Dynamic Prediction Algorithm for Low-Voltage Distribution Network Power Loss in a Smart City Based on Classification Decision Tree and Marketing Data

Abstract

Abstract When the current algorithm is used to predict the power loss of the low-voltage distribution network, the missing marketing data cannot be processed, which leads to the problem of relatively large root mean square error in the algorithm. To this end, this paper proposes a dynamic prediction algorithm for low-voltage distribution network power loss that combines classification decision trees and marketing data. First, use the classification decision tree to classify the marketing data, and select the missing marketing data. Second, the combined threshold filling method is used to fill the missing data. Finally, the process state characterization method is used to realize the dynamic prediction of the power loss of the low-voltage distribution network based on the complete marketing data. The experimental results show that the data missing ratio of the proposed algorithm is less than 0.2, the root mean square relative error is less than 0.02, and the fitness is higher than 0.08 on average, as with the comparison with the three methods of comparison. The results prove the future prediction to be implemented in a smart city.