Siyuan Zhan
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
Research Subjects
Siyuan Zhan has a total of 28 co-authors. Here are the top co-authors by number of shared publications:
- Guang Li
School of Engineering, The University of Manchester, Manchester, U.K.
9 shared publications - Jing Na
Kunming University of Science and Technology
4 shared publications - Bin Wang
Faculty of Mechanical and Electrical Engineering, Kunming University of Science and Technology, Kunming, China
3 shared publications - John Vincent Ringwood
Centre for Ocean Energy Research, Maynooth University, W23 F2H6 Maynooth, Ireland
3 shared publications - Zhijing Liao
School of Engineering, University of Manchester, Manchester, U.K.
3 shared publications - Christopher Edwards
College of Engineering, Mathematics and Physical Sciences, University of Exeter, Exeter, U.K.
2 shared publications
Latest Publications
- Open Access
- 2 October 2026
- Published by Elsevier in SSRN Electronic Journal
Dynamic Multi-UAV Offshore Wind Farm Inspection via Feasibility-Aware Task Reassignment and Localized Route Recovery
Inspection of large offshore wind farms (OWFs) requires efficient coverage of spatially distributed turbines within limited operational windows. Cooperative deployment of multiple unmanned aerial vehicles (UAVs) enables parallel inspection and provides operational redundancy; however, changing wind conditions, limited battery endurance, flight-safety constraints, and unexpected changes in vehicle availability and inspection demand can alter flight feasibility and mission costs during operation, requiring task assignments and flight routes to be adapted accordingly. This study therefore develops a wind-aware dynamic mission planning framework that incorporates wind-dependent flight performance, energy consumption, safe-return requirements, and dynamic mission events. A hierarchical strategy separates task reassignment from route recovery: a constraint-aware deep Q-network (CADQN) assigns released or newly generated tasks only to feasible UAVs, while a memory-guided protective genetic algorithm (MGPGA) reconstructs affected route segments while preserving unaffected feasible portions. A site-informed OWF scenario is used for evaluation. Results show that offshore wind conditions alter the battery-constrained reachable region and affect mission-level energy consumption and completion time. Compared with a coupled DQN strategy, the proposed framework reduces total flight distance, energy consumption, and mission completion time by 10.7%, 13.1%, and 12.6%, respectively, demonstrating effective adaptive mission recovery under dynamic offshore operating conditions.
- 2 October 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessDynamic Multi-UAV Offshore Wind Farm Inspection via Feasibility-Aware Task Reassignment and Localized Route Recovery
Inspection of large offshore wind farms (OWFs) requires efficient coverage of spatially distributed turbines within limited operational windows. Cooperative deployment of multiple unmanned aerial vehicles (UAVs) enables parallel inspection and provides operational redundancy; however, changing wind conditions, limited battery endurance, flight-safety constraints, and unexpected changes in vehicle availability and inspection demand can alter flight feasibility and mission costs during operation, requiring task assignments and flight routes to be adapted accordingly. This study therefore develops a wind-aware dynamic mission planning framework that incorporates wind-dependent flight performance, energy consumption, safe-return requirements, and dynamic mission events. A hierarchical strategy separates task reassignment from route recovery: a constraint-aware deep Q-network (CADQN) assigns released or newly generated tasks only to feasible UAVs, while a memory-guided protective genetic algorithm (MGPGA) reconstructs affected route segments while preserving unaffected feasible portions. A site-informed OWF scenario is used for evaluation. Results show that offshore wind conditions alter the battery-constrained reachable region and affect mission-level energy consumption and completion time. Compared with a coupled DQN strategy, the proposed framework reduces total flight distance, energy consumption, and mission completion time by 10.7%, 13.1%, and 12.6%, respectively, demonstrating effective adaptive mission recovery under dynamic offshore operating conditions.
