Thermal Error Compensation Modeling Based on Fuzzy C-means Clustering Algorithmand RBF Neural Network Modeling

The selection of thermal critical points and thermal error compensation modeling technique are crucial in deciding the effectiveness of thermal error compensation and important for improving machining accuracy of numerical control(NC) machine.In order to realize the compensation of the thermal error of NC machine,the temperature measurement points are optimized based on the fuzzy C-means(FCM) clustering algorithm and the number of temperature measurement points is cut down from 20 to 4,then the thermal error compensation model is established based on RBF neural network.The experiment result shows that the modeling method is proposed by this paper not only ensure the precision of the model,reduce the measurement points and avoid the correlation of the measurement points,but also improve the robustness of thermal error modeling.