Yulin Liu
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
Yulin Liu has a total of 37 co-authors. Here are the top co-authors by number of shared publications:
- Xinan Zhang
University of Western Australia
17 shared publications - Herbert Ho-Ching Iu
The University of Western Australia
15 shared publications - Tyrone Fernando
School of Engineering, University of Western Australia, Crawley, WA, Australia
15 shared publications - Tianhao Qie
Department of Electrical Engineering, City University of Hong Kong, Hong Kong
10 shared publications - Wendong Feng
School of Engineering, University of Western Australia, Crawley, WA, Australia
5 shared publications Ran Li
School of Engineering, University of Western Australia, Crawley, WA, Australia
4 shared publications
Latest Publications
- 26 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Emerging and Selected Topics in Industrial Electronics
- Vol. 7 (3) , 1036-1051
A Novel Robust and Adaptive Neural Network-Based Control Method for Proton Exchange Membrane Fuel Cells in DC Microgrids
Proton exchange membrane fuel cells (PEMFCs) are promising for DC microgrid applications, but their control is challenging due to complex dynamics, unpredictable disturbances, and limited sensor availability. Most advanced control methods rely on accurate models or extensive offline training, which can be costly and time-consuming. This work presents an online learning-based control method that does not require a detailed model and uses only three readily measurable variables: output voltage, voltage reference, and fuel flow rate. The approach employs two lightweight radial basis function neural net works, one to capture the slow-varying system dynamics and the other to estimate fast disturbances. Moreover, relative to existing prescribed performance control schemes, this paper introduces a novel prescribed function that is independent of the initial conditions of the nonlinear system and the reference signal. This design enables fast adaptation to disturbances and guarantees that, starting from an unknown initial control policy, the transient tracking error remains within the prescribed bounds for all time. The superior performance of the proposed method is validated through hardware-in-the-loop (HIL) experiments.
- 26 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Emerging and Selected Topics in Industrial Electronics
- Vol. 7 (3) , 1036-1051
A Novel Robust and Adaptive Neural Network-Based Control Method for Proton Exchange Membrane Fuel Cells in DC Microgrids
Proton exchange membrane fuel cells (PEMFCs) are promising for DC microgrid applications, but their control is challenging due to complex dynamics, unpredictable disturbances, and limited sensor availability. Most advanced control methods rely on accurate models or extensive offline training, which can be costly and time-consuming. This work presents an online learning-based control method that does not require a detailed model and uses only three readily measurable variables: output voltage, voltage reference, and fuel flow rate. The approach employs two lightweight radial basis function neural net works, one to capture the slow-varying system dynamics and the other to estimate fast disturbances. Moreover, relative to existing prescribed performance control schemes, this paper introduces a novel prescribed function that is independent of the initial conditions of the nonlinear system and the reference signal. This design enables fast adaptation to disturbances and guarantees that, starting from an unknown initial control policy, the transient tracking error remains within the prescribed bounds for all time. The superior performance of the proposed method is validated through hardware-in-the-loop (HIL) experiments.
- 21 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Sustainable Energy
- Vol. 17 (2) , 1847-1864
A Novel Learning-Based Optimal Control Strategy for SOFCs in DC Microgrids Using the Recurrent Equilibrium Network
This paper presents a feedback recurrent equilibrium network (REN)-based linear quadratic regulator for enhancing the control performance of solid oxide fuel cells (SOFCs) in DC microgrids. The REN is trained offline to approximate unmodeled system dynamics and compensate for disturbances, enabling efficient real-time implementation. The proposed method requires no more than 20 $\mu s$ computational time per control cycle and achieves high control accuracy, with the mean absolute percentage error (MAPE) not exceeding 0.2% in stack voltage regulation. The control stability is rigorously proven. Hardware-in-the-loop (HIL) experiments further demonstrate the practical applicability and superior performance of the proposed controller compared with various methods.