- 1 July 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Oceanic Engineering
- Vol. 51 (3) , 2091-2108
Numerical and Experimental Investigations on the Dynamic Coupling Characteristics of a Novel Semisubmersible Floating-Oscillating-Buoy Wave Energy Converter System
- 1 July 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Oceanic Engineering
- Vol. 51 (3) , 2091-2108
Numerical and Experimental Investigations on the Dynamic Coupling Characteristics of a Novel Semisubmersible Floating-Oscillating-Buoy Wave Energy Converter System
- Open Access
- 6 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Control Systems Technology
- Vol. 34 (3) , 1584-1591
Control-Informed Wave Prediction for Wave Energy Conversion
Energy maximization (EM) control for wave energy converters (WECs) is a noncausal control approach that explicitly or implicitly involves feedforward from future wave predictions. In the current noncausal control paradigm, the wave prediction problem and the control problem are considered separately. However, there is a lack of theoretical studies that elucidate the mechanism by which inaccuracies in wave prediction deteriorate EM control performance. This brief aims to address this gap by integrating control considerations into the development of wave predictors. Based on the assumption of a linear WEC, the brief first revisits the linear noncausal optimal control (LNOC) problem, establishing the connections between the control and wave prediction problems. Assuming a known wave spectrum, it is shown that the linear Gaussian process (GP)-based excitation force predictor, which optimizes the conditional covariance of excitation force prediction, also maintains statistical optimality in the WEC EM control problem. Subsequently, a novel time-series-based approach for direct acausal estimation is presented, which achieves the same performance as the optimal predictor but with significantly fewer parameters to determine. The implementation in practical scenarios with an unknown wave spectrum is also discussed. Finally, a numerical simulation, based on a benchmarked point absorber WEC (PA-WEC), is presented to demonstrate the implementation process and the efficacy of the developed predictor.
- 6 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Control Systems Technology
- Vol. 34 (3) , 1584-1591
Open AccessControl-Informed Wave Prediction for Wave Energy Conversion
Energy maximization (EM) control for wave energy converters (WECs) is a noncausal control approach that explicitly or implicitly involves feedforward from future wave predictions. In the current noncausal control paradigm, the wave prediction problem and the control problem are considered separately. However, there is a lack of theoretical studies that elucidate the mechanism by which inaccuracies in wave prediction deteriorate EM control performance. This brief aims to address this gap by integrating control considerations into the development of wave predictors. Based on the assumption of a linear WEC, the brief first revisits the linear noncausal optimal control (LNOC) problem, establishing the connections between the control and wave prediction problems. Assuming a known wave spectrum, it is shown that the linear Gaussian process (GP)-based excitation force predictor, which optimizes the conditional covariance of excitation force prediction, also maintains statistical optimality in the WEC EM control problem. Subsequently, a novel time-series-based approach for direct acausal estimation is presented, which achieves the same performance as the optimal predictor but with significantly fewer parameters to determine. The implementation in practical scenarios with an unknown wave spectrum is also discussed. Finally, a numerical simulation, based on a benchmarked point absorber WEC (PA-WEC), is presented to demonstrate the implementation process and the efficacy of the developed predictor.
- 27 June 2025
- Published by Elsevier in Ocean Engineering
- Vol. 338
Distributed nonlinear model predictive control of an array of wave energy converters
- 27 June 2025
- Published by Elsevier in Ocean Engineering
- Vol. 338
Distributed nonlinear model predictive control of an array of wave energy converters
- Open Access
- 29 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Control Systems Technology
- Vol. 32 (6) , 2164-2177
Model-Free Linear Noncausal Optimal Control of Wave Energy Converters via Reinforcement Learning
This article introduces a novel reinforcement learning (RL) method for wave energy converters (WECs), which directly generates linear noncausal optimal control (LNOC) policies on continuous action space. Unlike other existing WEC RL algorithms looking at the problem mainly from a learning perspective, the proposed RL approach adopts a control-theoretic approach by delving into the underlying WEC energy maximization (EM) optimal control problem (OCP). This leads to control-informed decisions on choosing the RL state, as well as developing the RL structure. The proposed model-free LNOC (MF-LNOC) offers substantial advantages, including significantly improved performance due to the use of noncausal information, a simplified RL with linear actor and quadratic critic structures, and remarkable fast convergence speeds, achieved using less than 150 s of data points, for a benchmarked point absorber, which can be further shortened using the replay technique. This reduction in training time allows for controller reconfiguration in pace with sea changes. Demonstrative numerical simulations are presented to verify the efficacy of the proposed methods. The proposed MF-LNOC also shows robustness against wave prediction inaccuracies and changing sea conditions. The MF-LNOC methodology can be highly attractive for WEC developers who want to design an efficient and reliable controller for WECs but also hope to avoid the challenge of establishing a control-oriented model that can preserve high fidelity over a wide range of sea conditions.
