Capacitive voltage transformer (CVT) is an essential power measurement equipment in the power grid, which generates errors in long-term operation. Therefore, it is necessary to quantify the measurement performance of CVT and predict its measurement deterioration trend. This study proposes a measurement performance index to characterize the ratio error of CVT and a trend prediction method for CVT measurement deterioration based on double Gaussian model-KAN fusion. First, approximation and detail coefficients are formed after multilayer wavelet transform on the secondary side voltage of CVT. The maximum approximate coefficient is selected from the approximate coefficients, and the State of Performance (SOP) representation ratio error of CVT is calculated using the maximum approximate coefficient. Then, a Variational Modal Decomposition Mean Difference (VMD-MD) method is proposed to decompose the SOP sequence of CVT in multiple layers. The residual decomposed from the SOP sequence is used to characterize the deterioration trend of SOP, and the double Gaussian model is used to model and predict it. The Intrinsic Mode Functions (IMFs) decomposed from the SOP sequence are used to characterize the deterioration fluctuation of SOP, and the KAN algorithm is used to predict it. Finally, all the predicted results are added to represent the deterioration trend of CVT. Using CVTs and SWCVT-3 CVT online test system of China Electric Power Research Institute, the three-phase voltage data with increasing ratio error are collected, and the proposed double Gaussian model-KAN fusion method is tested. During the experiment, the ratio error of CVT was characterized effectively by SOP, and the proposed double Gaussian model-KAN fusion method could accurately predict the CVT SOP deterioration trend.

In this paper, the problems of high insulation requirement and high cost of the traditional method in the measurement of MV-level impact voltage, as well as the easy aberration of the electric field caused by the spherical electric field measuring instrument are investigated. Firstly, the effects of spherical sensors on the distortion of electrostatic field and shock electric field are analyzed by simulation. In addition, a shock high-voltage standard field source device was developed for scale-quoted calibration. The device in the effective space can realize the field strength inhomogeneity $< 1.5 \%$ of the impact uniform field. The standard shock field calibration test shows that the improved measurement system exhibits excellent linearity characteristics (linearity better than $\pm 1.4 \%)$ under 2400 kV shock voltage. This significantly improves the accuracy and engineering applicability of high-voltage transient field strength measurements.

The growing demand for high-precision sensing technologies in power systems has po-sitioned the development of novel electric field measurement tools as a critical metro-logical challenge. This study presents a broadband photoelectric electric field sensor based on the Mach-Zehnder interference phenomenon in asymmetric straight wave-guides. Furthermore, a three-dimensional electric field measurement system was de-veloped through systematic investigation of multidimensional field measurement methodologies. Experimental results demonstrate that the proposed device achieves broadband electric field measurement with linearity up to 0.9996 and pow-er-frequency measurement accuracy exceeding 97%. The system demonstrates con-current measurement capability for both lightning impulse waveforms and operation-al power frequency electric fields. The synthesized three-dimensional measurement error remains below 4%, enabling precise omnidirectional electric field detection.

Efficient excitation of nitrogen-vacancy (NV) centers in diamonds and compelling collection of their fluorescence are essential for improving the performance of optical fiber-based NV center sensors. In this study, we obtained the optimized structural parameters of the tapered optical fiber with a tapered tip diameter of 90 mu m and a taper angle of 35.5782 degrees through ray tracing simulation, which significantly increased the fluorescence collection efficiency of the reflective all-optical fiber system to 5.6827%, which is approximately 9.02 times higher than that of the ordinary multimode fiber-coupled diamond, thus increasing the sensitivity of the fiber- integrated diamond magnetometer. Software simulations and numerical modeling demonstrate the optimized tapered fiber optic probe's high fluorescence excitation and collection capability. Experimental results confirm that this configuration improves fluorescence collection efficiency by more than 7.38 times and sensitivity by 4.14 times. The enhanced fluorescence collection directly contributes to the sensitivity of the fiber optic-based diamond magnetometer.

The current measurement method in the power industry today mainly relies on current sensors. With the development and progress of modern technology, the power system is facing unprecedented challenges. Traditional current sensing technology is difficult to adapt to the pace of the times, and the new generation of current sensing technology is constantly evolving. More and more attention by all walks of life and has been widely used. Tunnel magneto resistance current sensor has the characteristics of high linearity, high sensitivity, low cost, simple structure, etc., so it has become a very potential sensor product. Taking the current sensor developed by tunnel reluctance technology as the core, this paper introduces the working principle of tunnel reluctance current sensor, the advantages and disadvantages of various structures and their application fields, and the improvement methods of different defects. Furthermore, it summarizes the research status and future prospects of sensor array.

