Zhe Dong
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
Zhe Dong has a total of 57 co-authors. Here are the top co-authors by number of shared publications:
Xiaojin Huang
23 shared publicationsZuoyi Zhang
17 shared publications- Yujie Dong
Institute of Nuclear and New Energy Technology, Key Laboratory for Advanced Reactor Engineering and Safety of Ministry of Education, Tsinghua University 1 , Beijing 100084
10 shared publications Yujie Dong
7 shared publicationsBowen Li
6 shared publications- Xiaojin Huang
Collaborative Innovation Center of Advanced Nuclear Energy Technology of China, Key Laboratory of Advanced Reactor Engineering and Safety of Ministry of Education, Institute of Nuclear and New Energy Technology, Tsinghua University, Beijing 100084, China
6 shared publications
Latest Publications
- 2 November 2026
- Published by Elsevier in Neural Networks
- Vol. 206
High-order neural network extended state-observer with nuclear application
- 2 November 2026
- Published by Elsevier in Neural Networks
- Vol. 206
High-order neural network extended state-observer with nuclear application
- Open Access
Research on Frequency-Domain Learning-Based Fault Diagnosis Methods in Nuclear Power Plants via a Lightweight Complex-Valued Neural Network
Nuclear power plant (NPP) fault diagnosis is critical to ensure the safe, stable and economic operation of nuclear facilities. Existing deep learning-based NPP fault diagnosis methods primarily extract features from the temporal and spatial domain. Given recent advances in frequency-domain learning and lightweight models for time-series modeling, this paper proposes a frequency-domain learning-based fault diagnosis method for multi-modular high-temperature gas-cooled reactor (mHTGR) modules. It extracts temporal and spatial frequency-domain features using a lightweight complex-valued neural network (CVNN), which are then applied for fault classification and severity estimation. Under typical fault detection tasks for an mHTGR, comparative experiments verify that the proposed method outperforms existing methods that utilize long short-term memory (LSTM) and graph neural networks (GNNs) for spatiotemporal feature extraction in terms of classification accuracy, mean squared error (MSE) loss, and inference time.
Open AccessResearch on Frequency-Domain Learning-Based Fault Diagnosis Methods in Nuclear Power Plants via a Lightweight Complex-Valued Neural Network
Nuclear power plant (NPP) fault diagnosis is critical to ensure the safe, stable and economic operation of nuclear facilities. Existing deep learning-based NPP fault diagnosis methods primarily extract features from the temporal and spatial domain. Given recent advances in frequency-domain learning and lightweight models for time-series modeling, this paper proposes a frequency-domain learning-based fault diagnosis method for multi-modular high-temperature gas-cooled reactor (mHTGR) modules. It extracts temporal and spatial frequency-domain features using a lightweight complex-valued neural network (CVNN), which are then applied for fault classification and severity estimation. Under typical fault detection tasks for an mHTGR, comparative experiments verify that the proposed method outperforms existing methods that utilize long short-term memory (LSTM) and graph neural networks (GNNs) for spatiotemporal feature extraction in terms of classification accuracy, mean squared error (MSE) loss, and inference time.
- Top Cited Papers
Multimodular load following coordinated control of high temperature gas-cooled reactor nuclear cogeneration plants
- 2 January 2026
- Published by Springer Nature
- p. 259-269
Research on the Control Strategies of mHTGR Cogeneration Unit Under Abnormal Working Conditions
- 2 January 2026
- Published by Springer Nature
- p. 259-269
Research on the Control Strategies of mHTGR Cogeneration Unit Under Abnormal Working Conditions
- Open Access
Coordinated Control Optimization of Nuclear Steam Supply Systems via Multi-Agent Reinforcement Learning
Nuclear steam supply systems (NSSSs) are critical to achieving safe, efficient, and flexible nuclear power generation. While deep reinforcement learning (DRL) has shown potential in optimizing NSSS control, existing single-agent approaches apply the same optimization strategies to all subsystems. This simplification ignores subsystem-specific control requirements and limits both optimization efficacy and adaptability. To resolve this gap, we propose a multi-agent reinforcement learning (MARL) framework that independently optimizes each subsystem while ensuring global coordination. Our approach extends the current NSSS optimization framework from a single-agent model to a multi-agent one and introduces a novel MARL method to foster effective exploration and stability during optimization. Experimental findings demonstrate that our method significantly outperforms DRL-based approaches in optimizing thermal power and outlet steam temperature control. This research pioneers the application of MARL to NSSS optimization, paving the way for advanced nuclear power control systems.
Open AccessCoordinated Control Optimization of Nuclear Steam Supply Systems via Multi-Agent Reinforcement Learning
Nuclear steam supply systems (NSSSs) are critical to achieving safe, efficient, and flexible nuclear power generation. While deep reinforcement learning (DRL) has shown potential in optimizing NSSS control, existing single-agent approaches apply the same optimization strategies to all subsystems. This simplification ignores subsystem-specific control requirements and limits both optimization efficacy and adaptability. To resolve this gap, we propose a multi-agent reinforcement learning (MARL) framework that independently optimizes each subsystem while ensuring global coordination. Our approach extends the current NSSS optimization framework from a single-agent model to a multi-agent one and introduces a novel MARL method to foster effective exploration and stability during optimization. Experimental findings demonstrate that our method significantly outperforms DRL-based approaches in optimizing thermal power and outlet steam temperature control. This research pioneers the application of MARL to NSSS optimization, paving the way for advanced nuclear power control systems.
