Abstract
Reliable and precise multi-step-ahead tool wear state prediction is significant to modern industries for maintaining part quality and reducing cost. This study proposes a Clustering Feature-based Recurrent Fuzzy Neural Network (CFRFNN) for tool wear state monitoring and remaining useful life (RUL) prediction based on K-means Clustering, Recurrent Fuzzy Neural Network (RFNN) and Genetic Algorithm (GA). K-means Clustering method is utilized to realize tool wear state definition and input signal division, which reduces the dependence on the prior knowledge of tool wear degree and improves the prediction accuracy. Then, an enhanced RFNN model is designed and applied on the clustered features to predict tool wear state. The optimized GA technique is helpful for adaptive optimization of model parameters, which significantly improves convergence rate and prediction accuracy. The experiments on tool state prediction are performed to validate superiority of CFRFNN, and the results demonstrate that the proposed network could reasonably configure the complex non-stationary tool wear process and have high prediction accuracy of tool wear state.
Highlights
Tool wear is a pervasive problem in manufacturing process, which has an extremely negative effect on the performance and productivity of computerized numerical control (CNC) machines
The performance of clustering feature-based recurrent fuzzy neural network (CFRFNN) shows the improved accuracy and generalization ability in multi-step-ahead tool wear state prediction
We have proposed a clustering feature-based recurrent fuzzy neural network (CFRFNN) based on tool state management systems (TSMS) for multi-step-ahead tool wear state and remaining useful life (RUL) prediction
Summary
Tool wear is a pervasive problem in manufacturing process, which has an extremely negative effect on the performance and productivity of computerized numerical control (CNC) machines. The extracted features are utilized as input to different models, such as auto-regression [9], manifold learning [10], hidden Markov [11], sparse decomposition [12], and deep learning [13] Among these methods, deep learning models have been extensively applied in tool state diagnosis and prognosis due to the powerful capabilities in complex sequence modeling. In addition to the above deep learning models for fault monitoring and diagnosis, recurrent neural network (RNN) as crucial branches of deep learning models, have been widely applied in multi-step-ahead tool wear state and RUL prediction. A clustering feature-based recurrent fuzzy neural network (CFRFNN) is proposed for automatic identification and prediction of tool wear state.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
