Intelligent Diagnosis for Cutting Processes

Abstract

Because of the requirements of automated processing, the cutting status of a tool should be monitored in real-time to find out the processing quality in advance and detect cutting abnormalities in time. Thus, we developed an application called APP for tool-life management and damage detection to help operators control the timing of tool replacement and avoid time-delays and loss after damage to the product. The proposed system consists of two sub-systems: an electronic system (to capture signals) and a software system (APP). The electronic system has an accelerometer, a signal capture card, a current sensor, and a temperature sensor. The software system has operating interfaces, data visualization, and early warning systems. The principal component analysis uses a linear combination with the largest variance. The eigenvalues and eigenvectors can be obtained by drawing a slope diagram of the number of principal components by eigenvalues. The point when the slope changes significantly can be used as the selected monitoring characteristic parameter. To reduce the manpower required for follow-up processing, the proposed system can be combined with different controllers to achieve customized parameter adjustment. Unmanned control can ensure processing efficiency and quality.