State classification of CBN grinding with support vector machine

When grinding high-strength ferrous alloy with CBN wheel, attention is often paid to the variation of wheel surface condition to ensure work surface quality, since wheel sharpness is directly related to the ground surface. The on-line wheel condition can be obtained via process monitoring, which integrates the operations of process sensing, featured data extraction, and state assessment. When grinding with wheel state under control, work quality is ensured. On-line process state recognition usually relies on a pre-built classification model. This paper is to classify the intercepted grinding acoustic emission data using support vector machine (SVM) based on ground roughness variation during grindable period. An SVM model was constructed from the result of a grinding experiment and confirmed to find an 85% of prediction accuracy.