- 21 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Sustainable Energy
- Vol. 17 (2) , 1847-1864
A Novel Learning-Based Optimal Control Strategy for SOFCs in DC Microgrids Using the Recurrent Equilibrium Network
This paper presents a feedback recurrent equilibrium network (REN)-based linear quadratic regulator for enhancing the control performance of solid oxide fuel cells (SOFCs) in DC microgrids. The REN is trained offline to approximate unmodeled system dynamics and compensate for disturbances, enabling efficient real-time implementation. The proposed method requires no more than 20 $\mu s$ computational time per control cycle and achieves high control accuracy, with the mean absolute percentage error (MAPE) not exceeding 0.2% in stack voltage regulation. The control stability is rigorously proven. Hardware-in-the-loop (HIL) experiments further demonstrate the practical applicability and superior performance of the proposed controller compared with various methods.
- 2 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Electronics
- Vol. 41 (1) , 169-173
A Model-Independent Online Learning-Based Control Strategy for DC/AC Inverters
This letter proposes a novel control scheme for power electronic inverters using a barrier Lyapunov function guided radial basis function neural network controller, featuring online learning and real-time applicability. Unlike many existing adaptive neural network-based controller, the proposed method requires no knowledge of system parameters and does not require any offline training. The control law is updated entirely online with guaranteed convergence, ensuring bounded current tracking under uncertainties and disturbances. Its simple structure leads to extremely low computational complexity, making it one of the most efficient model-free controller currently applicable to real-time dc–ac inverter control. The effectiveness and robustness of the proposed controller are verified through application to a three-level neutral-point-clamped inverter.
- 2 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Electronics
- Vol. 41 (1) , 169-173
A Model-Independent Online Learning-Based Control Strategy for DC/AC Inverters
This letter proposes a novel control scheme for power electronic inverters using a barrier Lyapunov function guided radial basis function neural network controller, featuring online learning and real-time applicability. Unlike many existing adaptive neural network-based controller, the proposed method requires no knowledge of system parameters and does not require any offline training. The control law is updated entirely online with guaranteed convergence, ensuring bounded current tracking under uncertainties and disturbances. Its simple structure leads to extremely low computational complexity, making it one of the most efficient model-free controller currently applicable to real-time dc–ac inverter control. The effectiveness and robustness of the proposed controller are verified through application to a three-level neutral-point-clamped inverter.
- 16 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Smart Grid
- Vol. 16 (6) , 4348-4365
A Stable Recurrent Equilibrium Network-Based Controller for Proton Exchange Membrane Fuel Cell in DC Microgrids
This paper proposes a model-insensitive artificial intelligence control method for proton exchange membrane fuel cells (PEMFCs) in DC microgrids. The proposed method employs the recently developed recurrent equilibrium network (REN) to achieve model-independent PEMFC regulation. While REN has built-in stability, how to ensure the stability of the closed-loop control system formed by the REN and the target plant remains a challenge, which hinders the practical application of REN. To address this key obstacle, this study proposes a novel control strategy that provides a systematic closed-loop stability proof by incorporating the REN into PEMFC control systems through Youla parameterization. The guaranteed control stability makes the proposed method superior over most PEMFC control approaches that rely on offline trained neural network. Furthermore, compared to few online learning-based PEMFC control approaches with stability proof, the proposed method offers much lower computational burden. The superiority of the proposed method is validated through hardware-in-the-loop (HIL) experiments and digital signal processor (DSP).
- 16 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Smart Grid
- Vol. 16 (6) , 4348-4365
A Stable Recurrent Equilibrium Network-Based Controller for Proton Exchange Membrane Fuel Cell in DC Microgrids
This paper proposes a model-insensitive artificial intelligence control method for proton exchange membrane fuel cells (PEMFCs) in DC microgrids. The proposed method employs the recently developed recurrent equilibrium network (REN) to achieve model-independent PEMFC regulation. While REN has built-in stability, how to ensure the stability of the closed-loop control system formed by the REN and the target plant remains a challenge, which hinders the practical application of REN. To address this key obstacle, this study proposes a novel control strategy that provides a systematic closed-loop stability proof by incorporating the REN into PEMFC control systems through Youla parameterization. The guaranteed control stability makes the proposed method superior over most PEMFC control approaches that rely on offline trained neural network. Furthermore, compared to few online learning-based PEMFC control approaches with stability proof, the proposed method offers much lower computational burden. The superiority of the proposed method is validated through hardware-in-the-loop (HIL) experiments and digital signal processor (DSP).
