This paper takes the environmental protection island equipment for flue gas treatment in coal-fired power plants as the research object,and explores and develops equipment fault identification technology based on artificial intelligence big data technology to solve the problems of low fault identification accuracy and untimely fault discovery of environmental protection island equipment.Experimental testing and application results show that the recognition accuracy,recall rate,false alarm rate and missed alarm rate of the environmental protection equipment abnormality diagnosis model proposed in this article can reach a high level,and can realize abnormal status monitoring and control of environmental protection equipment in coal-fired power plants.Efficient and precise maintenance.