A Method for Extracting Fault Signal Eigenvalues ​​in the Process of Microgrid Fault Diagnosis

Abstract

The invention discloses a fault signal feature value extraction method during micro power grid fault diagnosis process and belongs to the field of micro power grid fault diagnosis. According to the method, on the basis of the changes of signal features before and after a fault and with the help of extreme value symmetrization processing, fault signals are detected from micro power grid voltage signals or current signals, the signal fluctuation is inhibited, fault signal features are kept, and the fault signals are normalized to idealize the fault signals to be processed. Then important feature values of the fault signals after normalization are extracted from multiple points of views, wherein the important feature values comprise the main frequency band energy values of the fault signals. Finally, self-adaptive feature value dynamic optimization method is adopted to automatically, multi-layeredly, and optimally process micro power grid fault feature values and provide stable fault feature values for later micro power grid fault detection process.

Description

technical field [0001] The invention belongs to the field of micro-grid fault diagnosis, in particular to a method for extracting characteristic values ​​of fault signals in the micro-grid fault diagnosis process. Background technique [0002] With the rapid development of the smart grid, there are more and more uncertain connections of distributed power sources, and the emergence of various types of load transformations and different controllers make the fault information of the micro-grid more and more complicated. Accurate diagnosis is increasingly difficult. Especially in recent years, these problems have become more prominent and become a very valuable research hotspot. [0003] The microgrid fault diagnosis process can be mainly divided into the extraction of fault eigenvalues, fault judgment and location. Among them, the extraction of fault eigenvalues ​​is the most important and the most difficult part. In the research of related fault methods at home and abroad, i...

AI Technical Summary

Problems solved by technology

[0004] Many existing fault signal eigenvalue extraction methods are based on ideal microgrid fault signals. However, there are many difficulties in obtaining ideal microgrid fault signals. Due to the various requirements of the actual microgrid, there are many controllers, which will also have many effects on the fault signal, so it is impossible to obtain the required ideal microgrid fault signal

Secondly, the sampling accuracy of the signal and various small disturbances will affect the accurate collection of the fault signal, so many existing fault signal eigenvalue extraction methods are difficult to be practical

Moreover, some existing fault signal eigenvalue extraction methods only extract a single fault signal eigenvalue, and due to the interference of small signals, it is difficult to accurately identify the fault, and sometimes wrong judgments occur

Images