- 7 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Smart Grid
- Vol. 16 (4) , 2774-2790
A Novel Barrier Lyapunov Function-Based Online Learning Control Method for Solid Oxide Fuel Cell in DC Microgrids
This paper proposes a novel barrier Lyapunov function (BLF)-based online learning control method to enhance the performance of solid oxide fuel cells (SOFCs) in DC microgrids. Leveraging the superior function approximation capability of the radial basis function neural network (RBFNN) and employing a dual RBFNN framework, where one network approximates long-term system dynamics and the other captures rapidly changing disturbances, the proposed method achieves excellent control performance while requiring only input-output data, without any prior knowledge of the system model. The incorporation of the BLF ensures that tracking errors never exceed predefined limits at any time. By precisely regulating the output of SOFC, the proposed control method ensures a stable voltage level in the DC microgrid, thus effectively mitigating fluctuations that may affect system performance and improving the overall reliability and efficiency of the microgrid. The superior performance of the proposed method is validated through hardware-in-the-loop (HIL) experiments.
- 7 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Smart Grid
- Vol. 16 (4) , 2774-2790
A Novel Barrier Lyapunov Function-Based Online Learning Control Method for Solid Oxide Fuel Cell in DC Microgrids
This paper proposes a novel barrier Lyapunov function (BLF)-based online learning control method to enhance the performance of solid oxide fuel cells (SOFCs) in DC microgrids. Leveraging the superior function approximation capability of the radial basis function neural network (RBFNN) and employing a dual RBFNN framework, where one network approximates long-term system dynamics and the other captures rapidly changing disturbances, the proposed method achieves excellent control performance while requiring only input-output data, without any prior knowledge of the system model. The incorporation of the BLF ensures that tracking errors never exceed predefined limits at any time. By precisely regulating the output of SOFC, the proposed control method ensures a stable voltage level in the DC microgrid, thus effectively mitigating fluctuations that may affect system performance and improving the overall reliability and efficiency of the microgrid. The superior performance of the proposed method is validated through hardware-in-the-loop (HIL) experiments.
- 7 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Electronics
- Vol. 40 (5) , 6593-6604
A Novel Adaptive Model Predictive Current Control for Three-Level Simplified Neutral-Point-Clamped Inverter With Carrier-Based Modulation
This article proposes an adaptive model predictive current control for three-level simplified neutral-point-clamped inverter (3L-SNPCI), which eliminates the model dependence problem of the conventional model predictive current control (MPCC). The system parameter is estimated by a new observer with high accuracy and fast convergence. The observed parameter is then utilized by the deadbeat MPCC to attain accurate reference voltage vector calculation, achieving low computational burden and constant switching frequency. Furthermore, a novel carrier-based space vector modulation (CBSVM) is proposed to synthesize the reference VV for 3L-SNPCI. Compared to the existing space vector modulation schemes of 3L-SNPCI, the proposed CBSVM reduces the hardware implementation cost while ensuring the neutral point voltage balance. Experimental results are presented to verify the validity of the proposed method.
- 7 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Electronics
- Vol. 40 (5) , 6593-6604
A Novel Adaptive Model Predictive Current Control for Three-Level Simplified Neutral-Point-Clamped Inverter With Carrier-Based Modulation
This article proposes an adaptive model predictive current control for three-level simplified neutral-point-clamped inverter (3L-SNPCI), which eliminates the model dependence problem of the conventional model predictive current control (MPCC). The system parameter is estimated by a new observer with high accuracy and fast convergence. The observed parameter is then utilized by the deadbeat MPCC to attain accurate reference voltage vector calculation, achieving low computational burden and constant switching frequency. Furthermore, a novel carrier-based space vector modulation (CBSVM) is proposed to synthesize the reference VV for 3L-SNPCI. Compared to the existing space vector modulation schemes of 3L-SNPCI, the proposed CBSVM reduces the hardware implementation cost while ensuring the neutral point voltage balance. Experimental results are presented to verify the validity of the proposed method.
