Gang Huang
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
Gang Huang has a total of 127 co-authors. Here are the top co-authors by number of shared publications:
Chuangxin Guo
8 shared publications- Wei Hua
Zhejiang University, Hangzhou, China
6 shared publications - Yongfu Li
Chongqing University of Posts and Telecommunications
6 shared publications - Zhou Zhou
Department of Engineering, Faculty of Environment, Science, and Economy (ESE), University of Exeter, Exeter EX4 4QF, UK
6 shared publications - Licheng Wang
University of Queensland, Brisbane, QLD, Australia
5 shared publications - Fei Wu
Zhejiang University
4 shared publications
Latest Publications
- Open Access
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1120-1128
SplitGaussian: Reconstructing Dynamic Scenes via Visual Geometry Decomposition
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1120-1128
Open AccessSplitGaussian: Reconstructing Dynamic Scenes via Visual Geometry Decomposition
- Open Access
- 2 February 2026
- Published by Springer Nature in Data Science and Engineering
- Vol. 11 (1) , 260-274
Dynamic Sub-Sequence Warping: A Representation-Based Similarity Measure for Long Time Series
Dynamic time warping (DTW), a typical elastic similarity measure that compares one-to-many points, has been proven effective for various time-series data mining tasks. However, it requires a quadratic time complexity $$O(n^2)$$ proportional to the length of time-series data, which undermines its applications involving long time series. In this paper, a representation-based similarity measure called Dynamic Sub-Sequence Warping (DSSW) is proposed. Instead of working on the raw data directly, we perform data representation to extract the distributional features of time series. Then, the similarity between two time series is measured by aligning the corresponding sub-sequences composed of the extracted features. We evaluate the proposed method through a supervised learning task on extensive real-world datasets. The results show that DSSW outperforms the prevalent DTW-based methods in terms of precision, and achieves one order of magnitude faster execution time on average compared with DTW.- 2 February 2026
- Published by Springer Nature in Data Science and Engineering
- Vol. 11 (1) , 260-274
Open AccessDynamic Sub-Sequence Warping: A Representation-Based Similarity Measure for Long Time Series
Dynamic time warping (DTW), a typical elastic similarity measure that compares one-to-many points, has been proven effective for various time-series data mining tasks. However, it requires a quadratic time complexity $$O(n^2)$$ proportional to the length of time-series data, which undermines its applications involving long time series. In this paper, a representation-based similarity measure called Dynamic Sub-Sequence Warping (DSSW) is proposed. Instead of working on the raw data directly, we perform data representation to extract the distributional features of time series. Then, the similarity between two time series is measured by aligning the corresponding sub-sequences composed of the extracted features. We evaluate the proposed method through a supervised learning task on extensive real-world datasets. The results show that DSSW outperforms the prevalent DTW-based methods in terms of precision, and achieves one order of magnitude faster execution time on average compared with DTW. - 19 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Artificial Intelligence
- Vol. 7 (7) , 3732-3747
Federated Semi-Supervised Learning with Annotation Heterogeneity
Federated Semi-Supervised Learning (FSSL) aims to learn a global model from different clients in an environment with both labeled and unlabeled data. Most of the existing FSSL work generally assumes that both types of data are available on each client. In this paper, we study a more general problem setup of FSSL with annotation heterogeneity, where each client can hold an arbitrary percentage (0%-100%) of labeled data. To this end, we propose a novel FSSL framework called Heterogeneously Annotated Semi-Supervised LEarning (HASSLE). Specifically, it employs a dual-model approach. Two models with the same architecture are trained separately: one uses labeled data only, and the other uses unlabeled data. This design enables the framework to be applied to clients with arbitrary labeling percentages. Furthermore, a mutual learning strategy called Supervised-Unsupervised Mutual Alignment (SUMA) is proposed for the dual models with global residual alignment and model proximity alignment. Subsequently, the dual models can implicitly learn from both types of data across different clients, although each dual model is only trained locally on a single type of data. Experiments verify that the dual models in HASSLE learned by SUMA can mutually learn from each other, thereby effectively utilizing the information of both types of data across different clients. Our code is available at https://github.com/shangxinyi/HASSLE.
