Multi-Step-Ahead Tool State Monitoring Using Clustering Feature-Based Recurrent Fuzzy Neural Networks

Reliable and precise multi-step-ahead tool wear state prediction is significant to modern industries for maintaining part quality reducing cost. This study proposes a Clustering Feature-based Recurrent Fuzzy Neural Network (CFRFNN) monitoring remaining useful life (RUL) based on K-means Clustering, (RFNN) Genetic Algorithm (GA). method utilized realize definition input signal division, which reduces the dependence prior knowledge of degree improves accuracy. Then, an enhanced RFNN model designed applied clustered features predict state. The optimized GA technique helpful adaptive optimization parameters, significantly convergence rate experiments are performed validate superiority CFRFNN, results demonstrate that proposed network could reasonably configure complex non-stationary process have high accuracy