Heng Lu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Heng Lu has a total of 6 co-authors. Here are the top co-authors by number of shared publications:
- Zhu Zhang
School of Electrical Engineering and Automation, Hefei University of Technology, Hefei, China
5 shared publications - Binbin Li
China and State Grid Anhui Electric Power Company Electric Power Research Institute, Hefei, China
4 shared publications - Lijian Ding
Department of Respiratory and Critical Care Medicine, The Affiliated Yangming Hospital of Ningbo University (Yuyao People’s Hospital), Ningbo University
3 shared publications - Shaorui Qin
China and State Grid Anhui Electric Power Company Electric Power Research Institute, Hefei, China
2 shared publications - JingXuan Wang
School of Electrical Engineering and Automation, Hefei University of Technology, Hefei, China
1 shared publication - XueRu Yang
School of Electrical Engineering and Automation, Hefei University of Technology, Hefei, China
1 shared publication
Latest Publications
- 5 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 75 (00189456) , 9004312
Measurement Error Situation Awareness Method for Electronic Voltage Transformers Based on Physics-Informed Machine Learning
Capacitive voltage divider-type electronic voltage transformers (EVTs) are widely used measurement devices in high-voltage power systems, where their measurement errors directly affect smart grids safety and economic operation. To address the issues of low robustness and insufficient accuracy in existing data-driven or single-physics-driven measurement error perception models, this paper proposes a state awareness method for capacitive voltage divider-type EVT measurement errors based on physics-informed machine learning (PIML). First, considering the generalization limitations of conventional temporal features, a physics-informed augmentation mechanism is designed to construct deep feature parameters that influence electronic voltage transformer measurement errors. Second, based on the physical processes governing EVT measurement error variations, physical loss constraints are formulated to establish a physics-informed machine learning model. Subsequently, an enhanced Shapley Additive Explanations (SHAP) interpretation mechanism is proposed to provide global explanations for these deep feature parameters, culminating in the development of an EVT measurement error situational awareness model. Actual engineering data demonstrate that the proposed situational awareness method for electronic voltage transformer measurement errors achieves a prediction accuracy of 0.99, representing an average improvement of 3.12% over existing models. The ablation experiment also shows that the model integrating physical information and physical constraints can improve accuracy by at least 2.77%. The explanation mechanism enables visualization of feature contribution and importance on a ‘single-instance-to-time-period’ basis. The explanation results indicate that the thermal-load coupling effect of EVTs warrants particular attention.
- 5 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 75 (00189456) , 9004312
Measurement Error Situation Awareness Method for Electronic Voltage Transformers Based on Physics-Informed Machine Learning
Capacitive voltage divider-type electronic voltage transformers (EVTs) are widely used measurement devices in high-voltage power systems, where their measurement errors directly affect smart grids safety and economic operation. To address the issues of low robustness and insufficient accuracy in existing data-driven or single-physics-driven measurement error perception models, this paper proposes a state awareness method for capacitive voltage divider-type EVT measurement errors based on physics-informed machine learning (PIML). First, considering the generalization limitations of conventional temporal features, a physics-informed augmentation mechanism is designed to construct deep feature parameters that influence electronic voltage transformer measurement errors. Second, based on the physical processes governing EVT measurement error variations, physical loss constraints are formulated to establish a physics-informed machine learning model. Subsequently, an enhanced Shapley Additive Explanations (SHAP) interpretation mechanism is proposed to provide global explanations for these deep feature parameters, culminating in the development of an EVT measurement error situational awareness model. Actual engineering data demonstrate that the proposed situational awareness method for electronic voltage transformer measurement errors achieves a prediction accuracy of 0.99, representing an average improvement of 3.12% over existing models. The ablation experiment also shows that the model integrating physical information and physical constraints can improve accuracy by at least 2.77%. The explanation mechanism enables visualization of feature contribution and importance on a ‘single-instance-to-time-period’ basis. The explanation results indicate that the thermal-load coupling effect of EVTs warrants particular attention.
