrated