基于深度学习的智能变电站变电运维安全管控方法

Key Points

• This research aims to improve safety management in smart substations using deep learning technologies.
• Developed a full-process management framework integrating data collection, hazard identification, safety warning, and decision support.
• Utilized CNN for detecting equipment appearance defects and RNN for abnormal operation status detection.
• Incorporated multi-source data fusion to enhance hazard identification accuracy.
• Achieved significant improvement in hazard identification accuracy through deep learning models.
• Developed a comprehensive warning model that integrates various underlying data for active safety management.
• Proposed optimization strategies for both technology and management levels.

Abstract