- 4 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-11
Online Error Inversion-Based Assessment of the Internal Insulation Performance of a Capacitor Voltage Transformer
The internal insulation parameters of capacitor voltage transformers (CVTs) are susceptible to the influence of environment and aging during operation, which leads to the degradation of its insulation performance and affects the safe and stable operation of power system. Since the current insulation testing method is essentially a periodic testing of power outage, which has the problems of difficulty of power outage and untimely testing, this article proposes an online assessment method of CVT internal insulation performance based on error inversion. This method innovatively takes the measurement error of CVT as a characteristic parameter to characterize the insulation state of CVT, establishes a feature set of insulation state composed of specific and angular differences, and then uses a convolutional neural network (CNN) and fuzzy analysis to form a data-knowledge double-layer-driven online assessment model of CVT’s internal insulation performance. The experimental results show that the proposed method has an accuracy of 99.76% for the identification of typical internal insulation states, and the average error of quantitative evaluation of internal insulation parameters is better than 3%, which effectively realizes the online evaluation of CVT internal insulation parameters.
- 4 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-11
Online Error Inversion-Based Assessment of the Internal Insulation Performance of a Capacitor Voltage Transformer
The internal insulation parameters of capacitor voltage transformers (CVTs) are susceptible to the influence of environment and aging during operation, which leads to the degradation of its insulation performance and affects the safe and stable operation of power system. Since the current insulation testing method is essentially a periodic testing of power outage, which has the problems of difficulty of power outage and untimely testing, this article proposes an online assessment method of CVT internal insulation performance based on error inversion. This method innovatively takes the measurement error of CVT as a characteristic parameter to characterize the insulation state of CVT, establishes a feature set of insulation state composed of specific and angular differences, and then uses a convolutional neural network (CNN) and fuzzy analysis to form a data-knowledge double-layer-driven online assessment model of CVT’s internal insulation performance. The experimental results show that the proposed method has an accuracy of 99.76% for the identification of typical internal insulation states, and the average error of quantitative evaluation of internal insulation parameters is better than 3%, which effectively realizes the online evaluation of CVT internal insulation parameters.
- 4 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 73 (00189456) , 1-11
Online Detection Method for Capacitor Voltage Transformer Measurement Error Adapted to Nonstationary Random Processes
Capacitor Voltage Transformers (CVTs) are measurement devices widely used in high-voltage power grids, and the long-term stability of their measurement errors affects the safe operation of power systems. To address the problem of insufficient real-time periodic calibration using standard transformers, this paper proposes an online detection method for CVT measurement error based on enhanced integrated principal component analysis (EIPCA). Considering that the secondary output voltage of CVTs is characterized by a non-stationary random distribution, the method combines the advantages of principal component analysis (PCA) and Kernel Principal Component Analysis (KPCA) to enhance its ability to process mixed linear and nonlinear data. It performs variational mode decomposition (VMD) of the linear residuals mapped in the cascade process to eliminate the modeling error caused by non-stationary data. Finally, the residual component characterizing the integrated error information of the CVT is separated from the primary voltage fluctuation, and the T 2 statistic is established as the index of measurement error detection. The experimental results show that the EIPCA-based online error detection method proposed in this paper can effectively detect the 0.2-level error of the CVT, significantly improving detection performance compared with traditional PCA.
- 4 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 73 (00189456) , 1-11
Online Detection Method for Capacitor Voltage Transformer Measurement Error Adapted to Nonstationary Random Processes
Capacitor Voltage Transformers (CVTs) are measurement devices widely used in high-voltage power grids, and the long-term stability of their measurement errors affects the safe operation of power systems. To address the problem of insufficient real-time periodic calibration using standard transformers, this paper proposes an online detection method for CVT measurement error based on enhanced integrated principal component analysis (EIPCA). Considering that the secondary output voltage of CVTs is characterized by a non-stationary random distribution, the method combines the advantages of principal component analysis (PCA) and Kernel Principal Component Analysis (KPCA) to enhance its ability to process mixed linear and nonlinear data. It performs variational mode decomposition (VMD) of the linear residuals mapped in the cascade process to eliminate the modeling error caused by non-stationary data. Finally, the residual component characterizing the integrated error information of the CVT is separated from the primary voltage fluctuation, and the T 2 statistic is established as the index of measurement error detection. The experimental results show that the EIPCA-based online error detection method proposed in this paper can effectively detect the 0.2-level error of the CVT, significantly improving detection performance compared with traditional PCA.