- 19 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Artificial Intelligence
- Vol. 7 (7) , 3732-3747
Federated Semi-Supervised Learning with Annotation Heterogeneity
Federated Semi-Supervised Learning (FSSL) aims to learn a global model from different clients in an environment with both labeled and unlabeled data. Most of the existing FSSL work generally assumes that both types of data are available on each client. In this paper, we study a more general problem setup of FSSL with annotation heterogeneity, where each client can hold an arbitrary percentage (0%-100%) of labeled data. To this end, we propose a novel FSSL framework called Heterogeneously Annotated Semi-Supervised LEarning (HASSLE). Specifically, it employs a dual-model approach. Two models with the same architecture are trained separately: one uses labeled data only, and the other uses unlabeled data. This design enables the framework to be applied to clients with arbitrary labeling percentages. Furthermore, a mutual learning strategy called Supervised-Unsupervised Mutual Alignment (SUMA) is proposed for the dual models with global residual alignment and model proximity alignment. Subsequently, the dual models can implicitly learn from both types of data across different clients, although each dual model is only trained locally on a single type of data. Experiments verify that the dual models in HASSLE learned by SUMA can mutually learn from each other, thereby effectively utilizing the information of both types of data across different clients. Our code is available at https://github.com/shangxinyi/HASSLE.
- 8 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industry Applications
- Vol. 62 (2) , 3588-3599
Region-Based Resilience Evaluation and Quantification Method for Networked Microgrids Under Extreme Weather
The resilience level of networked microgrids (NMGs) has become a critical topic due to the increasing frequency of extreme weather events and the growing integration of renewable energy sources. Existing resilience evaluation and quantification methods primarily employ performance curve-based metrics or graph theory-based metrics. However, metrics such as expected load shedding or other outcome-based indices often fail to capture the security margin of current operating points, offering limited guidance for operators during extreme events. To address this gap, this paper proposes a region-based resilience evaluation and quantification method for NMGs that enables more intuitive resilience assessment and supports proactive dispatch. Specifically, the concept of the networked microgrids resilience region (NMRR) is introduced. The NMRR is constructed by projecting the intersection of feasible regions under all foreseeable failure-induced topology reconfiguration scenarios onto dispatchable variables accessible to system operators. A theoretical derivation of the NMRR’s analytical expression is provided. To efficiently compute the NMRR within dispatch timescales, an outer progressive approximation algorithm is developed. Two NMRR-based metrics are then proposed to quantify the resilience level of NMGs and the security margin of operating points, offering actionable dispatch insights. Case studies show that the proposed method aligns with traditional resilience indices while offering greater intuitiveness and operational relevance.
- 8 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industry Applications
- Vol. 62 (2) , 3588-3599
Region-Based Resilience Evaluation and Quantification Method for Networked Microgrids Under Extreme Weather
The resilience level of networked microgrids (NMGs) has become a critical topic due to the increasing frequency of extreme weather events and the growing integration of renewable energy sources. Existing resilience evaluation and quantification methods primarily employ performance curve-based metrics or graph theory-based metrics. However, metrics such as expected load shedding or other outcome-based indices often fail to capture the security margin of current operating points, offering limited guidance for operators during extreme events. To address this gap, this paper proposes a region-based resilience evaluation and quantification method for NMGs that enables more intuitive resilience assessment and supports proactive dispatch. Specifically, the concept of the networked microgrids resilience region (NMRR) is introduced. The NMRR is constructed by projecting the intersection of feasible regions under all foreseeable failure-induced topology reconfiguration scenarios onto dispatchable variables accessible to system operators. A theoretical derivation of the NMRR’s analytical expression is provided. To efficiently compute the NMRR within dispatch timescales, an outer progressive approximation algorithm is developed. Two NMRR-based metrics are then proposed to quantify the resilience level of NMGs and the security margin of operating points, offering actionable dispatch insights. Case studies show that the proposed method aligns with traditional resilience indices while offering greater intuitiveness and operational relevance.
