Voltage transformer fault identification method based on time sequence layered clustering

Abstract

The present invention discloses a voltage transformer fault identification method based on time sequence layered clustering. The method comprises the following steps of: (A) obtaining a voltage transformer measurement value to form M voltage time sequences; (B) performing noise reduction of the voltage time sequences; (C) setting a layered clustering threshold value; (D) calculating distances between each two of the M voltage time sequences; (E) combining two voltage time sequence classes with the smallest distance smaller than a threshold value into one, and calculating the distances of eachtwo of the (M-1) voltage time sequence classes; (F) successively, performing gradual iteration until all the voltage time sequences are combined to one voltage time sequence class or the smallest distance is larger than the layered clustering threshold value; and (G) identifying a fault voltage transformer according to a voltage time sequence layered clustering result. The voltage transformer fault identification method based on time sequence layered clustering can perform online identification of voltage transformer faults without restriction of outage plans, is suitable for all the types ofvoltage transformers without restriction of the types and the structures of the voltage transformers, and is high in universality.

Description

technical field [0001] The invention relates to the field of power system big data analysis, in particular to a voltage transformer fault identification method based on time series hierarchical clustering. Background technique [0002] With the continuous deepening of the intelligence of the power system, the power system perception data as its basis becomes more and more important. The measurement data of voltage transformer is one of the important sources of sensing data, and it is the prerequisite for realizing intelligent analysis and control of power system. Incorrect measurement data may not only lead to misoperation and refusal of automatic devices, but may also mislead dispatchers to make wrong decisions, seriously affecting the safe and stable operation of power systems. However, voltage transformer faults cannot be completely avoided. Therefore, in recent years, voltage transformer fault identification has been extensively studied, and many methods have been forme...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to provide a voltage transformer fault identification method based on time series hierarchical clustering. One is to solve the problem that the existing fault identification method based on off-line test is limited by the power outage plan, and it is difficult to accurately reflect the operation of the voltage transformer in operation. state; at the same time, it solves the problem that the existing voltage transformer fault identification method based on online monitoring data is only suitable for specific types and structures of voltage transformers, and has poor universality

Images