Bus Load Forecasting Method Based on Cascade Generalization Training Strategy

Abstract

The invention discloses a bus load prediction method based on a cascaded generalization training strategy, which is characterized in that data preprocessing is performed on the bus load, original samples are constructed using characteristic variables, and a group of extreme learning machines are constructed to perform cascaded generalization training, and the training is completed Finally, a new feature space is obtained, which is composed of the output of these extreme learning machines and the corresponding real values; then the new feature space is linearly combined, and the output sequence in the new feature space is used as an observation, and the corresponding output weight As a state, the weights are recursively estimated using a Kalman filter. The invention can effectively improve the generalization ability of the model, so that the prediction accuracy of the bus load is greatly improved.

Description

technical field [0001] The invention relates to a busbar forecasting method, which belongs to the technical field of power system load forecasting. Background technique [0002] Bus load can be defined as the sum of the loads supplied by the substation's main transformer to a relatively small supply area. Usually the type of busbar load in an area is relatively single. For different regions and uses, there are different basis for dividing user types: (1) urban civil load; (2) industrial load; (3) commercial load; (4) office load; (5) rural load, etc. The accuracy of bus load forecasting is an important content of grid security early warning and intelligent dispatching technology research, and its forecasting accuracy directly affects grid security prediction analysis, grid transmission capacity calculation, and operation planning. [0003] There are two main types of bus load forecasting methods: the first is the forecasting method based on system load distribution. This ...

AI Technical Summary

Problems solved by technology

[0006] In order to avoid the shortcomings of the above-mentioned prior art, the present invention provides a bus load prediction method based on a cascading generalization training strategy to overcome or improve the poor stability and low prediction accuracy of the existing single intelligent algorithm prediction model , For problems such as weak generalization ability of non-stationary data, thus effectively ensuring the safety and reliability of the power grid in the power supply area of ​​the substation

Images