Periodic Reporting for period 1 - EDOWE (Economical Deep Offshore Wind Exploitation)
Reporting period: 2019-10-01 to 2021-09-30
Over the past years, offshore wind energy has been considered as one of the most promising replacements for fossil fuels. The problem being addressed in this project is how to develop an effective wind turbine control and fault diagnosis system, and to achieve (1) operational output power quality improvement with operational control techniques, (2) load mitigation as well as the increase of the structural lifetime with load control schemes and (3) early fault detection, diagnosis, and accommodation with condition monitoring architecture, and thus solve the difficulties of reliability, availability, and optimal operation of the offshore wind turbine to fulfil the highly challenging ambition set by European Union (EU).
Solving this problem is important for society because the current offshore wind turbine is lacking an effective and robust control and fault diagnosis system, which may result in operation interruptions and cause disastrous economic losses. The successful implementation of this project will improve the effectiveness, robustness, integration, and multi-scalability of the actual state of the art to make offshore wind energy more cost-effective. It will open new career opportunities in the field of offshore wind power generation in academic and industrial circles in Europe, and contribute to maintaining European leadership in the offshore wind industry.
The overall objective is to develop effective, robust, integrated and multi-scale control and fault diagnosis systems for offshore wind turbines, which will finally achieve cost-effective deep offshore wind exploitation.
Solving this problem is important for society because the current offshore wind turbine is lacking an effective and robust control and fault diagnosis system, which may result in operation interruptions and cause disastrous economic losses. The successful implementation of this project will improve the effectiveness, robustness, integration, and multi-scalability of the actual state of the art to make offshore wind energy more cost-effective. It will open new career opportunities in the field of offshore wind power generation in academic and industrial circles in Europe, and contribute to maintaining European leadership in the offshore wind industry.
The overall objective is to develop effective, robust, integrated and multi-scale control and fault diagnosis systems for offshore wind turbines, which will finally achieve cost-effective deep offshore wind exploitation.
My work performed from the beginning of the project to the end of the period and its main results achieved so far are summarized in the following two aspects: (1) Advanced wind turbine control with the aid of wind speed estimators, (2) Fault diagnosis system. All will actively contribute to the main goal of this project: development of the effective and robust control and fault diagnosis system for offshore wind turbines.
In detail, I focus on the offshore wind turbine system dynamics and develop new control and fault diagnosis strategies for economical offshore wind energy exploitation. In particular, I developed a fully data-driven control, which is called subspace predictive repetitive control, for load reductions and fault-tolerant control on a floating offshore wind turbine. High-fidelity simulations show that the proposed learning-based subspace predictive repetitive control is very effective at load reduction and fault-tolerant control. As a complementary but connected solution, the fault diagnosis system was also investigated, which aims to increase the safety and reliability of the floating offshore wind turbine.
In addiction, I built up a collaboration with researchers from Politechnical University in Milan, Italy and researched on the dynamical modelling and control system design for the floating offshore wind turbines. As a result, a novel wave-excited linear model is derived to represent the fundamental dynamics of the floating offshore wind turbine. Based on the linear model, different control algorithms are demonstrated on a 10MW DTU floating offshore wind turbine via hybrid software-in-the-loop wave basin experiments. We also develop a novel feedforward control to reject the wave disturbance of the floating offshore wind turbine.
For effective wind turbine control, I developed three novel wind speed estimators. The first one is called subspace predictive repetitive estimator, which is inspired by the subspace predictive repetitive control method, to estimate the periodic wind state. Second, the global convergence of the regular torque balance estimator is proved based on the circle criterion in my paper, which opens up the inclusion of time-delayed systems and/or uncertainties.
In detail, I focus on the offshore wind turbine system dynamics and develop new control and fault diagnosis strategies for economical offshore wind energy exploitation. In particular, I developed a fully data-driven control, which is called subspace predictive repetitive control, for load reductions and fault-tolerant control on a floating offshore wind turbine. High-fidelity simulations show that the proposed learning-based subspace predictive repetitive control is very effective at load reduction and fault-tolerant control. As a complementary but connected solution, the fault diagnosis system was also investigated, which aims to increase the safety and reliability of the floating offshore wind turbine.
In addiction, I built up a collaboration with researchers from Politechnical University in Milan, Italy and researched on the dynamical modelling and control system design for the floating offshore wind turbines. As a result, a novel wave-excited linear model is derived to represent the fundamental dynamics of the floating offshore wind turbine. Based on the linear model, different control algorithms are demonstrated on a 10MW DTU floating offshore wind turbine via hybrid software-in-the-loop wave basin experiments. We also develop a novel feedforward control to reject the wave disturbance of the floating offshore wind turbine.
For effective wind turbine control, I developed three novel wind speed estimators. The first one is called subspace predictive repetitive estimator, which is inspired by the subspace predictive repetitive control method, to estimate the periodic wind state. Second, the global convergence of the regular torque balance estimator is proved based on the circle criterion in my paper, which opens up the inclusion of time-delayed systems and/or uncertainties.
