Data-driven regression method for load flow of power system based on ELM-integrated online learning

Abstract

The invention provides a data-driven regression method for load flow of a power system based on ELM-integrated online learning, belonging to the technical field of new energy power system load flow calculation based on data driving. The method solves the problems that under large-scale access of new energy, power grid load flow calculation depends on an accurate model and parameters of a system, a load flow calculation result needs to be obtained through multiple iterations, a topological structure and model parameters of the system are inaccurate, and load flow iteration solving consumes considerable time in the prior art. According to a technical scheme in the invention, a power system load flow mapping model based on ELM-integrated online learning is constructed through a large amount of load flow time sequence section data. The method has the advantages that a new-energy-containing power distribution network system can rapidly solve system load flow without depending on an element model and the parameters of a power grid, a system operation load flow state can be rapidly sensed, and the method is suitable for analyzing a power grid operation state in the scene of frequent power fluctuation of a new-energy-containing power grid.

Description

technical field [0001] The present invention relates to the technical field of data-driven new energy power system power flow calculation, in particular to a power system power flow data-driven regression method based on Extreme Learning Machine (Extreme Learning Machine, ELM) integrated online learning. Background technique [0002] In order to promote the healthy development of the energy system, the new power system with new energy as the main body will become the main form of the future power grid. With the rapid growth of new energy installed capacity and extensive grid connection, system power fluctuations are more complex and diverse. It is necessary to quickly solve the system power flow, clarify the system operating status, and avoid safe operation accidents. [0003] Traditional power flow calculations rely on accurate models and parameters of the system, and require multiple iterations to obtain power flow calculation results. Inaccurate system topology and model ...

AI Technical Summary

Problems solved by technology

[0010] The purpose of the present invention is to provide a power system power flow data-driven regression method based on ELM integrated online learning, which solves the problem that the power flow calculation of the power grid under the large-scale access of new energy depends on the accurate model and parameters of the system, and requires multiple iterations to obtain the power flow calculation results , the system topology and model parameters are inaccurate, and the power flow iterative solution is time-consuming. It realizes the fast solution of the system power flow containing the new energy distribution network system that does not depend on the grid component model and parameters, and quickly perceives the power flow state of the system.

Images