Power system transient frequency prediction model updating method based on data inheritance idea

Abstract

The invention provides a power system transient frequency prediction model updating method based on a data inheritance idea. The method comprises the following steps of firstly, training a historicaltransient fault sample by utilizing an extreme learning machine (ELM) algorithm, and establishing a historical frequency prediction model; and if the power system has a transient fault during the process of monitoring the running state of the system in real time, collecting the system running data before and after the fault, and then updating the historical frequency prediction model according toa newly added transient fault sample. According to the method, the data inheritance thought is applied to the frequency prediction model updating process, the frequency prediction model can be rapidlyupdated in a power system transient frequency prediction scene with few historical samples, the frequency prediction performance is improved, the emergency control measures can be made and adopted intime, and the method has important significance for maintaining the stable operation of a power system.

Description

technical field [0001] The invention relates to the technical field of power system situation awareness and trend prediction, in particular to a power system transient frequency prediction model update method based on the idea of ​​data inheritance. Background technique [0002] With a large number of new energy sources such as photovoltaics and wind power connected to the grid and gradually replacing conventional power sources, the equivalent inertia of the system is gradually reduced, resulting in more serious system transient problems, especially transient frequency safety problems, which is not conducive to the safe and stable operation of the power system. Rapid prediction of power system transient frequency is of great significance for timely taking emergency control measures and improving the safety and stability of power system. [0003] In recent years, artificial intelligence and machine learning have developed rapidly. Machine learning methods can effectively make...

AI Technical Summary

Problems solved by technology

However, based on the traditional machine learning method to predict the transient frequency of the power system, all the historical data need to be used when the model is updated, resulting in repeated training of a lot of historical data, which makes the model update efficiency inefficient

Images