The progress beyond the current state of the art lies in two aspects:
(1) The current theoretical and algorithmic research has difficulty in addressing all the challenges of the aforementioned effectiveness, robustness, integration and multi-scalability of control and monitoring systems for floating offshore wind turbines. In detail, all of the existing methods fail to develop an online system, because the computational complexity of these models is too high to allow for real-time dynamical simulations. Moreover, only a limited variety of methods are reported to model the fault diagnosis and optimal control system, especially for the floating offshore wind turbines currently in the literature, which might be attributed to the complicated coupling effects of the platform motions, generator speed, and blade pitch, all of which are highly dependent on the floating platform characteristics. The expected results derived from our project will fully address all the challenges of the effectiveness, robustness, integration and multi-scalability of the wind turbine control and fault diagnosis system for the floating offshore wind turbine. For instance, in this project, I develop a fully data-driven control algorithm called subspace predictive repetitive control for load control and fault accommodation. Then an adaptive control law is formulated to aim at the load reduction or fault accommodation. In addition, I build up a mixed model and signal-based fault diagnosis algorithm, which is able to take into account all the critical faults and failures in floating offshore wind turbines. All of these algorithms are demonstrated via high-fidelity simulations.
(2) Simulation and experimental research are available to evaluate the performance of the newly developed controller and monitoring system by establishing numerical simulations or experimental tests. However, existing literature on simulation and experimental research is still scarce. Most of the relevant work, to the best of the researcher’s knowledge, mainly considers the onshore wind turbine benchmark, with limited consideration of the FOWTs. In this project, the expected results will address the simulation and experimental research knowledge gap. First, we build up a new benchmark model for the floating offshore wind turbine, which opens up the inclusion of different kinds of critical faults and failures. Based on the benchmark model, the simulation research on the wind turbine control and fault diagnosis system can be implemented. Furthermore, we did hybrid software-in-the-loop wave basin experiments in MARIN in the Netherlands to validate the newly developed control system of the floating offshore wind turbine under several fault scenarios.
Based on the progress and the results derived from the EDOWE project, it is promising to apply the developed new control and condition monitoring techniques to the real floating offshore wind turbine in the near future. With an effective and optimal control and condition monitoring strategy developed in this project, the safety, reliability, robustness and cost-effectiveness of the wind turbine will be increased, which leads to an economical viable and more cost-effective floating wind energy exploitation, and thus a green future and zero-carbon emission. This is the socio-economic impact and the wider societal implications of this project.
(1) The current theoretical and algorithmic research has difficulty in addressing all the challenges of the aforementioned effectiveness, robustness, integration and multi-scalability of control and monitoring systems for floating offshore wind turbines. In detail, all of the existing methods fail to develop an online system, because the computational complexity of these models is too high to allow for real-time dynamical simulations. Moreover, only a limited variety of methods are reported to model the fault diagnosis and optimal control system, especially for the floating offshore wind turbines currently in the literature, which might be attributed to the complicated coupling effects of the platform motions, generator speed, and blade pitch, all of which are highly dependent on the floating platform characteristics. The expected results derived from our project will fully address all the challenges of the effectiveness, robustness, integration and multi-scalability of the wind turbine control and fault diagnosis system for the floating offshore wind turbine. For instance, in this project, I develop a fully data-driven control algorithm called subspace predictive repetitive control for load control and fault accommodation. Then an adaptive control law is formulated to aim at the load reduction or fault accommodation. In addition, I build up a mixed model and signal-based fault diagnosis algorithm, which is able to take into account all the critical faults and failures in floating offshore wind turbines. All of these algorithms are demonstrated via high-fidelity simulations.
(2) Simulation and experimental research are available to evaluate the performance of the newly developed controller and monitoring system by establishing numerical simulations or experimental tests. However, existing literature on simulation and experimental research is still scarce. Most of the relevant work, to the best of the researcher’s knowledge, mainly considers the onshore wind turbine benchmark, with limited consideration of the FOWTs. In this project, the expected results will address the simulation and experimental research knowledge gap. First, we build up a new benchmark model for the floating offshore wind turbine, which opens up the inclusion of different kinds of critical faults and failures. Based on the benchmark model, the simulation research on the wind turbine control and fault diagnosis system can be implemented. Furthermore, we did hybrid software-in-the-loop wave basin experiments in MARIN in the Netherlands to validate the newly developed control system of the floating offshore wind turbine under several fault scenarios.
Based on the progress and the results derived from the EDOWE project, it is promising to apply the developed new control and condition monitoring techniques to the real floating offshore wind turbine in the near future. With an effective and optimal control and condition monitoring strategy developed in this project, the safety, reliability, robustness and cost-effectiveness of the wind turbine will be increased, which leads to an economical viable and more cost-effective floating wind energy exploitation, and thus a green future and zero-carbon emission. This is the socio-economic impact and the wider societal implications of this project.