Jingyi Yuan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Jingyi Yuan has a total of 64 co-authors. Here are the top co-authors by number of shared publications:
- Yang Weng
School of Electrical, Computer and Energy Engineering, Arizona State University, Tempe, AZ, USA
9 shared publications - Erik Blasch
Air Force Research Laboratory
4 shared publications - Jiaqi Wu
Arizona State University
2 shared publications - Adrian S. Wong
Air Force Research Laboratory, Materials and Manufacturing Directorate, Composites, Ceramics, Metallics, and Materials Awareness Division, Materials State Awareness Branch, 2230 10th Street, B655, Ste 1, WPAFB, Dayton, Ohio 45433, United States of America
1 shared publication - Alice Cicirello
Department of Engineering, University of Cambridge
1 shared publication - Amrita Kumar
Acellent Technologies Inc., Sunnyvale, CA, United States of America
1 shared publication
Latest Publications
- 1 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Smart Grid
- Vol. 17 (3) , 2064-2076
Cyclic Regularization of Input Convex Neural Networks for Robust Voltage Regulation
Traditional centralized voltage regulation in distribution grids is typically formulated using the DistFlow model, which requires complete system knowledge, including topology and line parameters. However, due to limited observability, complete system information is often unavailable or unreliable at the grid edge. Previous works introduce data-driven approaches to overcome this problem, and input convex neural networks (ICNNs) have been applied to preserve convexity, enabling tractable optimization. However, discriminative learning models like ICNNs often overfit historical data, leading to poor generalization to unseen operating conditions. Overfitting may lead to biased control, particularly under shifting operating conditions caused by high penetrations of distributed energy resources (DERs). Challenges to regularizing such data-driven surrogates lie in the absence of a generative mapping from voltage states to control variables like power injections. To address these limitations, we propose a Regularized ICNN (RegICNN) incorporating a cyclic neural architecture for physics-based regularization. In RegICNN, we 1) enable a form of generative regularization to capture bidirectional mappings between voltage and power for unseen operation points; 2) further introduce a quadratic regularization term tailored to power flow physics; and 3) theoretically analyze the injection-voltage mapping under the DistFlow model, showing its convexity conditions and justifying the use of ICNNs as optimum data-driven surrogates for voltage control. Specifically, RegICNN learns a convex neural surrogate that approximates the injection-voltage mapping while preserving feasibility and global optimality for gradient-based control. Experiments on 3-, 12-, 116-, and 123-bus distribution feeders demonstrate that the RegICNN-based surrogate significantly improves modeling accuracy and control robustness when facing changes in operating points, making it a promising approach for data-driven voltage control in modern, dynamic distribution grids.
- 1 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Smart Grid
- Vol. 17 (3) , 2064-2076
Cyclic Regularization of Input Convex Neural Networks for Robust Voltage Regulation
Traditional centralized voltage regulation in distribution grids is typically formulated using the DistFlow model, which requires complete system knowledge, including topology and line parameters. However, due to limited observability, complete system information is often unavailable or unreliable at the grid edge. Previous works introduce data-driven approaches to overcome this problem, and input convex neural networks (ICNNs) have been applied to preserve convexity, enabling tractable optimization. However, discriminative learning models like ICNNs often overfit historical data, leading to poor generalization to unseen operating conditions. Overfitting may lead to biased control, particularly under shifting operating conditions caused by high penetrations of distributed energy resources (DERs). Challenges to regularizing such data-driven surrogates lie in the absence of a generative mapping from voltage states to control variables like power injections. To address these limitations, we propose a Regularized ICNN (RegICNN) incorporating a cyclic neural architecture for physics-based regularization. In RegICNN, we 1) enable a form of generative regularization to capture bidirectional mappings between voltage and power for unseen operation points; 2) further introduce a quadratic regularization term tailored to power flow physics; and 3) theoretically analyze the injection-voltage mapping under the DistFlow model, showing its convexity conditions and justifying the use of ICNNs as optimum data-driven surrogates for voltage control. Specifically, RegICNN learns a convex neural surrogate that approximates the injection-voltage mapping while preserving feasibility and global optimality for gradient-based control. Experiments on 3-, 12-, 116-, and 123-bus distribution feeders demonstrate that the RegICNN-based surrogate significantly improves modeling accuracy and control robustness when facing changes in operating points, making it a promising approach for data-driven voltage control in modern, dynamic distribution grids.
