Distributed power generation island detection improvement method based on data mining

Abstract

The invention belongs to the technical field of power grid fault diagnosis, and particularly relates to a distributed power generation island detection improvement method based on data mining, which adopts a meta-learning mode in machine learning to determine a threshold value of island detection, and the method comprises the following steps: step 1, adopting a RELIEF algorithm to identify key features of island detection before classification, wherein the key features comprise steady state quantity features and transient state quantity features; and step 2, performing a classification algorithm based on the base learner and the meta learner, and implementing distributed power generation island detection. The step 2 comprises the following steps: step 21, data stream mining: adopting an increment mode, and considering time and space efficiency of the algorithm; step 22, meta learning: utilizing complementarity of different classifiers to improve adaptability of data mining and machine learning; and step 23, online self-learning: adopting a sliding data window and a multi-classifier integration method to realize online self-learning, and keeping the classification precision at a relatively high level all the time.

Description

technical field [0001] The invention belongs to the technical field of power grid fault diagnosis, and in particular relates to an improved method for detecting islands of distributed power generation based on data mining. Background technique [0002] An important issue in distributed generation systems is the islanding detection problem. An isolated island means that after a power grid failure, the switch between the distributed generation system and the large power grid is disconnected, resulting in a local load and local DG with a certain capacity, forming an isolated island that operates independently, and the power company cannot control it. It is generally believed that all distributed power supplies must be disconnected immediately when an island occurs, because DG generally operates in the current control mode, and the formation of an island will cause the voltage to be unstable, damage equipment, and bring safety hazards to maintenance personnel. However, with the...

AI Technical Summary

Problems solved by technology

However, the simulation model used is quite different from the actual situation.

[0006] 3. Using the method of data mining, the process of obtaining the threshold value of islanding by using specific feature criteria, however, the selection of feature quantities is relatively conventional, the classification method is not optimal, and the robustness of a single classification algorithm is relatively weak

[0010] 2. Using a single classifier without considering the phenomenon of inductive bias

This necessary assumption about the objective function is called inductive bias. In the past, in the study of power systems in data mining, the influence of single classifier inductive bias on the classification prediction results was rarely considered. Because of this limitation, excellent algorithms will be limited in some case becomes unused

[0012] 3. The off-line classifier training method is adopted, without considering the phenomenon of concept drift

[0014] Therefore, for the active island detection scheme of the transmission, although the detection accuracy is high and the detection blind area is small, it usually has a negative impact on the power quality of the power grid and the local microgrid; the passive detection method is simple to implement, but there are blind areas and weaknesses that cannot be identified

Although the blind area is small, the switch detection method is expensive, complex in design, and requires high communication reliability, so it has not been widely used in distributed power generation systems.

Images