- 29 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Control Systems Technology
- Vol. 32 (6) , 2164-2177
Open AccessModel-Free Linear Noncausal Optimal Control of Wave Energy Converters via Reinforcement Learning
This article introduces a novel reinforcement learning (RL) method for wave energy converters (WECs), which directly generates linear noncausal optimal control (LNOC) policies on continuous action space. Unlike other existing WEC RL algorithms looking at the problem mainly from a learning perspective, the proposed RL approach adopts a control-theoretic approach by delving into the underlying WEC energy maximization (EM) optimal control problem (OCP). This leads to control-informed decisions on choosing the RL state, as well as developing the RL structure. The proposed model-free LNOC (MF-LNOC) offers substantial advantages, including significantly improved performance due to the use of noncausal information, a simplified RL with linear actor and quadratic critic structures, and remarkable fast convergence speeds, achieved using less than 150 s of data points, for a benchmarked point absorber, which can be further shortened using the replay technique. This reduction in training time allows for controller reconfiguration in pace with sea changes. Demonstrative numerical simulations are presented to verify the efficacy of the proposed methods. The proposed MF-LNOC also shows robustness against wave prediction inaccuracies and changing sea conditions. The MF-LNOC methodology can be highly attractive for WEC developers who want to design an efficient and reliable controller for WECs but also hope to avoid the challenge of establishing a control-oriented model that can preserve high fidelity over a wide range of sea conditions.
- Open Access
- 27 June 2023
- Published by Wiley in International Journal of Robust and Nonlinear Control
- Vol. 35 (7) , 2694-2716
Terminal weight and constraint design for wave energy converter economic model predictive control problems
The wave energy converter (WEC) control problem aims to make the best use of wave excitation to maximize energy capture and ensure safe operation across a broad range of sea states. This falls into the recently developed economic model predictive control (EMPC) framework, subject to wave excitation being treated as a predictable additive disturbance in the control problem. However, there are few theoretical developments on EMPC theory that can be directly used, as the persistent disturbances bring nontrivial problems to optimal operation, safety etc. Whilst the disturbance is beneficial for the energy production objective, it may cause safe operation problems. This article develops a systematic WEC EMPC design approach for WECs with essential linear hydrodynamical characteristics. By introducing terminal weight and terminal state constraint design into the WEC EMPC structure, important features such as convexity, satisfaction of safety constraints, and recursive feasibility can be guaranteed. Contrary to our intuition, we show that (1) the proposed EMPC with a terminal weight provides an unbiased estimate of the ideal conceptual infinite horizon solution, and (2) the WEC operational range, in terms of the operational sea states, is extended by imposing a terminal constraint. Finally, numerical simulations are provided to verify the efficacy of the proposed approach.
- 27 June 2023
- Published by Wiley in International Journal of Robust and Nonlinear Control
- Vol. 35 (7) , 2694-2716
Open AccessTerminal weight and constraint design for wave energy converter economic model predictive control problems
The wave energy converter (WEC) control problem aims to make the best use of wave excitation to maximize energy capture and ensure safe operation across a broad range of sea states. This falls into the recently developed economic model predictive control (EMPC) framework, subject to wave excitation being treated as a predictable additive disturbance in the control problem. However, there are few theoretical developments on EMPC theory that can be directly used, as the persistent disturbances bring nontrivial problems to optimal operation, safety etc. Whilst the disturbance is beneficial for the energy production objective, it may cause safe operation problems. This article develops a systematic WEC EMPC design approach for WECs with essential linear hydrodynamical characteristics. By introducing terminal weight and terminal state constraint design into the WEC EMPC structure, important features such as convexity, satisfaction of safety constraints, and recursive feasibility can be guaranteed. Contrary to our intuition, we show that (1) the proposed EMPC with a terminal weight provides an unbiased estimate of the ideal conceptual infinite horizon solution, and (2) the WEC operational range, in terms of the operational sea states, is extended by imposing a terminal constraint. Finally, numerical simulations are provided to verify the efficacy of the proposed approach.
- 17 February 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Sustainable Energy
- Vol. 14 (3) , 1737-1746
Modelling and Control Tank Testing Validation for Attenuator Type Wave Energy Converter – Part III: Model Predictive Control and Robustness Validation
Following the tank testing results of linear passive damping control and linear non-causal optimal control (LNOC) in the Part I and Part II papers, this paper presents further tank testing results focusing on two aspects: Firstly, a model predictive controller (MPC) is designed based on the model developed by system identification in Part I to optimally handle actuator saturation limits. The MPC control is demonstrated to significantly improve the energy output by 9.86% up to 463.42% compared with a well-tuned passive damper and can also outperform the LNOC presented in Part II in a range of irregular unidirectional waves. Secondly, the robustness of the MPC controller and the LNOC controller is validated in more realistic sea conditions when directional spreading waves and side waves are superimposed onto the dominant directional waves. The test results show that both MPC and LNOC can still significantly outperform the passive damping controller in all the sea conditions. These tank testing results pave a solid way for future sea trial testing of these advanced non-causal optimal control strategies developed for wave energy converters in sea trials.
