Generator rotor turn-to-turn short circuit online monitoring method based on PCC-SVM

Abstract

The invention relates to a generator rotor turn-to-turn short circuit on-line monitoring method based on PCC-SVM, and the method comprises the steps: calculating the correlation among a plurality of operation parameters of a generator through a Pearson's correlation coefficient according to the operation historical data of the generator in a period of time, so as to obtain one or more types of characteristic parameters; adding a label indicating whether a turn-to-turn short circuit fault occurs to the characteristic parameters to obtain a training sample set; training a support vector machine model by using the training sample set data to obtain a fault discrimination model; and analyzing the to-be-analyzed operating parameters of the generator by using the fault judgment model to obtain the probability of the occurrence of the turn-to-turn short circuit fault, and if the probability of the occurrence of the turn-to-turn short circuit fault continuously exceeds a preset safety threshold within a period of time, giving an alarm prompt. The detection method provided by the invention does not need technical experience to analyze parameters, and is suitable for common field technicians.

Description

technical field [0001] The invention relates to the field of electrical equipment defect detection and early warning, in particular to a PCC-SVM-based on-line monitoring method for interturn short circuit of a generator rotor. Background technique [0002] Turn-to-turn short-circuit defects of rotor windings are one of the most common defects in turbogenerators. The resulting sudden failures not only disturb the production plan of the power plant, but also bring great pressure to the production operation. If the fault is not discovered in time, if it continues to develop, it will even cause serious faults such as grounding of the rotor winding and magnetization of the large shaft, resulting in more serious economic losses. Therefore, it is necessary to conduct on-line detection and evaluation of the inter-turn insulation state of large turbogenerator rotor windings, and to provide early warning of inter-turn short circuit defects. [0003] The accurate judgment of inter-tur...

AI Technical Summary

Problems solved by technology

However, when the traditional analytical method calculates the theoretical value of the excitation current, it does not consider the change of the saturation state of the internal magnetic field of the generator, and ignores the influence of the change of the synchronous reactance on the calculation of the excitation current value. The calculation of the excitation current is not accurate enough, which may lead to errors. High judgment rate

Images