QU Zhaoyang, XU Qi, QU Nan, et al. Online Calibration Method for Current Transformer Based on Compensated Modified Particle Swarms[J]. 2026, 46(7): 2939-2949.
DOI:
QU Zhaoyang, XU Qi, QU Nan, et al. Online Calibration Method for Current Transformer Based on Compensated Modified Particle Swarms[J]. 2026, 46(7): 2939-2949. DOI: 10.13334/j.0258-8013.pcsee.241965.
Online Calibration Method for Current Transformer Based on Compensated Modified Particle Swarms
Aiming at the problem of large calculation error of current transformer online calibration caused by the unsynchronization of metering devices and the volatility of line loss current
this paper proposes a current transformer online calibration method based on compensation correction particle swarm. Firstly
a current transformer online calibration model based on the derivation of the error relationship is constructed
and the factors affecting the error calculation are considered from the dimensions of data
error relationship and solution to realize the online calibration of the current transformer; secondly
an adaptive sliding window preprocessing method for the average current data is proposed
and by introducing the relative rate of change of the current and the threshold of the sliding window
the average current of the smooth interval is calculated
and the problem of sudden change of current data is solved; then
the method of compensating correction particle swarm based on the online calibration of the current transformer is proposed. Next
an error relationship derivation method is given to account for the periodicity of the current
and on this basis
an error solving algorithm based on compensation correction particle swarm is designed to suppress the effect of loss current volatility on the error solving by using the particle swarm algorithm with constant compensation and remove the anomalies and center-of-mass-weighted correction of the error
so as to get the error results of the current transformer; finally
the validity of the model is verified by case analysis. Finally
the validity of the model is verified by case analysis.