- 23 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 21 (3) , 2580-2589
A Novel Online Learning-Based Optimal Control Algorithm for Enhancing Solid Oxide Fuel Cells Performance in Islanded DC Microgrids
Solid oxide fuel cells (SOFCs) offer a promising solution for enhancing reliability and sustainability in microgrid power supply with the growing penetration of renewable energy sources. The proposed method addresses key challenges in existing SOFC control approaches, including model dependence, usage of nonoptimal control policy, reliance on an offline-trained neural network (NN), and complex design. Compared with model-based methods, this method uses NN and policy iteration technology to learn system dynamics and approximate optimal control policy, thereby eliminating model dependence. Compared with offline learning-based methods, this method achieves online policy evaluation and NN updating to eliminate tedious offline training and data acquisition processes. Compared with the online learning-based SOFC control approaches, this method employs a fixed-weight recurrent NN to avoid slow or even no convergence caused by recursive least squares-based NN weights updating process, reducing design complexity without sacrificing control performance. The superiority of the proposed method is validated through hardware-in-the-loop tests.
- 23 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 21 (3) , 2580-2589
A Novel Online Learning-Based Optimal Control Algorithm for Enhancing Solid Oxide Fuel Cells Performance in Islanded DC Microgrids
Solid oxide fuel cells (SOFCs) offer a promising solution for enhancing reliability and sustainability in microgrid power supply with the growing penetration of renewable energy sources. The proposed method addresses key challenges in existing SOFC control approaches, including model dependence, usage of nonoptimal control policy, reliance on an offline-trained neural network (NN), and complex design. Compared with model-based methods, this method uses NN and policy iteration technology to learn system dynamics and approximate optimal control policy, thereby eliminating model dependence. Compared with offline learning-based methods, this method achieves online policy evaluation and NN updating to eliminate tedious offline training and data acquisition processes. Compared with the online learning-based SOFC control approaches, this method employs a fixed-weight recurrent NN to avoid slow or even no convergence caused by recursive least squares-based NN weights updating process, reducing design complexity without sacrificing control performance. The superiority of the proposed method is validated through hardware-in-the-loop tests.
- 20 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Emerging and Selected Topics in Power Electronics
- Vol. 13 (2) , 2234-2245
A Novel Adaptive Model Predictive Current Control for Three-Level Neural-Point-Clamped Inverter With RL Load
Parameter uncertainty has been a key challenge for the model predictive current control (MPCC) of a three-level neutral-point clamped (3L-NPC) inverter. To solve this issue, this article proposes an adaptive MPCC (AMPCC), which employs a novel online observer to estimate the system parameter with guaranteed convergence. The observed parameter is then utilized in the deadbeat MPCC to accurately calculate the duty ratios of voltage vectors (VVs). Moreover, to eliminate the tedious weighting factor selection process, a new simplified neutral-point (NP) voltage balancing method is proposed to rearrange the distribution time of the complementary VVs in each control period, resulting in precise dc-link voltage balance. Noticeably, the proposed NPV balancing method also attains very fast dynamic responses when the input voltage changes. Consequently, the proposed AMPCC obtains superior control performance with the constant switching frequency, while imposing a low computational burden. Thorough experimental assessments have been carried out to verify the effectiveness of the proposed method.
