A method for ultra-short-term power load forecasting and early warning

Abstract

The invention discloses an ultra-short-term power load forecasting and early warning method based on a Kalman filter and wavelet echo state network. In order to solve the problem that noise and the like are contained in power load data, a Kalman filtering method is adopted to conduct real-time estimating on 'collected data', with the help of a forgetting factor, the weight of old-fashioned data is weakened, and prediction accuracy is improved. Before ultra-short-term load forecasting is conducted, firstly, a principal component is used for analyzing and determining main working procedures for influencing the change of a power load, the main working procedures are used as the input of a power load capacity prediction model, afterwards, wavelets are used for decomposing the loads of different spectral characteristics (high frequency, follow-up and stability) of the power load, echo state network singe power loads are respectively established for predicting and modeling, various forecasting components are integrated to obtain a total load variation trend, and ultimately an early warning test is conducted on the prediction model specified by a user.

Description

Technical field: [0001] The invention relates to a power load forecasting and early warning method, in particular to an ultra-short-term power load forecasting and early warning method based on Kalman filtering and wavelet echo state network, belonging to the field of information technology. Background technique: [0002] Large-scale industrial enterprises are large consumers of electricity load. The production process is composed of many interrelated production links. The power consumption of each link is determined by its power consumption characteristics and production operation conditions. The total load of the power grid is closely related to the production and operation conditions of the enterprise. The particularity of production results in the high frequency and large amplitude of the power load in some processes, showing significantly different characteristics from the usual regional power grid. Although the company-owned power plant has sufficient power generation capac...

AI Technical Summary

Problems solved by technology

For large industrial enterprises, the load forecasting work starts with the development of the electricity market, the necessary data accumulation is not sufficient, and there are big problems in the data quality

And because the production plan is used for predictive modeling, it will be affected by on-site factors, and the information required for prediction may be unstable.

In addition, predictive modeling needs to consider the internal production links of many enterprises, and the use of a large amount of forecasting information makes it difficult to avoid the occurrence of missing data, which brings great difficulties to load forecasting

[0004] In the prior art, a general forecasting method is used to forecast the power load. The forecasting principle is based on the time series method. This scheme has the following problems: the factors affecting the power load of industrial enterprises are complex, and the power consumption characteristics of the process sections are different. The load fluctuates greatly, the general modeling method is very dependent on the data, and the data fluctuation has a great impact on the accuracy

However, in this invention, all maintenance, power consumption and production plans of each user in the enterprise power system are input as training features, which increases the complexity of the model, and this method can only use a large amount of historical scheduling data for long-term power consumption such as It is difficult to predict the power consumption of several days, weeks, and months, and it is impossible to predict the dynamic ultra-short-term fluctuations, such as the prediction of the next 15 minutes to half an hour.

Images