Abstract: In recent years, with the development of social economy, customers have increasingly high requirements for power supply reliability. To improve power supply reliability and reduce the impact of power failure on the majority of power consumers, it is imperative to further promote the intelligent fault diagnosis technology of power equipment. Deep learning has received extensive attention in many fields due to its powerful automatic data feature extraction capability and the end-to-end training mode integrated with classifiers. This paper studies the traditional fault diagnosis methods and deep learning related theories, analyzes and proposes three application models of power equipment fault diagnosis based on deep learning, which is of great significance to the application research of deep learning in power equipment fault diagnosis in the future.