Abstract: The intelligent maintenance of coal mine electromechanical equipment is an important part of the construction of intelligent mines. The research status of intelligent fault diagnosis and predictive maintenance of coal mine electromechanical equipment is summarized from four aspects, coal mine electromechanical equipment fault mechanism, equipment condition monitoring, signal analysis and processing, and fault diagnosis and prediction algorithm. ① Equipment fault mechanism research is mainly for different equipment, which uses different methods to establish equipment fault analysis models and applies incentives to the models to obtain dynamic response to equipment fault so as to provide a basis for subsequent fault diagnosis. ② Equipment condition monitoring research has established a relatively complete condition monitoring system for coal mine electromechanical equipment, which can obtain equipment parameters accurately and timely and provide data support for equipment fault diagnosis. ③ Signal analysis and processing research not only uses traditional time domain, frequency domain and time-frequency domain analysis methods, but also combines multiple methods for signal processing and characteristic extraction, which improves the efficiency of signal processing and the reliability of processing results. ④ Fault diagnosis and prediction algorithms mainly use artificial neural networks, including machine learning, deep learning and other intelligent algorithms to establish equipment fault diagnosis and prediction models so as to achieve intelligent diagnosis and prediction of faults. It is pointed out the problems of intelligent maintenance research of coal mine electromechanical equipment are as follows. ① The study of equipment fault mechanism lacks the study of fault mechanism under multiple fault compound state. And more research is needed on the chain reaction caused by the fault of a certain part of the equipment. ② The fault data obtained in the simulated environment cannot fully reflect the actual operating conditions of the equipment, and it is necessary to continuously collect on-site monitoring data for a long time, preferably the whole life cycle data of the equipment. ③ At present, there is few research using the combined algorithm, and the research object is more limited to a certain part or component of the equipment. Finally, the development trend of intelligent maintenance of coal mine electromechanical equipment is proposed as follows. ① It is suggested to study the application of intelligent sensors with higher sensitivity to monitor equipment, combine the methods such as random resonance and blind source separation to extract weak characteristic signals from strong noise to identify early equipment faults in a timely manner, thus realizing predictive maintenance. ② The use of independent diagnosis method can no longer meet the diagnosis needs of actual equipment. The diagnosis and prediction based on multiple sensor information fusion technology can identify all faults in the equipment accurately and effectively. ③ It is proposed to use the migration learning algorithm as a 'bridge' to establish the correlation between simulation, test data and field data, and provide data support and guarantee for solving the problem of the difference between simulation and test conditions and field conditions.