Target Detection Method for Energy Storage and Power Supply Service Cabin Based on Improved YOLOv5s

A target detection method for energy storage power supply service cabin based on improved YOLOv5s is proposed to address the issues of low accuracy and low efficiency in target detection of energy storage devices. Firstly, the overall architecture of the energy storage and power supply service cabin is designed, including battery packs, various sensors, monitoring systems, and other units, with sufficient information exchange to meet the needs of emergency power supply, marketing business services, and other services. Then, the adaptive feature extraction module and Focal-loss function are used to optimize YOLOv5s and apply it to the detection of target images inside the cabin. Finally, the improved YOLOv5s network will be embedded into the energy storage and power supply service cabin management system to detect abnormal operating conditions inside the cabin and issue warnings. Based on the operation sample set of the energy storage and power supply service cabin, the proposed method is experimentally demonstrated. The results shows that its average detection accuracy reaches 94.52%, and the detection time is 8.73s, which can quickly and accurately achieve target detection.