Abstract: In response to the current situation of long fault processing cycles, heavy maintenance tasks, and high maintenance costs in equipment management, a framework for equipment fault diagnosis and health management has been constructed, including the device layer, perception layer, data processing and storage layer, data analysis layer, and application layer. In the data analysis layer, a equipment fault diagnosis and health management model was established by combining BP neural network and genetic algorithm. Finally, Taking the vibration data of electromechanical equipment as an example, the prediction results of the equipment fault diagnosis model were analyzed, thus verifying the feasibility of the model. The research results indicate that this model can improve the accuracy of equipment fault diagnosis,and has good fault diagnosis performance. The predicted trend of health status of the equipment is basically consistent with the actual health status changes, with a coincidence rate greater than 90%. The research results can provide relevant strategies for equipment fault diagnosis and health management in manufacturing enterprises, effectively eliminating fault problems and reducing management costs.