- 24 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 73 (00189456) , 1-12
Research on Data-Driven Self-Diagnosis for Measurement Errors in Capacitor Voltage Transformers
Capacitor voltage transformers (CVTs) are widely used in high-voltage power systems. Their main problem is poor stability in measurement errors. The existing approach of utilizing standard voltage transformers for regular offline error calibration is insufficient in real-time and cannot meet the requirements of intelligent substations for online operations monitoring of key equipment. Therein, this paper proposes a data-driven self-diagnosis method based on metering errors of operational CVTs. This method synchronously collects the secondary signals of three-phase CVT outputs. Firstly, under the rigid constraints of three-phase symmetrical power system operations, the residual subspace parameters that characterize the measurement errors are extracted using principal component analysis (PCA). Then, bilateral threshold detection is combined with a multi-dimensional comprehensive assessment to determine the boundary conditions of the abnormal overall error state of the three-phase CVT and its deterioration direction. Finally, the coded multi-dimensional detection results are input into the learning vector quantized (LVQ) classifier to rapidly diagnose CVT measurement error deterioration. The experimental results show that the proposed method effectively identifies error fluctuations at 0.02% with a detection accuracy for the CVT amplitude error overshoot reaching 0.1%. The method can accurately locate the faulty phases and direction of error deterioration in real-time, realizing the demand for self-diagnosing for CVT metrological errors at the 0.2-level.
- 24 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 73 (00189456) , 1-12
Research on Data-Driven Self-Diagnosis for Measurement Errors in Capacitor Voltage Transformers
Capacitor voltage transformers (CVTs) are widely used in high-voltage power systems. Their main problem is poor stability in measurement errors. The existing approach of utilizing standard voltage transformers for regular offline error calibration is insufficient in real-time and cannot meet the requirements of intelligent substations for online operations monitoring of key equipment. Therein, this paper proposes a data-driven self-diagnosis method based on metering errors of operational CVTs. This method synchronously collects the secondary signals of three-phase CVT outputs. Firstly, under the rigid constraints of three-phase symmetrical power system operations, the residual subspace parameters that characterize the measurement errors are extracted using principal component analysis (PCA). Then, bilateral threshold detection is combined with a multi-dimensional comprehensive assessment to determine the boundary conditions of the abnormal overall error state of the three-phase CVT and its deterioration direction. Finally, the coded multi-dimensional detection results are input into the learning vector quantized (LVQ) classifier to rapidly diagnose CVT measurement error deterioration. The experimental results show that the proposed method effectively identifies error fluctuations at 0.02% with a detection accuracy for the CVT amplitude error overshoot reaching 0.1%. The method can accurately locate the faulty phases and direction of error deterioration in real-time, realizing the demand for self-diagnosing for CVT metrological errors at the 0.2-level.
- 1 January 2024
- Published by AIP Publishing in Review of Scientific Instruments
- Vol. 95 (1)
Self-detection method for measurement error of capacitor voltage transformer considering conversion error
The measurement error of capacitor voltage transformers (CVTs) has poor stability under the complex environment of substations. Conventionally, error detection is performed by regularly comparing the output of standard transformers, which lacks real-time performance. Moreover, CVTs are prone to operating in an out-of-tolerance state. Thus, this study first analyzes the basic principle of the CVT measurement error self-detection method based on principal component analysis under the constraints of three-phase symmetrical operating characteristics of a power system. Then, the impact of the signal conversion error of analog-to-digital conversion equipment on the self-detection results of CVT measurement errors in engineering applications is analyzed, and a multi-layer wavelet analysis signal denoising method is proposed to enhance the self-detection ability of CVT measurement errors. Finally, simulations show that the proposed method can identify an error change of 0.1%, meeting the self-detection requirements for the measurement error at 0.2 class.
- 1 January 2024
- Published by AIP Publishing in Review of Scientific Instruments
- Vol. 95 (1)
Self-detection method for measurement error of capacitor voltage transformer considering conversion error
The measurement error of capacitor voltage transformers (CVTs) has poor stability under the complex environment of substations. Conventionally, error detection is performed by regularly comparing the output of standard transformers, which lacks real-time performance. Moreover, CVTs are prone to operating in an out-of-tolerance state. Thus, this study first analyzes the basic principle of the CVT measurement error self-detection method based on principal component analysis under the constraints of three-phase symmetrical operating characteristics of a power system. Then, the impact of the signal conversion error of analog-to-digital conversion equipment on the self-detection results of CVT measurement errors in engineering applications is analyzed, and a multi-layer wavelet analysis signal denoising method is proposed to enhance the self-detection ability of CVT measurement errors. Finally, simulations show that the proposed method can identify an error change of 0.1%, meeting the self-detection requirements for the measurement error at 0.2 class.