Features Extraction of Tool Wear and its Detection using Neural Network

가공 재질에 따른 공구 마멸의 특성 추출과 신경회로망을 이용한 마멸 검출

A16061, SB41 and SM45C was used for developing tool wear monitoring system in face milling. First of all, Neural networks of which input are 8 $_{th}$ order AR morel parameters, frequency band energies, cutting conditions was used to monitor tool wear for each material. Finally, A unified neural network, which has tensile strengths of each material as an additional input, was constructed to consider the effect three materials on the features of tool wear. It was verified that tensile strength is the one of properties of workpiece materials.s.