- Top Cited PapersOpen Access
- 12 March 2026
- Published by IOP Publishing in Measurement Science and Technology
- Vol. 37 (10) , 103001
Roadmap: Integrating artificial intelligence in structural health monitoring systems
Advances in computing and machine learning (ML) methods have led to a rapid rise in artificial intelligence (AI) research and applications in many fields. AI research benefitted from advances in computation hardware, collection and distribution of large data sets, and proliferation of software techniques. AI techniques include ML for provable results, deep learning for data exploration, reinforcement learning for control, and active learning for adaptive systems. Likewise, AI algorithms can handle large amounts of data, construct unknown representations, and provide a direct link between data and classification for decision making. These unmatched capabilities have been seen as a path to solving hard engineering problems, including that of structural health monitoring (SHM). SHM consists of automating the condition assessment task of civil, health, mechanical, and aerospace systems using measurements obtained from temporary or permanently installed sensors. Often, the systems of interest are geometrically large and/or technically complex, which complicates the development and application of physics-based methods. It follows that AI is seen as a key potential contributor enabling SHM in field applications for data-driven analysis. As with many research endeavors, many concepts using AI for SHM have been explored in the literature. Nevertheless, very few AI methods have been deployed in the context of SHM, which may be due to the lack of available data supporting their capabilities, limited integrated AI-SHM systems capable of providing results to users and operators with decision-making capabilities, or certification of AI methods for safety-critical applications. The objective of this Roadmap publication is to discuss the integration of AI at the system level enabling SHM, including associated challenges and opportunities such as those found in common metrics of concern (e.g. transparency, interpretability, explainability, security, certifiability, etc), with a particular focus on providing a path to research and development efforts that could yield impactful field applications. The overview of available methods and directions will provide the readers with applicability of AI for certain SHM designs (software), availability of common data sets for further AI comparisons (data), and lessons learned in implementation (hardware).
- 12 March 2026
- Published by IOP Publishing in Measurement Science and Technology
- Vol. 37 (10) , 103001
Top Cited PapersOpen AccessRoadmap: Integrating artificial intelligence in structural health monitoring systems
Advances in computing and machine learning (ML) methods have led to a rapid rise in artificial intelligence (AI) research and applications in many fields. AI research benefitted from advances in computation hardware, collection and distribution of large data sets, and proliferation of software techniques. AI techniques include ML for provable results, deep learning for data exploration, reinforcement learning for control, and active learning for adaptive systems. Likewise, AI algorithms can handle large amounts of data, construct unknown representations, and provide a direct link between data and classification for decision making. These unmatched capabilities have been seen as a path to solving hard engineering problems, including that of structural health monitoring (SHM). SHM consists of automating the condition assessment task of civil, health, mechanical, and aerospace systems using measurements obtained from temporary or permanently installed sensors. Often, the systems of interest are geometrically large and/or technically complex, which complicates the development and application of physics-based methods. It follows that AI is seen as a key potential contributor enabling SHM in field applications for data-driven analysis. As with many research endeavors, many concepts using AI for SHM have been explored in the literature. Nevertheless, very few AI methods have been deployed in the context of SHM, which may be due to the lack of available data supporting their capabilities, limited integrated AI-SHM systems capable of providing results to users and operators with decision-making capabilities, or certification of AI methods for safety-critical applications. The objective of this Roadmap publication is to discuss the integration of AI at the system level enabling SHM, including associated challenges and opportunities such as those found in common metrics of concern (e.g. transparency, interpretability, explainability, security, certifiability, etc), with a particular focus on providing a path to research and development efforts that could yield impactful field applications. The overview of available methods and directions will provide the readers with applicability of AI for certain SHM designs (software), availability of common data sets for further AI comparisons (data), and lessons learned in implementation (hardware).
