ANALYSIS OF DATA OF ELECTRIC ENERGY METERING MANAGEMENT SYSTEM BY CNN ALGORITHM OF MECHATRONICS

Abstract

- With the development of science and technology, electromechanical integration and the Convolutional Neural Network (CNN) have developed rapidly. At present, one of the more widely used fields is the electric energy metering management system. Data analysis is one of the focuses of research in this field. Therefore, this paper introduces CNN algorithm and explains the advantages and disadvantages of the CNN algorithm in previous studies and the direction of optimization. Secondly, the target detection algorithm and data analysis are described, and the application of the target detection algorithm to image information processing and information analysis in the current research is introduced. Additionally, two methods are proposed for optimizing the CNN algorithm, and the optimization model is re-optimized by introducing the migration model. Finally, comparative experiments are conducted to verify the effectiveness and rationality of this model. The experimental results show that the detection rate of the two optimization methods is higher than that of the traditional model. The detection rate of CNN based on Region Proposal Network (RPN) is higher than that based on Region of Interest (ROI) pooling. Simulation experiments are carried out in different power metering management systems in the second experiment. The RPN-CNN model was introduced into the migration model. In system 1, the maximum difference between the detection rate and the traditional model is 0.2. In system 2, the maximum difference in detection rate is 0.12, which verifies the effectiveness of this model. Additionally, the stability of the RPN-CNN is better than that of the traditional model in the slope comparison of the curve, which proves the feasibility of the model. Therefore, this paper has certain reference significance for the data analysis of the power metering management system.