Under the background of increasing requirements for safety, automation and intelligence in mining operations, real-time monitoring and fault diagnosis of underground electrical automation equipment have become particularly critical. In order to meet the demand for equipment status monitoring in the complex underground environment, this paper designs a set of intelligent monitoring system architecture for electrical equipment based on edge computing, which contains four main sections: real-time monitoring, data processing, data analysis, and control center. In terms of equipment fault diagnosis, this paper studies GRU neural networks in detail, combines the intelligent monitoring system designed in this paper with GRU neurons, and constructs the equipment fault diagnosis model in this paper. The equipment fault diagnosis model in this paper is tested and analyzed. The precision, recall, and accuracy of this paper’s model for fault recognition are 0.899, 0.913, and 0.935, respectively, indicating that this paper’s model has excellent performance in the field of electrical equipment fault recognition.