Research of Key Technologies for Automatic Acceptance of Electric Power E-commerce Materials Based on Image and Data-driven Approaches

Various emerging technologies and concepts are constantly being integrated into the power system to improve the efficiency of power grid operation. The types of power equipment and materials involved in multiple links such as power generation, energy storage, transmission, distribution and power consumption are increasing, which brings great challenges to the comprehensive management of power e-commerce. Especially in the acceptance stage of materials, the traditional manual acceptance method is inefficient and easily affected by human factors. This study aims to develop an automatic acceptance system for power e-commerce materials based on image recognition and data-driven to improve the intelligence level, accuracy and efficiency of the material acceptance process. By combining image recognition technology, GPS location information and weight sensor data, a new automatic acceptance method is proposed. The system uses sliding window filtering and error removal algorithms to preprocess data, and generates symmetric key encryption to transmit images to ensure the security of data transmission. Furthermore, the CNN model is used to review the matching degree of acceptance materials to ensure that the accepted materials are the items required by the purchase order. The experimental results show that the system can achieve a high correct acceptance rate under various working conditions, effectively reducing the cost of manual participation and improving the intelligence level of the power material management process.