The calibration of ultra-high voltage direct current (UHVDC) transmission systems, as an advanced electric power system that transmits electric energy with high efficiency and over long distances, is crucial to ensure the safe and stable operation of the system. However, due to the complexity and high technical requirements, it faces difficulties in calibration accuracy, calibration methods, and equipment. For this reason, this article proposes a digital twin model containing a physical system, data interaction, and a digital system, which adopts a two-layer feature extraction structure to recognize key devices through parameter perception. Then, the feature information of key devices is parametrically identified and the relative topological features are learned. Experimental results show that mean squared error, root mean squared error, and mean absolute error indexes are reduced by 71.08%, 46.32%, and 64.39%, respectively, and the R2 score is improved by 3.8% compared with the original method. It provides an important guarantee for the safe operation and reliability of the UHVDC system.

Impulse current measurement technology is widely used in various applications, including lightning protection monitoring in power systems, welding current measurement in aircraft and shipbuilding industries, as well as high-current measurement in pulsed power systems. With the advancement of industrial technology, the measurement range of impulse currents has continuously expanded, reaching levels as high as mega-amperes (MA). The calibration of the scale factor for impulse current measurement devices is determined through comparison with standard measurement devices. Developing high-accuracy impulse current measurement devices and accurately judging their characteristics are prerequisites for ensuring the precise calibration of impulse current values. This paper introduces two different types of high-impulse current measurement devices. Experimental studies were conducted on the scale factor and response characteristics of the sensors. The scale factor extension calibration method for sensors under high currents of more than 100 kA has also been introduced. Test results indicate that the developed impulse current measurement devices can serve as standard measurement devices for high impulse current measurement.

The measurement accuracy of current transformers is crucial for power system protection and trade fairness. The high penetration of renewable energy into the power grid has affected the transient performance of power systems, posing significant challenges for accurate current transformer measurement. To address this issue, this paper proposes a prediction model for transformer measurement accuracy based on an adaptive dual-modal decomposition strategy and a hybrid deep learning architecture. The framework integrates an enhanced Adaptive Time-Varying Filter (A-TVF), an enhanced Adaptive Variational Mode Decomposition (A-VMD), the Residual Error Index (REI), and the Maximum Information Coefficient (MIC). First, A-TVF preprocesses the collected data by setting REI as the optimization objective to adaptively adjust filter construction parameters, including the B-spline order, bandwidth threshold, and decomposition number, and decomposes the collected ratio error sequence to reduce the non-stationarity of the original sequence. Subsequently, indices such as PE and Kurt are used to screen the decomposed sub-sequences and reconstruct the complex components. Then, A-VMD is applied to further decompose the complex components, minimizing MIC by adaptively determining the decomposition number, penalty factor, convergence accuracy, and fidelity parameters. Afterward, the complexity of the subcomponents obtained from the secondary decomposition is calculated, and the entire sequence is reconstructed. Finally, a hierarchical prediction model integrating Temporal Convolutional Networks (TCN), Bidirectional Gated Recurrent Units (BiGRU), and a Multi-Head Attention mechanism (MHA) is employed to predict the reconstructed components and generate the final results. Experimental results demonstrate that the proposed adaptive dual-modal decomposition method significantly improves prediction performance: compared with non-decomposition models, RMSE, MAE, and SMAPE were reduced by an average of 50.12%, 46.09%, and 37.70% in global decomposition scenarios, and by 25.92%, 23.69%, and 19.96% in rolling decomposition scenarios, respectively. These results validate the effectiveness of the proposed method in reducing data complexity and improving the accuracy and stability of Ratio Error predictions.

This study introduces an all-fiber integrated nitrogen-vacancy (NV) center quantum sensor system for high-precision current sensing. By optimizing the structural parameters of the tapered optical fiber probe, we achieved a fluorescence collection intensity 7.1 times higher than that of a conventional multimode fiber-coupled system. Experimental results demonstrate that under optimal laser and microwave conditions, the magnetic field sen\/ sitivity reaches 1.83 nT/ Hz under the actual operating condition. Current testing shows that the R 2 value of the magnetic field, calculated from the resonance frequency splitting and the tested current, is 0.997, with a +/- 0.7 % instability, indicating stable and precise current measurements. The full fiber integration of the system, achieved through fusion splicing, results in a compact and miniaturized design, making it ideal for outdoor current detection applications.

