|
1.INTRODUCTIONThe errors in current transformers are mainly caused by the excitation current required to build up the flux in the core. The non-linear nature of the excitation current makes it difficult to reduce it in order to improve the accuracy of the current transformer [1-3]. The current error in uncompensated current transformers is mostly negative, and appropriate compensation methods can reduce the ratio difference and phase difference. Current transformer error compensation methods can be divided into two categories: active compensation and passive compensation. Passive compensation is not widely used because it cannot meet the needs of wide range measurement [4-6]. The active compensation method reduces the excitation current in the primary winding by providing magnetic potential or electromotive force to the compensated current transformer, thus reducing the error of the current transformer, which can achieve the purpose of wide range measurement [7-8]. In this paper, under the condition that the adiabatic performance, metering performance and cost of the current transformer for wide-range measurement are fully considered, the adaptive control idea is introduced by means of segmentation and dimensionality reduction to track the change of the error and further adjust the gain of the compensation signal to achieve closed-loop measurement, which makes the measurement effect more ideal. 2.1Segmentation and dimensionality reduction methodsPiecewise aggregate approximation (PAA) [9-10] is a data dimensionality reduction method, the core idea of which is to evenly divide the time series and use the segmented mean to approximate the original series to achieve the purpose of data dimensionality reduction. For a given segment of sequence X of length n: The sequence after division into w segments is represented as: where In this paper, based on the PAA algorithm, a new nonlinear model is constructed by introducing the hyperbolic tangent function as a moving influence factor, and the expression of the hyperbolic tangent function is as follows: f(x) is monotonically increasing in the domain of definition, the function value tends to 1 infinitely, has a fast convergence rate and accuracy, and can satisfy the condition of moving the influence factor. After the introduction of the moving factor, in order to divide the sequence Si (Si = {si,1, si,2⋯ si,n}) of length n into sequences of length N, a sliding window of size ω is set, and ω is defined as follows: Placing the sliding window at the front of the sequence Si, the sliding window of size ω is moved along the direction of the time axis, and the product of the mean of each window and f(x) is calculated to obtain the index vector according to the direction of the time axis as follows: The sequence after the improved PAA algorithm division becomes: 2.2Adaptive control of the compensation currentIn this paper, gain adaptive control is used as a way of filtering the processed signal to obtain the fundamental error compensation signal e(t). The phase of e(t) is adjusted using a phase shifter so that it is in the same phase as the standard output voltage Ym of the reference model. To achieve error-free rectification, a multiplier is used to obtain a DC signal of (Ym − Y) sign(Ym), which corresponds to an adjustable gain Kc. Multiplying Kc with the compensation current achieves adaptive adjustment of the parameters. The overall structure of the adopted active compensation method for current transformers is shown in Figure 1. 2.3Circular magnetic split compensationIn the toroidal core, the magnetic split can be installed on both sides of the core, but the structure is more complex, can be made into a ring installed in the periphery of the core, its schematic and schematic diagram as shown in Figure 2, the magnetic split is circular, so called circular magnetic split compensation, circular magnetic split by cold-rolled silicon steel strip rolled, the cross-sectional area is very small, generally only a few pieces of silicon steel. The compensating structure of the circular magnetic tap is to wind the primary and secondary windings on the main core T1, while on the magnetic tap T2, the secondary winding is wound with less Nb turns, Nb is called the number of turns of the magnetic tap compensation. The ampere-turns of the excitation current in the circular magnetic tap are: where Hb is the magnetic field strength of the magnetic split, and lb is the average magnetic circuit length of the magnetic split. The secondary induced electric potential of the circular magnetic split is here Bb is the magnetic splitting density, Sb is the magnetic splitting cross section and k is the core stacking factor. 3.WIDE RANGE CURRENT TRANSFORMER CALIBRATION DEVICESThe current transformer calibration device for wide range metering consists of regulating power supply, current boosting unit, standard current transformer, transformer calibrator, current load box, etc. Through the regulating power supply, current boosting unit and high current conductor to build a primary circuit, the current transformer for wide range metering and standard current transformer will be connected to the primary test circuit in series. The secondary winding of the low-voltage current transformer for wide-range metering is connected to the current load box and the transformer calibrator, and the standard current transformer is connected to the transformer calibrator. Among them, the primary winding of the primary winding parallel principle as shown in Figure 3, will be P1C2 and C1P2 parallel, the total current 2I1 from the P1 end into, and then divided into 2 branches into the current transformer, and finally from the P2 end summary out, each primary winding current are I1, the secondary winding current is I2. for the primary winding parallel current transformer, the total primary winding is 1 turn, the total current flowed for 2I1, so the ratio in this case is the actual number of turns of the secondary winding of the current transformer. Let the number of turns of the secondary winding of the current transformer be N2, then the ratio is: The error measurement module uses the vector decomposition method, which means that the difference current between the transformer under test and the standard transformer is directly collected, vector decomposed and calculated to obtain the error of the transformer under test. The error measurement module uses electronic and computer technology, which is highly automated and simple to operate. The basic principle of the error measurement unit is shown in the following block diagram. The unit consists of an analog circuit part and a digital circuit part. The analog circuit part mainly completes the signal processing of differential current, differential pressure and dial indicator, so as to separate the differential current and differential pressure signals in phase and quadrature. The digital circuit part mainly completes the A/D conversion of the processed signal, and processes the converted data, so as to obtain the final measurement result as shown in Figure 4. 4.EXPERIMENTAL VALIDATION AND ANALYSISThe information on the low-voltage current transformers for wide-range metering used in this paper is as follows. Rated ratio: 500/5 Operating current range: 0.1%-200% of rated current Rated load: 5VA Accuracy class: 0.2SS Its error limit table is shown as follows in Table 1. Table 1.Error limits.
