Abstract: In view of problems existing in current coal preparation equipments fault prediction, such as equipments monitoring data conversion method was not unified, single fault classification model cannot meet requirements of multiple fault types prediction, and the fault classification model cannot be updated and modified online. Based on remote operation and maintenance mode, a remote fault prediction system for coal preparation equipment was designed. By preprocessing sensor data of coal preparation equipment, full information is generated and input to various fault classification models. Overall fault index is weighted and fault prediction result is obtained through comprehensive decision. It can provide technical support for the whole life cycle management of intelligent coal preparation plant equipments. The application results verify effectiveness of the system.