- Open Access
- 24 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 4071-4082
DipDNN: Preserving Inverse Consistency and Approximation Efficiency for Invertible Learning
Consistent bi-directional inferences are the key for many machine learning applications. Without consistency, inverse learning-based inferences can cause fuzzy images, erroneous control signals, and cascading failure in SCADA systems. Since standard deep neural networks (DNNs) are not inherently invertible to offer consistency, some past methods reconstruct DNN architecture analytically for one-to-one correspondence but compromise key features such as universal approximation. Other work maintains the capability of universal approximation in DNNs via iterative numerical approximation. However, these methods limit their applications significantly due to Lipschitz conditions and issues of numerical convergence. The dilemma of the analytical and numerical methods is the incompatibility between nonlinear layer compositions and bijective function construction for inverse modeling. Based on the observation, we propose decomposed-invertible-pathway DNNs (DipDNN). It relaxes the redundant reconstruction of nested DNN in the former methods and eases the Lipschitz constraint. As a result, we strictly guarantee the consistency of global inverse modeling without harming DNN's capability for universal approximation. As numerical stability and generalizability are keys for controlling critical infrastructures, we integrate contractive property with a parallel structure for inductive biases, leading to stable performance. Numerical results show that DipDNN performs significantly better than past methods, thanks to its enforcement of inverse consistency, numerical stability, and physical regularization.
- 24 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 4071-4082
Open AccessDipDNN: Preserving Inverse Consistency and Approximation Efficiency for Invertible Learning
Consistent bi-directional inferences are the key for many machine learning applications. Without consistency, inverse learning-based inferences can cause fuzzy images, erroneous control signals, and cascading failure in SCADA systems. Since standard deep neural networks (DNNs) are not inherently invertible to offer consistency, some past methods reconstruct DNN architecture analytically for one-to-one correspondence but compromise key features such as universal approximation. Other work maintains the capability of universal approximation in DNNs via iterative numerical approximation. However, these methods limit their applications significantly due to Lipschitz conditions and issues of numerical convergence. The dilemma of the analytical and numerical methods is the incompatibility between nonlinear layer compositions and bijective function construction for inverse modeling. Based on the observation, we propose decomposed-invertible-pathway DNNs (DipDNN). It relaxes the redundant reconstruction of nested DNN in the former methods and eases the Lipschitz constraint. As a result, we strictly guarantee the consistency of global inverse modeling without harming DNN's capability for universal approximation. As numerical stability and generalizability are keys for controlling critical infrastructures, we integrate contractive property with a parallel structure for inductive biases, leading to stable performance. Numerical results show that DipDNN performs significantly better than past methods, thanks to its enforcement of inverse consistency, numerical stability, and physical regularization.
- Open Access
- 29 March 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Systems
- Vol. 39 (1) , 2251-2263
Enhance Unobservable Solar Generation Estimation via Constructive Generative Adversarial Networks
Power distribution grids experiences proliferation of solar photovoltaics (PV) at the system edge. However, its counterpart of sparse meter deployment provides insufficient monitoring of PVs, for which the potential violations challenge the operators for energy management and stable operation. Some previous works use satellite imagery to detect distributed PVs for the easy access of data. However, their PV localization methods rely on label-rich area with unitary background/environment to implement well; even further/harder, they do not provide precise metered-PV detection and quantification to estimate/know PV generation outputs in unobservable area, which is essential to prevent the edge from excessive two-way power flow and other violations. Thus, we combine the two steps of detecting PV existence and quantify PV amount into one classification task. To boost the classification performance in unobservable edge area, we construct a generative adversarial network that simultaneous augments the diversity of labelled PV satellite images and embed distinct PV characteristics/features for training the classifier. Furthermore, the PV localization and quantification result is combined with geographic information, historical weather conditions and neighboring generation patterns to estimate power output at the system edge. We validate the proposed approaches on PV systems in the southwest of the U.S. Experiment results show high accuracy and robustness in predicting distributed solar power without sufficient prior information.
- 29 March 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Systems
- Vol. 39 (1) , 2251-2263
Open AccessEnhance Unobservable Solar Generation Estimation via Constructive Generative Adversarial Networks
Power distribution grids experiences proliferation of solar photovoltaics (PV) at the system edge. However, its counterpart of sparse meter deployment provides insufficient monitoring of PVs, for which the potential violations challenge the operators for energy management and stable operation. Some previous works use satellite imagery to detect distributed PVs for the easy access of data. However, their PV localization methods rely on label-rich area with unitary background/environment to implement well; even further/harder, they do not provide precise metered-PV detection and quantification to estimate/know PV generation outputs in unobservable area, which is essential to prevent the edge from excessive two-way power flow and other violations. Thus, we combine the two steps of detecting PV existence and quantify PV amount into one classification task. To boost the classification performance in unobservable edge area, we construct a generative adversarial network that simultaneous augments the diversity of labelled PV satellite images and embed distinct PV characteristics/features for training the classifier. Furthermore, the PV localization and quantification result is combined with geographic information, historical weather conditions and neighboring generation patterns to estimate power output at the system edge. We validate the proposed approaches on PV systems in the southwest of the U.S. Experiment results show high accuracy and robustness in predicting distributed solar power without sufficient prior information.
