中图分类号：TP181；TP277文献标识码：ADOI:10.19339/j.issn.1674-2583.2026.02.008
中文引用格式：邢宇鹏，秦存金，田清远，等. 电机故障预测与智能诊断系统研究及应用［J］.集成电路应用，2026，43（2）：38-44.
英文引用格式：Xing Yupeng，Qin Cunjin，Tian Qingyuan,et al. Research and application of motor fault pragnosis and intelligent diagnosis system［J］.Application of IC，2026，43（2）：38-44.

Research and application of motor fault pragnosis and intelligent diagnosis system

Abstract： In modern industrial scenarios, the continuous and stable operation of electric motors is essential to ensuring production efficiency and equipment safety. To address the limitations of traditional fault diagnosis methods, such as reliance on veteran experience, delayed response, and insufficient generalization capability, this paper proposes and implements an intelligent motor fault diagnosis and prognosis system centered on vibration data and integrated with multi-source information.Aiming at typical industrial motor applications, the system employs a hybrid GMM-SVM framework for operating state identification, and introduces an LSTM network to model fault evolution and trend prognosis. By combining model-level weight fusion with a threshold-based decision mechanism, the robustness and practical applicability of fault prognosis are significantly improved.Experimental evaluation and on-site industrial validation demonstrate that the proposed system satisfies the requirements of trend-oriented prognosis in industrial environments. Compared with conventional diagnostic approaches, it achieves a notably higher fault identification accuracy and effectively supports prognostic maintenance and intelligent motor management. The proposed framework provides a scalable and practical technical solution for industrial equipment health management.

Key words : vibration analysis; multisource data fusion; prognostic maintenance; fault diagnosis; machine learning