Research on Electricity Customer Behavior Prediction Based on Improved Time Convolutional Network

In the process of practical application of the power customer behavior tendency prediction method, there are problems that the prediction time is too long and the accuracy of the behavior tendency prediction is poor. For this reason, an improved time convolution network based power customer behavior tendency prediction method is designed. It mines customer electricity demand, calculates the electricity purchase ratio of electricity users, divides the type of electricity customer behavior tendency, calculates the capacity utilization rate of power enterprises, uses the improved time convolution network to optimize the prediction model, and realizes the design of the electricity customer behavior tendency prediction method. The experimental results show that the average time-consuming of the power customer behavior tendency prediction method designed in this paper is 5. 438s, and the accuracy rate of the power customer behavior tendency prediction is 95.9%, indicating that the power customer behavior tendency prediction method designed in this paper has better performance.