Power distribution network sag frequency estimation method based on optimized K nearest neighbor method and process immunity

Abstract

The invention discloses a power distribution network sag frequency estimation method based on an optimized K-nearest neighbor method and process immunity, which is characterized in that an optimal K-nearest neighbor parameter is searched based on a ten-fold cross validation method, and accurate distribution characteristics can be obtained for a line with relatively complex fault distribution; the defects that a traditional method is insufficient in approximation capability, too complex in model, poor in anti-interference capability and the like are overcome, and therefore a more objective and real voltage sag frequency estimation result is obtained. Based on the protection cooperation condition of a feeder automation system of a power distribution network, the voltage sag mode suffered by a user is analyzed and summarized, and for the defect that the influence of an existing method on voltage sag + short-time voltage interruption and continuous voltage sag is not considered sufficiently, a user process parameter change equation before and after aggregation is established and solved, so that the user sag mode is obtained. According to the method, the problems of under-estimation and over-estimation of the sag in the traditional method are avoided, and meanwhile, the defect that short-time interruption is not brought into statistics in the prior art is overcome.

Description

technical field [0001] The invention relates to the technical field of electricity quality evaluation, in particular to a distribution network sag frequency estimation method based on an optimized K-nearest neighbor method and process immunity. Background technique [0002] Precision equipment and industrial processes are often interrupted and tripped due to voltage sags, causing huge economic losses to users. Estimation of the frequency of voltage sags is the prediction of the single characteristics and times of possible voltage sags, and is an important basis for evaluating process interruptions and economic losses of sensitive industrial users. Existing methods can evaluate the frequency of voltage sags at the public connection point of the power grid, usually the busbar of a substation. However, for users directly connected to the busbar of the distribution network, it is difficult to directly apply the evaluation results to predict the consequences of industrial process...

AI Technical Summary

Problems solved by technology

[0014] In view of the above problems, the object of the present invention is to provide a distribution network sag frequency estimation method based on the optimized K nearest neighbor method and process immunity, avoiding the defects of insufficient approximation ability, overly complex model and poor anti-interference ability of the traditional method, thereby Obtain more objective and realistic voltage sag amplitude estimation results; at the same time, avoid underestimation and overestimation of such sags by traditional methods, and reduce the error of traditional frequency estimation results in evaluating the number of production interruptions

Images