- 20 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Emerging and Selected Topics in Power Electronics
- Vol. 13 (2) , 2234-2245
A Novel Adaptive Model Predictive Current Control for Three-Level Neural-Point-Clamped Inverter With RL Load
Parameter uncertainty has been a key challenge for the model predictive current control (MPCC) of a three-level neutral-point clamped (3L-NPC) inverter. To solve this issue, this article proposes an adaptive MPCC (AMPCC), which employs a novel online observer to estimate the system parameter with guaranteed convergence. The observed parameter is then utilized in the deadbeat MPCC to accurately calculate the duty ratios of voltage vectors (VVs). Moreover, to eliminate the tedious weighting factor selection process, a new simplified neutral-point (NP) voltage balancing method is proposed to rearrange the distribution time of the complementary VVs in each control period, resulting in precise dc-link voltage balance. Noticeably, the proposed NPV balancing method also attains very fast dynamic responses when the input voltage changes. Consequently, the proposed AMPCC obtains superior control performance with the constant switching frequency, while imposing a low computational burden. Thorough experimental assessments have been carried out to verify the effectiveness of the proposed method.
- 8 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Emerging and Selected Topics in Industrial Electronics
- Vol. 6 (4) , 1319-1328
A Novel Neural Network-Based Control Method for Proton Exchange Membrane Fuel Cell in DC Microgrids
This paper proposes a novel neural network-based control algorithm for the proton exchange membrane fuel cell (PEMFC) in DC microgrids. A novel recurrent equilibrium network (REN) is developed to improve the control performance in the presence of uncertainties and disturbances. In theory, the contraction of the REN is guaranteed, regardless of the inputs or initial conditions. Furthermore, the REN is directly parameterized, thereby simplifying the neural network training process without sacrificing learning performance. These features make REN particularly suitable for machine learning and system identification tasks. In summary, the proposed REN-based PI control method has very low computational complexity, excellent control performance, and verified stability. The effectiveness of the proposed control algorithm is validated through hardware-in-the-loop experiments
- 8 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Emerging and Selected Topics in Industrial Electronics
- Vol. 6 (4) , 1319-1328
A Novel Neural Network-Based Control Method for Proton Exchange Membrane Fuel Cell in DC Microgrids
This paper proposes a novel neural network-based control algorithm for the proton exchange membrane fuel cell (PEMFC) in DC microgrids. A novel recurrent equilibrium network (REN) is developed to improve the control performance in the presence of uncertainties and disturbances. In theory, the contraction of the REN is guaranteed, regardless of the inputs or initial conditions. Furthermore, the REN is directly parameterized, thereby simplifying the neural network training process without sacrificing learning performance. These features make REN particularly suitable for machine learning and system identification tasks. In summary, the proposed REN-based PI control method has very low computational complexity, excellent control performance, and verified stability. The effectiveness of the proposed control algorithm is validated through hardware-in-the-loop experiments
- 30 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 25 (12) , 21597-21607
A Novel Machine Learning-Based Online Optimal Control Strategy for Fuel Cell in Electrified Transportation System
Proton exchange membrane fuel cells (PEMFCs) have been widely used as clean energy storage devices in electrified transportation system. The key challenges in the existing control approaches of PEMFCs include model dependence, the usage of non-optimal control policy and the reliance on offline-trained neural networks. To address these challenges, this paper proposes a novel machine learning-based optimal control strategy for the PEMFC in electrified transportation system. Furthermore, the proposed method employs a recurrent neural network (RNN) to successfully avoid the problem of slow or even no convergence that may be caused in recursive least square-based neural network weights updating process. It offers excellent control performance with guaranteed convergence and stability. The superiority of the proposed method is validated through Hardware-in-the-Loop (HIL) tests.
- 30 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 25 (12) , 21597-21607
A Novel Machine Learning-Based Online Optimal Control Strategy for Fuel Cell in Electrified Transportation System
Proton exchange membrane fuel cells (PEMFCs) have been widely used as clean energy storage devices in electrified transportation system. The key challenges in the existing control approaches of PEMFCs include model dependence, the usage of non-optimal control policy and the reliance on offline-trained neural networks. To address these challenges, this paper proposes a novel machine learning-based optimal control strategy for the PEMFC in electrified transportation system. Furthermore, the proposed method employs a recurrent neural network (RNN) to successfully avoid the problem of slow or even no convergence that may be caused in recursive least square-based neural network weights updating process. It offers excellent control performance with guaranteed convergence and stability. The superiority of the proposed method is validated through Hardware-in-the-Loop (HIL) tests.