- 18 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 26 (6) , 7465-7475
A Hierarchical Controller for Connected Truck Platoon: Analysis and Verification
This paper proposes a novel hierarchical controller for connected truck platoons. To this end, the predecessor following topology is used to characterize the communication connectivity between connected trucks. Then, a longitudinal efficient controller consisting of upper-level and lower-level controllers is proposed. In particular, the upper-level controller is designed based on the kinematic model to handle the car-following interactions between connected trucks and delays in communication and input. The lower-level controller comprises a feedforward and a feedback control law. The feedforward control law converts the desired acceleration from the upper-level controller into the vehicle throttle or braking pressure using the inverse dynamic model, while the feedback control law compensates for the control error caused by unknown vehicle parameters. In addition, in the linear region, the internal stability is analyzed based on the second-order kinematic model using s-domain analysis and linearization method, respectively. Then, the string stability is proved. The influence of parameters on the stability performance is extensively discussed using the stability diagram. Finally, the feasibility of the proposed controller is verified via co-simulations in PreScan and TruckSim, in terms of acceleration, velocity, and spacing error profiles.
- 18 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 26 (6) , 7465-7475
A Hierarchical Controller for Connected Truck Platoon: Analysis and Verification
This paper proposes a novel hierarchical controller for connected truck platoons. To this end, the predecessor following topology is used to characterize the communication connectivity between connected trucks. Then, a longitudinal efficient controller consisting of upper-level and lower-level controllers is proposed. In particular, the upper-level controller is designed based on the kinematic model to handle the car-following interactions between connected trucks and delays in communication and input. The lower-level controller comprises a feedforward and a feedback control law. The feedforward control law converts the desired acceleration from the upper-level controller into the vehicle throttle or braking pressure using the inverse dynamic model, while the feedback control law compensates for the control error caused by unknown vehicle parameters. In addition, in the linear region, the internal stability is analyzed based on the second-order kinematic model using s-domain analysis and linearization method, respectively. Then, the string stability is proved. The influence of parameters on the stability performance is extensively discussed using the stability diagram. Finally, the feasibility of the proposed controller is verified via co-simulations in PreScan and TruckSim, in terms of acceleration, velocity, and spacing error profiles.
- 26 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-12
Federated Unilateral-Mutual Learning for Privacy-Preserving Nonintrusive Load Monitoring
Accurate monitoring of appliance energy usage is essential for efficient energy management and savings. The popularity of nonintrusive load monitoring (NILM) has increased, thanks to deep learning algorithms that reduce hardware dependency through computational efficiency. However, collecting sufficient data from various households or buildings for deep learning compromises privacy, posing significant real-world risks. Federated learning has recently been applied to bridge this gap by keeping the energy usage data localized. Nevertheless, the assumption of data homogeneity in federated learning can be problematic in real-world scenarios. This paper proposes a novel Federated Unilateral-Mutual Learning (FUML) framework, addressing the critical measurement science challenge of data heterogeneity while ensuring privacy for NILM. First, a unilateral learning module is employed to capture the inner patterns within participants. Then, a mutual learning module is proposed to facilitate the exchange of pattern differences among participants. Finally, a federated aggregation module is utilized to coordinate the trained models. Through the above modules, the FUML framework extends standard federated learning into a three-stage process to accommodate heterogeneous data. The training algorithm for the proposed method is presented, and its effectiveness is validated through extensive case studies across three real-world datasets. The results show that the proposed method outperforms existing federated learning methods by a substantial margin for privacy-preserving NILM.
