State monitoring method, device and equipment for pumped storage unit and medium

• Main Authors: PEI JUN, QIU XIAOBO, XU KAIWEI, LIANG YEQUAN, LIU HUAN, CHEN YUNYUN, HUANG XIANGWEI, YU JIALIANG, WAN BO, HU WENXING, ZHANG JIANXIONG, GONG YU, YAN HANQIU, YE LI, YANG MINGXUAN, YU YAXIONG, LUO SHUSHENG, LIU XUAN, LI QING, HUANG ZHONGJIE, PENG WEIWEI
• Format: Patent
• Language: Chinese; English
• Published: 05.05.2023
• Subjects: BLASTING; CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; HEATING; LIGHTING; MACHINES OR ENGINES FOR LIQUIDS; MECHANICAL ENGINEERING; OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR; PHYSICS; PRODUCING MECHANICAL POWER; WEAPONS; WIND, SPRING WEIGHT AND MISCELLANEOUS MOTORS

The invention discloses a state monitoring method, device and equipment for a pumped storage unit and a medium. The method comprises the steps of periodically obtaining working condition signal data of the pumped storage unit; inputting the working condition signal data into a pre-trained pumped storage group classification network model, and determining the condition grade of the pumped storage group; and the state grade of the pumped storage unit is fed back to a user, so that the working state of the pumped storage unit is monitored. The problems that although a deep teacher model corresponding to a pumped storage unit classification network model has good unit working condition judgment capacity, the deep teacher model is large in size, large in occupied memory, low in reasoning speed, difficult to deploy and the like are solved. The performance of the deep teacher model is distilled to the shallow student model, so that the shallow student network can learn more robust features, and the classification ac