Research On Tool Condition Monitoring And Diagnosis System Of CNC Machine Tool

In industrial production, cutting tool state affects work-piece quality. Effective tool state monitoring system can reduce production cost, improve production efficiency and product quality. On the basis of analyzing the present situation of the research on cutting tool state monitoring and diagnosis system, the cutting force signal of milling is studied.In this paper, starting from analysis of system requirement, the experimental platform is set up to monitor the cutting tool state of the NC machine tool, and the orthogonal experiment scheme is designed. The cutting force signal of the milling cutter is collected under different wear state, which provides the original data for signal processing and feature extraction.In order to extract the features, which can reflect the change of tool state, but is not sensitive to changes in cutting parameters, the paper has done the following work. Through the time domain, frequency domain and wavelet packet analysis of cutting force signal in three directions, peak value, root mean square value and wavelet packet energy were extracted. In view of changes in the cutting conditions of cutting tools, the sensitivity analysis of the features is made, and the average relative change rate is defined. On the basis of the analysis of these features, the method of feature ratio of three-direction cutting force is put forward, and the feature selection is optimized. In order to further study the feature ratio, the sensitivity of the four essential factors of milling is analyzed, and the feature vector combined optimized feature ratio and features is utilized.In order to diagnose the tool state, the tool state monitoring and diagnosis model based on BP neural network and support vector machine are constructed respectively. The two models have good recognition accuracy, with comparison of the two methods, it is found that the BP network has the disadvantage of slow convergence speed, and the support vector machine is more suitable for the identification of tool condition diagnosis.Based on the relevant technology of the system, the prototype system is designed.