Traceability of alternating current (AC) quantum voltmeters (QVM) based on the programmable Josephson voltage standard (PJVS) is hampered by the stability of standard AC sources, with their critical boundary conditions remaining uncharacterized. we address this limitation by quantifying how AC source stability —including amplitude/phase jitter, noise, and frequency deviation—affects measurement errors of AC QVM, demonstrating that amplitude errors of AC QVM below 5 μV can be achieved under the characteristics of conventional power signals—such as an amplitude jitter and noise of 1 mV, a phase jitter of ±1°, and a frequency deviation of ±0.2 Hz. A theoretical error model is developed and the LabVIEW simulation is carried out to verified that issue. Experimental validation confirms that AC QVM can meet the requirements of AC voltage traceability applications in certain complex signal scenarios, advancing their practical utility beyond controlled laboratory environments.

Accurate current measurement is the basis for power system state sensing and control, and the application of tunneling magnetoresistance (TMR) in current measurement is gaining more attention due to its high sensitivity and low power consumption. However, the interference of the application environment brings temperature drift and nonlinearity problems to the TMR sensor. This paper proposes a signal processing method based on adaptive wavelet threshold denoising, and then designs a TMR current sensor with antiinterference capability. The signal output from the TMR sensor is subjected to wavelet denoising by adaptive wavelet thresholding, which effectively improves the signal-to-noise ratio of the signal and significantly reduces the mean square error. Simulation and measured data show that the current measurement error is 0.05 %, which effectively reduces the nonlinear error compared with other signal processing methods. This method can significantly improve the measurement accuracy of TMR sensors in complex environments, providing an effective technical support for high-performance current monitoring.

In order to solve the problem of transient anomaly identification on the DC side of distributed photovoltaics, a measurement device that can measure the superimposed transient voltage of DC voltage at low voltage was developed in this paper, and its measurement method was proposed. In this paper, through theoretical analysis, the circuit model is built based on the resistor-capacitance parallel voltage divider model, and further simulation is carried out to obtain the response characteristics of the measurement device. Secondly, the measurement method of the measuring device is introduced. Finally, the measuring device is calibrated, and it is obtained that the error of measuring the DC signal is within ±0.5%, and the measurement error of the transient voltage signal is better than ±1%. It can be used to measure the transient voltage signal of DC voltage superposition to realize the transient anomaly identification of photovoltaic DC side.

To address the issue of inaccurate high-frequency voltage signal measurements in the military and power industries, and to further enhance the measurement capabilities of high-frequency pulse voltage, this paper proposes a capacitive voltage divider technique based on wave impedance matching. First, it introduces the theory and role of wave impedance matching in measurement circuits. Next, the design process of a capacitive voltage divider based on wave impedance matching is described, taking into account the distributed parameters of the divider and conducting circuit simulations. Finally, a step wave response test is conducted on the developed 200 kV capacitive voltage divider to verify its wide-frequency measurement capabilities. The results show that the capacitive voltage divider designed with wave impedance matching has a step wave stabilization time of less than 200 ns in the response test. The step wave quickly stabilizes after passing through the voltage divider and matches the theoretical calculations. Compared to a voltage divider not designed with wave impedance matching, the stabilization time improves by at least 100 ns, achieving accurate measurement of higher frequency transient voltage signals.

The measurement accuracy of voltage transformers (VTs) is crucial for power system protection and trade fairness. However, the large-scale integration of renewable energy into the grid affects the transient performance of power systems, presenting significant challenges for the accurate measurement of VT ratio errors. This paper presents a hybrid prediction framework incorporating a layered signal decomposition and reconstruction approach, along with feature selection and data augmentation, to forecast transformer ratio errors under limited data conditions. First, an energy entropy-optimized adaptive variational mode decomposition method is developed, which introduces convergence constraints for parameter selection, reducing randomness during the decomposition process and is validated by power spectrum analysis. Next, a permutation entropy-kurtosis joint index is used to reconstruct primary modal components, followed by secondary decomposition with an adaptive filter to further simplify the components. Then, the Maximum Information Coefficient (MIC) quantifies the nonlinear relationship between environmental factors and ratio errors, reducing redundant features. At the same time, the Autocorrelation Function (ACF) processes time series data to identify temporal dependencies and select effective features. Additionally, a Wasserstein Generative Adversarial Network (WGAN) generates fault samples, thereby enhancing the model’s ability to handle small sample conditions. Finally, 15 benchmark prediction models are constructed to validate the effectiveness of the proposed prediction framework. Experimental results demonstrate that the proposed framework effectively reduces the complexity of the original data, showing significant improvements across four performance metrics, and enhancing the accuracy and stability of VT ratio error predictions. This method provides a robust solution for VT performance monitoring and data-driven transformation in renewable energy-integrated grids.