Using the proposed adaptive error compensation method, the experimental results are shown in Table 2. Table 2.Test results for low voltage current transformers for wide range metering.
As can be seen from the table, the error characteristics, insulation structure and accuracy level of the wide metering low-voltage current transformer studied in this paper meet the error limits under different load conditions, it can also be seen that after the error compensation by the method proposed in this paper, the error of the low-voltage current transformer for wide range metering meets the 0.2SS level under both light and full load conditions. 5.CONCLUSIONThe adaptive error compensation method for wide-range current transformers using segmental dimensionality reduction can achieve adaptive adjustment of the compensation current gain, while based on the segmental dimensionality reduction method to achieve the purpose of dimensionality reduction for high-dimensional, massive signal sequences, effectively widening the range of the transformer and improving the measurement accuracy of the transformer. The test shows that this method can be used to make low-voltage current transformers for wide range metering meet the error level of 0.2SS under both light and full load conditions. ACKNOWLEDGMENTSThis study has been supported by the Foundations of State Grid Jiangsu Electric Power Co., Ltd under grant No. J2021209 (Research, development and application of wide range current transformer and its verify unit). REFERENCESXiao, X., Song, H. and Li, H.,
“A high accuracy AC+DC current transducer for calibration,”
Sensors, 22
(6), 2214
–2214
(2022). https://doi.org/10.3390/s22062214 Google Scholar
Ai, Y., Shi, Y., Chen, S., Zhang, Z., Zhao, S. and Liu, J.,
“High-voltage isolated multiple output auxiliary power supply with output voltage self-regulation,”
Energies, 15
(6), 2106
–2106
(2022). https://doi.org/10.3390/en15062106 Google Scholar
Chen, K., Ren, S., Guo, Y., Chen, N. and Lee, W.,
“A redundancy mechanism design for hall-based electronic current transformers,”
Energies, 10
(3), 312
(2017). https://doi.org/10.3390/en10030312 Google Scholar
Bezerra, F. V. V., Cavalcante, G. P. S., Barros, F. J. B., Tostes, M. E. L., Bezerra, U. H.,
“Methodology for predictive assessment of failures in power station electric bays using the load current frequency spectrum,”
Energies, 13
(19), 5123
–5123
(2020). https://doi.org/10.3390/en13195123 Google Scholar
D’Avanzo, G., Femine, A. D., Gallo, D., Landi, C. and Luiso, M.,
“Impact of inductive current transformers on synchrophasor measurement in presence of modulations,”
Measurement, 155
(C), 107535
–107535
(2020). https://doi.org/10.1016/j.measurement.2020.107535 Google Scholar
Ballal, M. S., Wath, M. G. and Suryawanshi, H. M.,
“Measurement current transformer error compensation by ANN methodology,”
Journal of The Institution of Engineers (India): Series B: Electrical, Electronics & Telecommunication and Computer Engineering, 101
(3), 261
–271
(2020). Google Scholar
Su, Q., Quan, W., Cai, G. and Li, J.,
“An improved adaptive backstepping approach on static var compensator controller of nonlinear power systems,”
International Journal of Adaptive Control and Signal Processing, 32
(5), 700
–712
(2018). https://doi.org/10.1002/acs.v32.5 Google Scholar
Ueno, S., Enokizono, M. and Inoue, S.,
“Vector B-H tester for simple evaluation of vector magnetic characteristics,”
Journal of Electrical Engineering, 69
(6), 407
–410
(2018). https://doi.org/10.2478/jee-2018-0061 Google Scholar
Fotso, V. S. S., Nguifo, E. M. and Vaslin, P.,
“Grasp heuristic for time series compression with piecewise aggregate approximation,”
RAIRO-Operations Research, 53
(1), 243
–259
(2019). https://doi.org/10.1051/ro/2018089 Google Scholar
Ren, H., Liao, X., Li, Z. and Ahmari, A. A.,
“Anomaly detection using piecewise aggregate approximation in the amplitude domain,”
Applied Intelligence, 48
(5), 1097
–1110
(2018). https://doi.org/10.1007/s10489-017-1017-x Google Scholar
|