Journal of Electrical Engineering and Automation

With the continuous improvement of industrial automation, traditional fault diagnosis methods gradually reveal their limitations. Especially in complex environments, higher requirements are imposed on accuracy and real-time performance. The fault diagnosis method of industrial automation control system based on deep learning can effectively improve the accuracy of fault detection and prediction by utilizing the powerful features of deep neural networks. This study deeply explores the application of deep learning in the fault diagnosis of automation control systems, analyzes common network models and algorithms, compares the performance of different models in industrial fault diagnosis, and proposes a multi-level diagnosis framework based on deep learning. Deep learning can greatly optimize the efficiency of fault diagnosis and improve the safety and stability of industrial production.

Deep learning; Industrial automation; Fault diagnosis; Neural network; Control system