The long-term monitoring stability of electronic current transformers is crucial for accurately obtaining the current signal of the power grid. However, it is difficult to accurately distinguish between the fluctuation of non-stationary random signals on the primary side of the power grid and the gradual error of the transformers themselves. A current transformer error prediction model, CNN-MHA-BiLSTM, based on the golden jackal optimization (GJO) algorithm, which is used to obtain the optimal parameter values, bidirectional long short-term memory (BiLSTM) network, convolutional neural networks (CNNs), and multi-head attention (MHA), is proposed to address the difficulty of measuring error evaluation. This model can be used to determine the operation of transformers and can be widely applied to assist in determining the stability of transformer operation and early faults. First, CNN is used to mine the vertical detail features of error data at a certain moment, improving the speed of error prediction. Furthermore, a cascaded network with BiLSTM as the core is constructed to extract the horizontal historical features of the error data. The GJO algorithm is used to adjust the parameters of the BiLSTM model; optimize the hidden layer nodes, training frequency, and learning rate; and integrate MHA mechanism to promote the model to pay attention to the characteristic changes of the data in order to improve the accuracy of error prediction. Finally, this method is applied to the operation data of transformer in substations, and four time periods of data are selected to verify the model effectiveness of the current transformer dataset. The analysis results of single step and multi-step examples indicate that the proposed model has significant advantages in terms of accuracy and stability in error prediction.

Optical fiber current transformers (FOCTs) are affected by various external factors, resulting in the deterioration or even failure of devices, causing changes in critical state quantities, and reducing the accuracy and reliability of products. To solve this problem, based on the neural network algorithm, this paper starts from the four deterioration characteristics of FOCT SLD junction temperature, SLD output optical power, phase modulator half-wave voltage, and optical fiber sensing ring temperature, and identifies the deterioration of key optical components of FOCT, which provides a basic model and data support for the online monitoring and early warning to improving the stability and reliability of FOCT in long-term operation.

The ultra-high voltage direct current (UHVDC) system is widely constructed due to its suitability for large-capacity long-distance transmission, and its losses have become an important part of the power grid losses. However, the overall and subcomponent losses of the power system are difficult to measure by measuring devices or accurately calculated by existing algorithms. In this paper, a convolutional neural network (CNN)-based digital twin model for UHVDC system loss measurement is proposed. The parameters and data of the physical system are fed into the digital space system through the data interaction. In the digital twin model of the digital space system, a CNN network with added topology information extraction layers and key equipment parameter adaptive perception layers excavates the deep multidimensional correlation features of the digital space mapping data. Then a multi-task joint processing layer fuses the extracted deep features and missing information to calculate the total and sub-component losses. The training and testing results based on actual UHVDC project data show that the mean absolute error (MAE) is 0.9, while R 2 is 0.9999, which proves the accuracy of the proposed digital twin model is superior to mainstream deep learning models. This model is embedded in digital twin devices and applied in actual UHVDC converter stations.

Based on the characteristics of optical couplers, a time domain demodulation algorithm for Sagnac Interferometer current transformers is proposed, which can reduce external interference and is insensitive to the polarization state of the input light.

An accelerator-based facility, such as an FEL injector, has stringent requirements on the quality of electron beam, and the electron beam is directly determined by beam injector, which is generally a source to provide driven beams with the energy of several MeVs. Since such space-charge dominated relativistic beams are sensitive and easy to be deteriorated during transportation, it is necessary to carry out online monitoring of beam quality under commissioning, so as to achieve accurate measurement of beam parameters such as beam spot size, beam emittance and beam energy spread, and beam current. The whole measuring facility is composed of magnetic components such as analysis magnet and quadrupole magnet. It is necessary to control and monitor the magnet current, and then display the data uniformly to the operating interface and feed back to the user through different control and measuring devices in the system. This paper introduces the basic principle of beam measuring device and the layout design of measuring system. Based on EPICS system and LabVIEW software, a distributed three-layer architecture control and measuring system is developed, which can realize feedback control, remote monitoring and signal acquisition. The whole system is stable and reliable in operation, convenient in use, high in beam parameter accuracy and accurate in calculation, which improves the efficiency of online monitoring.