Research on Intelligent Prediction of Time-of-Use Price on Power Sale

Aiming at the low accuracy forecasting of traditional Time-of-Use (ToU) price, a key intelligent price forecasting technology based on Internet is proposed. Taking the historical ToU price data on the Internet as input, the middle pass filter function and wavelet decomposition are used to preprocess the historical data. After converting the data into values in the [0, 1] interval by using the normalization method, input the neural network. Using the input data, the neural network is trained to determine the weight of the forecast, complete the forecast of ToU price, and output the forecast result. By comparing the two technologies and the real value, the relative error of the designed intelligent prediction technology is 1/5 of the relative error of the traditional prediction technology, indicating that the designed intelligent prediction technology is more accurate.