Feature extraction method based on fractional order Hilbert cepstrum

Abstract

The invention particularly relates to a feature extraction method based on fractional order Hilbert cepstrum. The method comprises the steps: S1, collecting current data generated when different electrical loads operate independently; S2, carrying out the windowing preprocessing on the collected current data; S3, carrying out the fractional order Hilbert transformation on the processed current data, and mapping the data to a fractional space; S4, optimizing the fractional order Hilbert transform order, and determining an optimal order; S5, performing calculating to obtain cepstrum characteristics of different electrical loads under the optimal order; S6, substituting the fractional order cepstrum characteristics of the different electrical loads under the optimal order into a support vector machine for load identification to obtain identification rates of the different electrical loads. According to the fractional order cepstrum characteristics provided by the invention, the recognition rate of different electrical loads is effectively improved, and a better classification effect is achieved under the condition that the load characteristics are similar.

Description

technical field [0001] The invention relates to a non-invasive load feature extraction method, in particular to a feature extraction method based on fractional Hilbert cepstrum. Background technique [0002] Non-intrusive load monitoring (Non-intrusive Load Monitoring, NILM) technology can obtain refined user internal load category and usage status data by decomposing and identifying the total load data of users, which is an effective solution to the problem of intelligent power load monitoring. way. Non-intrusive load monitoring does not require the installation of load monitoring devices on the user side or substantial upgrades to smart meters. It relies on the data acquisition devices and communication networks of the existing power consumption information acquisition system, and uses advanced data communication technology to obtain refined user information. Electric load data, and then use the powerful data processing capabilities of the electric power cloud platform to...

AI Technical Summary

Problems solved by technology

However, when the characteristics of the electric load are similar, it cannot bring the ideal recognition effect

[0003] Title of prior art document: A non-intrusive load monitoring method based on cepstrum analysis Published to: ("Electronic Technology", 2018, discloses the use of traditional cepstrum analysis to monitor the load of multiple electrical appliances and a single electrical appliance when they are running separately identification, but there is no research on the identification of loads with similar power

Images