- 17 February 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Sustainable Energy
- Vol. 14 (3) , 1737-1746
Modelling and Control Tank Testing Validation for Attenuator Type Wave Energy Converter – Part III: Model Predictive Control and Robustness Validation
Following the tank testing results of linear passive damping control and linear non-causal optimal control (LNOC) in the Part I and Part II papers, this paper presents further tank testing results focusing on two aspects: Firstly, a model predictive controller (MPC) is designed based on the model developed by system identification in Part I to optimally handle actuator saturation limits. The MPC control is demonstrated to significantly improve the energy output by 9.86% up to 463.42% compared with a well-tuned passive damper and can also outperform the LNOC presented in Part II in a range of irregular unidirectional waves. Secondly, the robustness of the MPC controller and the LNOC controller is validated in more realistic sea conditions when directional spreading waves and side waves are superimposed onto the dominant directional waves. The test results show that both MPC and LNOC can still significantly outperform the passive damping controller in all the sea conditions. These tank testing results pave a solid way for future sea trial testing of these advanced non-causal optimal control strategies developed for wave energy converters in sea trials.
- 17 February 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Sustainable Energy
- Vol. 14 (3) , 1758-1768
Modelling and Control Tank Testing Validation for Attenuator Type Wave Energy Converter - Part II: Linear Noncausal Optimal Control and Deterministic Sea Wave Prediction Tank Testing
Tank testing results from applying Linear non-causal optimal control (LNOC) to the Mocean M100 wave energy converter (WEC) are presented in this paper. Based on the two state-space models derived from Part I, two LNOC algorithms designed to increase power capture of the $1/20$ scaled M100 WEC prototype are tested separately in real-time under various irregular wave conditions. The deterministic sea wave prediction (DSWP) algorithm is employed to provide the required wave elevation/excitation force predictions for the LNOC using real-time wave measurements. Experimental data shows that the DSWP can predict the incoming waves with sufficiently high accuracy in various sea states. Compared with a well-tuned passive damper, LNOC based on a hydrodynamic model increases the power capture by up to 128% while LNOC based on a system identification (SysID) model increases the power capture by up to 323%. The technologies are transferrable. The testing procedure and the results also provide informative guidance for the tank testing of other advanced optimal controllers and their implementations on other types of WECs.
- 17 February 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Sustainable Energy
- Vol. 14 (3) , 1758-1768
Modelling and Control Tank Testing Validation for Attenuator Type Wave Energy Converter - Part II: Linear Noncausal Optimal Control and Deterministic Sea Wave Prediction Tank Testing
Tank testing results from applying Linear non-causal optimal control (LNOC) to the Mocean M100 wave energy converter (WEC) are presented in this paper. Based on the two state-space models derived from Part I, two LNOC algorithms designed to increase power capture of the $1/20$ scaled M100 WEC prototype are tested separately in real-time under various irregular wave conditions. The deterministic sea wave prediction (DSWP) algorithm is employed to provide the required wave elevation/excitation force predictions for the LNOC using real-time wave measurements. Experimental data shows that the DSWP can predict the incoming waves with sufficiently high accuracy in various sea states. Compared with a well-tuned passive damper, LNOC based on a hydrodynamic model increases the power capture by up to 128% while LNOC based on a system identification (SysID) model increases the power capture by up to 323%. The technologies are transferrable. The testing procedure and the results also provide informative guidance for the tank testing of other advanced optimal controllers and their implementations on other types of WECs.
- Open Access
- 3 November 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Control Systems Technology
- Vol. 31 (3) , 1443-1450
A Fast Model Predictive Control Framework for Multi-Float and Multi-Mode-Motion Wave Energy Converters
Recently, multi-float and multi-mode-motion wave energy converters (M-WECs) have been developed to improve energy conversion capability. Although model predictive control (MPC) can be very effective to solve the constrained energy maximization control problem of point absorber WECs, the increased complexity of the M-WEC hydrodynamics can bring significant challenges due to computational demand. This brief proposes a novel computational-efficient fast MPC (FMPC) design method for the M-WECs requiring complex linear hydrodynamic models. The controller design objective is to maximize the energy conversion with some available wave forecasting information and to satisfy state and control input constraints to ensure safe operation. The main advantage of the proposed FMPC is the reduced computational burden with a negligible impact on performance. A demonstrative numerical simulation based on a 1:50 laboratory-scale M-WEC design, M4, for which linear hydrodynamics has been verified experimentally, is presented to verify the efficacy of the proposed control method in terms of both computational load and energy output.