- 18 March 2023
- Published by Springer Nature
- p. 253-269
Physically Invertible System Identification for Monitoring System Edges with Unobservability
- 18 March 2023
- Published by Springer Nature
- p. 253-269
Physically Invertible System Identification for Monitoring System Edges with Unobservability
- Open Access
- 8 August 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Smart Grid
- Vol. 14 (1) , 354-364
Spatial-Temporal Deep Learning for Hosting Capacity Analysis in Distribution Grids
The widespread use of distributed energy sources (DERs) raises significant challenges for power system design, planning, and operation, leading to wide adaptation of tools on hosting capacity analysis (HCA). Traditional HCA methods conduct extensive power flow analysis. Due to the computation burden, these time-consuming methods fail to provide online hosting capacity (HC) in large distribution systems. To solve the problem, we first propose a deep learning-based problem formulation for HCA, which conducts offline training and determines HC in real time. The used learning model, long short-term memory (LSTM), implements historical time-series data to capture periodical patterns in distribution systems. However, directly applying LSTMs suffers from low accuracy due to the lack of consideration on spatial information, where location information like feeder topology is critical in nodal HCA. Therefore, we modify the forget gate function to dual forget gates, to capture the spatial correlation within the grid. Such a design turns the LSTM into the Spatial-Temporal LSTM (ST-LSTM). Moreover, as voltage violations are the most vital constraints in HCA, we design a voltage sensitivity gate to increase accuracy further. The results of LSTMs and ST-LSTMs on feeders, such as IEEE 34-, 123-bus feeders, and utility feeders, validate our designs.
- 8 August 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Smart Grid
- Vol. 14 (1) , 354-364
Open AccessSpatial-Temporal Deep Learning for Hosting Capacity Analysis in Distribution Grids
The widespread use of distributed energy sources (DERs) raises significant challenges for power system design, planning, and operation, leading to wide adaptation of tools on hosting capacity analysis (HCA). Traditional HCA methods conduct extensive power flow analysis. Due to the computation burden, these time-consuming methods fail to provide online hosting capacity (HC) in large distribution systems. To solve the problem, we first propose a deep learning-based problem formulation for HCA, which conducts offline training and determines HC in real time. The used learning model, long short-term memory (LSTM), implements historical time-series data to capture periodical patterns in distribution systems. However, directly applying LSTMs suffers from low accuracy due to the lack of consideration on spatial information, where location information like feeder topology is critical in nodal HCA. Therefore, we modify the forget gate function to dual forget gates, to capture the spatial correlation within the grid. Such a design turns the LSTM into the Spatial-Temporal LSTM (ST-LSTM). Moreover, as voltage violations are the most vital constraints in HCA, we design a voltage sensitivity gate to increase accuracy further. The results of LSTMs and ST-LSTMs on feeders, such as IEEE 34-, 123-bus feeders, and utility feeders, validate our designs.
- Open Access
- 30 December 2021
- Published by IMR Press in Frontiers in Bioscience-Landmark
- Vol. 26 (12) , 1746-1759
Speech depression recognition based on attentional residual network
Background: Depressive disorder is a common affective disorder, also known as depression, which is characterized by sadness, loss of interest, feelings of guilt or low self-worth and poor concentration. As speech is easy to obtain non-offensively with low-cost, many researchers explore the possibility of depression prediction through speech. Adopting speech signals to recognize depression has important practical significance. Aiming at the problem of the complex structure of the deep neural network method used in the recognition of speech depression and the traditional machine learning methods need to manually extract the features and the low recognition rate. Methods: This paper proposes a model that combines residual thinking and attention mechanism. First, depression corpus is designed based on the classic psychological experimental paradigm self-reference effect (SRE), and the speech dataset is labeled; then the attention module is introduced into the residual, and the channel attention is used to learn the features of the channel dimension, the spatial attention feedback the features of the spatial dimension, and the combination of the two to obtain the attention residual unit; finally the stacking unit constructs a speech depression recognition model based on the attention residual network. Results: Experimental results show that compared with traditional machine learning methods, this model obtains better results in the recognition of depression, which can meet the need for actual recognition application of depression. Conclusions: In this study, we not only predict whether person is depressed, but also estimate the severity of depression. In the designed corpus, the depression binary classification of an individual is given based on the severity of depression which is measured using BDI-II scores. Experimental results show that spontaneous speech can obtain better results than automatic speech, and the classification of speech features corresponding to negative questions is better than other tasks under negative emotions. Besides, the recognition accuracy rate of both male and female subjects is higher than that under other emotions.