- 26 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-12
Federated Unilateral-Mutual Learning for Privacy-Preserving Nonintrusive Load Monitoring
Accurate monitoring of appliance energy usage is essential for efficient energy management and savings. The popularity of nonintrusive load monitoring (NILM) has increased, thanks to deep learning algorithms that reduce hardware dependency through computational efficiency. However, collecting sufficient data from various households or buildings for deep learning compromises privacy, posing significant real-world risks. Federated learning has recently been applied to bridge this gap by keeping the energy usage data localized. Nevertheless, the assumption of data homogeneity in federated learning can be problematic in real-world scenarios. This paper proposes a novel Federated Unilateral-Mutual Learning (FUML) framework, addressing the critical measurement science challenge of data heterogeneity while ensuring privacy for NILM. First, a unilateral learning module is employed to capture the inner patterns within participants. Then, a mutual learning module is proposed to facilitate the exchange of pattern differences among participants. Finally, a federated aggregation module is utilized to coordinate the trained models. Through the above modules, the FUML framework extends standard federated learning into a three-stage process to accommodate heterogeneous data. The training algorithm for the proposed method is presented, and its effectiveness is validated through extensive case studies across three real-world datasets. The results show that the proposed method outperforms existing federated learning methods by a substantial margin for privacy-preserving NILM.
- 16 December 2024
- Published by Elsevier in Pattern Recognition Letters
- Vol. 188, 117-124
Improving generative trajectory prediction via collision-free modeling and goal scene reconstruction
- 16 December 2024
- Published by Elsevier in Pattern Recognition Letters
- Vol. 188, 117-124
Improving generative trajectory prediction via collision-free modeling and goal scene reconstruction
- 11 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Systems
- Vol. 40 (1) , 1164-1167
Identifying the Largest RoCoF and Its Implications
The rate of change of frequency (RoCoF) is a critical factor in ensuring frequency security, particularly in power systems with low inertia. Currently, most RoCoF security constrained optimal inertia dispatch methods predominantly rely on aggregated frequency models which are tractable in optimization problems. These models, however, do not account for the disparities in post-contingency frequency dynamics across different regions of a power system. Specifically, regional buses can exhibit significantly larger RoCoFs than that predicted by an aggregated frequency model, particularly in systems characterized by unevenly distributed inertia. To bridge this gap, in this paper, the post-contingency nodal RoCoF model is first derived on the basis of augmented direct current power flow equations and the intrinsic nature that (virtual) synchronous generators will keep their rotor angles constant immediately after a disturbance. Then, the maximal initial RoCoF is mathematically proven to be at one of buses with inertia. This finding together with the nodal RoCoF model contribute to the establishment of the optimal nodal inertia dispatch method in a convex and concise form. The effectiveness of our finding and the proposed nodal inertia dispatch method are further verified by simulation results of two typical interconnected power systems under different scenarios.
- 11 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Systems
- Vol. 40 (1) , 1164-1167
Identifying the Largest RoCoF and Its Implications
The rate of change of frequency (RoCoF) is a critical factor in ensuring frequency security, particularly in power systems with low inertia. Currently, most RoCoF security constrained optimal inertia dispatch methods predominantly rely on aggregated frequency models which are tractable in optimization problems. These models, however, do not account for the disparities in post-contingency frequency dynamics across different regions of a power system. Specifically, regional buses can exhibit significantly larger RoCoFs than that predicted by an aggregated frequency model, particularly in systems characterized by unevenly distributed inertia. To bridge this gap, in this paper, the post-contingency nodal RoCoF model is first derived on the basis of augmented direct current power flow equations and the intrinsic nature that (virtual) synchronous generators will keep their rotor angles constant immediately after a disturbance. Then, the maximal initial RoCoF is mathematically proven to be at one of buses with inertia. This finding together with the nodal RoCoF model contribute to the establishment of the optimal nodal inertia dispatch method in a convex and concise form. The effectiveness of our finding and the proposed nodal inertia dispatch method are further verified by simulation results of two typical interconnected power systems under different scenarios.