- 3 November 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Control Systems Technology
- Vol. 31 (3) , 1443-1450
Open AccessA Fast Model Predictive Control Framework for Multi-Float and Multi-Mode-Motion Wave Energy Converters
Recently, multi-float and multi-mode-motion wave energy converters (M-WECs) have been developed to improve energy conversion capability. Although model predictive control (MPC) can be very effective to solve the constrained energy maximization control problem of point absorber WECs, the increased complexity of the M-WEC hydrodynamics can bring significant challenges due to computational demand. This brief proposes a novel computational-efficient fast MPC (FMPC) design method for the M-WECs requiring complex linear hydrodynamic models. The controller design objective is to maximize the energy conversion with some available wave forecasting information and to satisfy state and control input constraints to ensure safe operation. The main advantage of the proposed FMPC is the reduced computational burden with a negligible impact on performance. A demonstrative numerical simulation based on a 1:50 laboratory-scale M-WEC design, M4, for which linear hydrodynamics has been verified experimentally, is presented to verify the efficacy of the proposed control method in terms of both computational load and energy output.
- 27 June 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Control Systems Technology
- Vol. 31 (2) , 543-554
An Offline Closed-Form Optimal Predictive Power Management Strategy for Plug-In Hybrid Electric Vehicles
This article develops an optimal predictive power management strategy (OP-PMS) for plug-in hybrid electric vehicles (PHEVs) to manage the remaining battery level throughout a defined journey. Apart from using the battery before charging it again at the destination, this approach also supports transit through an Ultra Low Emission Zone (ULEZ), where a PHEV needs to operate in pure electric mode to avoid emissions. This type of PHEV power management strategy (PMS) design problem is a noncausal control problem, where the power demand of the remaining driving cycle and the final battery energy level targets influence the current power management action. Rather than solving the whole optimization problem online, this article presents a simplification that leads to a closed-form analytic solution with parameters that can be computed offline. This article makes three key contributions. First, a novel input-linearization method is developed, which converts the OP-PMS into a convex quadratic programming (QP) problem that captures the essential model nonlinearities using an input transformation and a quadratic cost function. Second, the influence of future power demands and final targets on current energy management policy is explicitly quantified from a dynamic optimal control perspective. Third, the noncausal convex QP is approximated with a closed-form analytic solution that is calculated offline. This leads to a real-time implementable OP-PMS with coefficients that can be precalculated and stored in the engine control unit. To demonstrate the viability of this approach, numerical examples are provided based on a generic PHEV model to verify the efficacy of the proposed OP-PMS.
- 27 June 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Control Systems Technology
- Vol. 31 (2) , 543-554
An Offline Closed-Form Optimal Predictive Power Management Strategy for Plug-In Hybrid Electric Vehicles
This article develops an optimal predictive power management strategy (OP-PMS) for plug-in hybrid electric vehicles (PHEVs) to manage the remaining battery level throughout a defined journey. Apart from using the battery before charging it again at the destination, this approach also supports transit through an Ultra Low Emission Zone (ULEZ), where a PHEV needs to operate in pure electric mode to avoid emissions. This type of PHEV power management strategy (PMS) design problem is a noncausal control problem, where the power demand of the remaining driving cycle and the final battery energy level targets influence the current power management action. Rather than solving the whole optimization problem online, this article presents a simplification that leads to a closed-form analytic solution with parameters that can be computed offline. This article makes three key contributions. First, a novel input-linearization method is developed, which converts the OP-PMS into a convex quadratic programming (QP) problem that captures the essential model nonlinearities using an input transformation and a quadratic cost function. Second, the influence of future power demands and final targets on current energy management policy is explicitly quantified from a dynamic optimal control perspective. Third, the noncausal convex QP is approximated with a closed-form analytic solution that is calculated offline. This leads to a real-time implementable OP-PMS with coefficients that can be precalculated and stored in the engine control unit. To demonstrate the viability of this approach, numerical examples are provided based on a generic PHEV model to verify the efficacy of the proposed OP-PMS.