关键词: 故障预测, 健康管理, 频域滤波器, 主成分分析法, 数控机床, 故障诊断

Abstract: To improve life and efficiency of Numerical Control (NC) machine tool, key techniques of Prediction and Health Management (PHM) was studied. Fault prediction and life assessment method based on hybrid intelligence was proposed, and a PHM system of NC machine tool typed Changzheng 718 was constructed. Taking screw life prediction as example, implementation scheme for PHM application was proposed. An adaptive frequency filter using Constant Modulus Algorithm (CMA) was put forward to eliminate noise, principal component analysis was used to select the most sensitive features which were extracted by time-domain, frequency analysis and wavelet packet transformation, the fault and life prediction model was set up by hybrid model. Practice results showed that the proposed method could realize fault prediction and condition assessment of NC machine tool, it was with industrial prospects.

Key words: fault prediction, health management, frequency filter, principal component analysis, numerical control machine tool, fault diagnosis