Load identification method based on electric power fingerprint features and integrated learning mechanism

Abstract

The invention discloses a load identification method based on electric power fingerprint features and a Stacking integrated learning mechanism. The method comprises the following steps: acquiring voltage and current high-frequency waveform data of various electric appliances of a user family; firstly, using various time domain and frequency domain analysis methods to calculate various electric power fingerprint steady-state characteristics of a load; secondly, screening the most representative features out through an FCBF method; and finally, constructing a load identification model fusing a plurality of machine learning methods by using a Stacking mechanism. According to the method, the idea of integrated learning is fully utilized, the advantages of various machine learning methods are effectively combined, compared with a single-model method, the model recognition precision is improved, and the method has the advantages of being high in accuracy, good in generalization performance and capable of flexibly fusing various models.

Description

technical field [0001] The invention relates to the field of load identification and classification, in particular to a load identification method based on power fingerprint features and an integrated learning mechanism. Background technique [0002] Most of the existing load identification technologies use traditional machine learning methods such as k-nearest neighbor algorithm, decision tree, support vector machine, etc. With the development of integrated learning, the research on integrated learning ideas such as bagging, boosting, and stacking in the field of load identification is also increasing. widely. At present, the research on feature fusion in the field of load recognition has become a hot spot, but the combination of multiple features is prone to feature redundancy, which may reduce the performance of the model. Contents of the invention [0003] The technical problem to be solved by the present invention is to provide a load identification method based on p...

AI Technical Summary

Problems solved by technology

[0003] The technical problem to be solved by the present invention is to provide a load identification method based on power fingerprint features and an integrated learning mechanism to solve the problem that the prior art load identification adopts multi-feature combination and is prone to feature redundancy, which may reduce the model performance instead. and other technical issues

Images