- 30 December 2021
- Published by IMR Press in Frontiers in Bioscience-Landmark
- Vol. 26 (12) , 1746-1759
Open AccessSpeech depression recognition based on attentional residual network
Background: Depressive disorder is a common affective disorder, also known as depression, which is characterized by sadness, loss of interest, feelings of guilt or low self-worth and poor concentration. As speech is easy to obtain non-offensively with low-cost, many researchers explore the possibility of depression prediction through speech. Adopting speech signals to recognize depression has important practical significance. Aiming at the problem of the complex structure of the deep neural network method used in the recognition of speech depression and the traditional machine learning methods need to manually extract the features and the low recognition rate. Methods: This paper proposes a model that combines residual thinking and attention mechanism. First, depression corpus is designed based on the classic psychological experimental paradigm self-reference effect (SRE), and the speech dataset is labeled; then the attention module is introduced into the residual, and the channel attention is used to learn the features of the channel dimension, the spatial attention feedback the features of the spatial dimension, and the combination of the two to obtain the attention residual unit; finally the stacking unit constructs a speech depression recognition model based on the attention residual network. Results: Experimental results show that compared with traditional machine learning methods, this model obtains better results in the recognition of depression, which can meet the need for actual recognition application of depression. Conclusions: In this study, we not only predict whether person is depressed, but also estimate the severity of depression. In the designed corpus, the depression binary classification of an individual is given based on the severity of depression which is measured using BDI-II scores. Experimental results show that spontaneous speech can obtain better results than automatic speech, and the classification of speech features corresponding to negative questions is better than other tasks under negative emotions. Besides, the recognition accuracy rate of both male and female subjects is higher than that under other emotions.
- Open Access
- 27 August 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Systems
- Vol. 37 (2) , 1151-1161
Support Matrix Regression for Learning Power Flow in Distribution Grid With Unobservability
Increasing renewable penetration in distribution grids calls for improved monitoring and control, where power flow (PF) model is the basis for many advanced functionalities. However, unobservability makes the traditional way infeasible to construct PF analysis via admittance matrix for many distribution grids. While data-driven approaches can approximate PF mapping, direct machine learning (ML) applications may suffer from several drawbacks. First, complex ML models like deep neural networks lack the degradability and explainability to the true system model, leading to overfitting. There are also asynchronization issues among different meters without GPS chips. Last but not least, bad data is quite common in the distribution grids. To resolve these problems all at once, we propose a variational support matrix regression (SMR). It provides structural learning to (1) embed kernels to regularize physical form in observable area while achieving good approximation at unobservable area, (2) integrate temporal information into matrix regression for asynchronized data imputation, and (3) define support matrix for margins to be robust against bad data. We test the performance for mapping rule learning via IEEE test systems and a utility distribution grid. Simulation results show high accuracy, degradability from data-driven model to physical model, and robustness to data quality issues.
- 27 August 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Systems
- Vol. 37 (2) , 1151-1161
Open AccessSupport Matrix Regression for Learning Power Flow in Distribution Grid With Unobservability
Increasing renewable penetration in distribution grids calls for improved monitoring and control, where power flow (PF) model is the basis for many advanced functionalities. However, unobservability makes the traditional way infeasible to construct PF analysis via admittance matrix for many distribution grids. While data-driven approaches can approximate PF mapping, direct machine learning (ML) applications may suffer from several drawbacks. First, complex ML models like deep neural networks lack the degradability and explainability to the true system model, leading to overfitting. There are also asynchronization issues among different meters without GPS chips. Last but not least, bad data is quite common in the distribution grids. To resolve these problems all at once, we propose a variational support matrix regression (SMR). It provides structural learning to (1) embed kernels to regularize physical form in observable area while achieving good approximation at unobservable area, (2) integrate temporal information into matrix regression for asynchronized data imputation, and (3) define support matrix for margins to be robust against bad data. We test the performance for mapping rule learning via IEEE test systems and a utility distribution grid. Simulation results show high accuracy, degradability from data-driven model to physical model, and robustness to data quality issues.