Node load prediction method taking power grid topology constraints into consideration

Abstract

The invention discloses a node load prediction method taking power grid topology constraints into consideration. The idea of small-area estimation is introduced into a three-dimensional node load prediction system, small-area nodes existing in load prediction of a power system are pointed out, the relationship between observations is taken into consideration, a correlation equation is formed, and state variables are introduced into a three-dimensional load prediction model through measurement on the branches of the power system directly so as to improve the robustness of load prediction. The method of the invention takes power grid topology constraints into consideration, can effectively consider the containment relationship between load nodes of the power system and indirectly estimate the load of a small-area node with a small amount of sample effective information so as to improve the prediction of load prediction of small-area nodes. The model is a universal model which is not only applicable to small-area nodes, but also has a good effect for a normal system and can provide technical support for the intelligent development of power system dispatching.

Description

technical field [0001] The invention relates to the field of electrical engineering, in particular to a node load forecasting method for power system short-term operation regulation taking into account power grid topology constraints. Background technique [0002] Power system node load forecasting is the basis of short-term operation regulation of power system. It provides basis for power grid safety and economic dispatch, provides load trend data for power system safety warning, and provides a data source for detecting data mutations for system state estimation. Under the new situation, with the continuous advancement of the competition mechanism of the electricity market, the electric load is diversified in the form of electric vehicles and energy storage, and the fluctuation law of the node load is becoming increasingly complex, and the number of small load nodes with poor regularity is increasing. Traditional power system node load forecasting methods are facing challen...

AI Technical Summary

Problems solved by technology

[0006] The node load prediction described in the above patents has a good application effect when the load regularity of each node is good, but with the gradual popularization of wind power, photovoltaics, and electric vehicles in the power system, many problems with poor regularity have been formed. For small load nodes, the load forecasting of the power system faces two problems: one is that the load of each node can be predicted independently and the prediction accuracy basically meets the requirements of dispatching, but the results of multi-layer load forecasting are not coordinated; the other is that the load of some nodes can be accurately predicted , but another part of the node prediction reliability is low. In this context, if the traditional node load forecasting method is still used, it is bound to have a serious impact on the short-term operation of the power system and threaten the safe operation of the system.

Images