Pilot protection method, device and storage medium

Abstract

The present application provides a longitudinal protection method, device and storage medium, which acquire time-domain signal data of a target element at a preset sampling frequency; perform fusion processing on the time-domain signal data of multiple first sampling periods to obtain a first time domain domain signal combination data; based on the machine learning model, determine whether the target element is faulty according to the first time domain signal combination data; when it is determined that the target element is faulty according to the first time domain signal combination data, based on the machine learning model, according to the second time The domain signal combination data determines whether the target element is faulty in the second sampling period, and the second sampling period is the sampling period after it is determined that the fault occurs; and when it is determined that the target element has the same type of fault in a plurality of consecutive second sampling periods, Control the pilot protection system to perform the protection action on the target element. Using machine learning models to achieve fault identification of power system components to significantly improve the protection performance of smart grids.

Description

technical field [0001] The present application relates to the technical field of power system relay protection, and in particular, to a pilot protection method, device and storage medium. Background technique [0002] With the ubiquitous access of information technology, information sharing has become a prominent feature of smart grids. Using the time-domain signal of the power acquisition information shared in the smart substation and the smart grid to directly perform data fusion to describe the physical phenomenon. [0003] Existing longitudinal protection methods such as proportional braking differential protection have been used in engineering for many years, and have good application effects in traditional power grids, but need to extract frequency domain components, and also rely on setting values ​​and complex criteria. In the future smart grid, considering the extensive access of large-scale intermittent energy sources and power electronic equipment, the operation ...

AI Technical Summary

Problems solved by technology

[0003] Existing longitudinal protection methods such as proportional brake differential protection have been used in engineering for many years, and have good application effects in traditional power grids, but they need to extract frequency domain components and depend on setting values ​​and complex criteria

In the future smart grid, considering the wide access of large-scale intermittent energy sources and power electronic equipment, the operation of the power system will face various uncertainties, showing the aliasing of dynamic processes with various time constants, resulting in the distortion of system fault characteristics. Huge changes, which can make it difficult for the protection function under the existing design mode to have sufficient coping capacity

Images