- Open Access
An Improved Dynamic Matrix Control Algorithm and Its Application in Cold Helium Temperature Control of a Modular High-Temperature Gas-Cooled Reactor (mHTGR)
As a model predictive control (MPC) technique, dynamic matrix control (DMC) has gained widespread industrial adoption due to its straightforward model construction and clear physical interpretation. However, its effectiveness relies on the accuracy of the predictive model, where measurement inaccuracies or excessive noise in step-response coefficients may significantly degrade control performance. This study enhances robustness of DMC by implementing finite impulse response (FIR) filters on measured step-response coefficients while providing theoretical proof of its stability. The improved algorithm is applied to cold helium temperature control of the modular High-Temperature Gas-Cooled Reactor (mHTGR). A cascade control structure is adopted, where the inner loop uses a PID controller to ensure system stability, while the outer loop uses DMC to adjust the setpoint of the hot helium temperature, thereby controlling the cold helium temperature. Numerical simulation results demonstrate significant improvements in temperature control performance and enhanced robustness of the modified DMC method.
Open AccessAn Improved Dynamic Matrix Control Algorithm and Its Application in Cold Helium Temperature Control of a Modular High-Temperature Gas-Cooled Reactor (mHTGR)
As a model predictive control (MPC) technique, dynamic matrix control (DMC) has gained widespread industrial adoption due to its straightforward model construction and clear physical interpretation. However, its effectiveness relies on the accuracy of the predictive model, where measurement inaccuracies or excessive noise in step-response coefficients may significantly degrade control performance. This study enhances robustness of DMC by implementing finite impulse response (FIR) filters on measured step-response coefficients while providing theoretical proof of its stability. The improved algorithm is applied to cold helium temperature control of the modular High-Temperature Gas-Cooled Reactor (mHTGR). A cascade control structure is adopted, where the inner loop uses a PID controller to ensure system stability, while the outer loop uses DMC to adjust the setpoint of the hot helium temperature, thereby controlling the cold helium temperature. Numerical simulation results demonstrate significant improvements in temperature control performance and enhanced robustness of the modified DMC method.
- Top Cited PapersOpen Access
- 21 March 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-14
Testing the feasibility of multi-modular design in an HTR-PM nuclear plant
Small modular reactors are the nuclear fission reactors with electric power output less than 300 MWe. To apply the enhanced safety of small modular reactors to build large-scale nuclear plants with any desired power ratings, the multi-modular scheme is recommended to be adopted, where multiple reactor modules are utilized to drive the common load equipment for power generation or cogeneration. The feasibility of multi-modular scheme is not verified until several plant-wide tests were carried out recently on the high temperature gas-cooled reactor pebble-bed module (HTR-PM) nuclear plant. The HTR-PM plant consists of two inherently safe nuclear reactors of 200 MWt, adopts the scheme of two reactor modules driving a common steam turbine, and operates commercially since December 6, 2023. In this paper, the responses of key process variables of HTR-PM plant in the tests of power ramping, turbine trip and reactor trip are provided, and the related multi-modular coordinated control method is also proposed. This result manifests the feasibility of multi-modular scheme practically, and shows the promising future of building large-scale nuclear plants with a system of small modular reactors.
- 21 March 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-14
Top Cited PapersOpen AccessTesting the feasibility of multi-modular design in an HTR-PM nuclear plant
Small modular reactors are the nuclear fission reactors with electric power output less than 300 MWe. To apply the enhanced safety of small modular reactors to build large-scale nuclear plants with any desired power ratings, the multi-modular scheme is recommended to be adopted, where multiple reactor modules are utilized to drive the common load equipment for power generation or cogeneration. The feasibility of multi-modular scheme is not verified until several plant-wide tests were carried out recently on the high temperature gas-cooled reactor pebble-bed module (HTR-PM) nuclear plant. The HTR-PM plant consists of two inherently safe nuclear reactors of 200 MWt, adopts the scheme of two reactor modules driving a common steam turbine, and operates commercially since December 6, 2023. In this paper, the responses of key process variables of HTR-PM plant in the tests of power ramping, turbine trip and reactor trip are provided, and the related multi-modular coordinated control method is also proposed. This result manifests the feasibility of multi-modular scheme practically, and shows the promising future of building large-scale nuclear plants with a system of small modular reactors.
- 18 November 2024
- Published by Elsevier in Renewable and Sustainable Energy Reviews
- Vol. 209
The definition of entropy production metric with application in passivity-based control of thermodynamic systems
- 18 November 2024
- Published by Elsevier in Renewable and Sustainable Energy Reviews
- Vol. 209
The definition of entropy production metric with application in passivity-based control of thermodynamic systems