Abstract

The vast potential of extracting wave energy from the ocean has significant challenges that still need to be overcome. Many different types of devices are being developed worldwide and all of them have advantages as well as drawbacks, so that none of them has proven to be the best solution in terms of cost-effectiveness, ease in deployment, visual and environmental impact. Apart from the design, a huge impact on the power output (whatever the type) is the control action used. The main requirements for this are: real-time execution, insensitiveness to model inaccuracies and obviously better performance than just a linear damping.

One of the key motivation in the present research lays in here: to find control strategies that perform well, and that can be possibly generalized for a range of wave energy converters. Swirl Generators Limited (SGL) is the Industry alongside this project and they are currently deploying a wave energy device called WaveRAM, that consists of a floating Oscillating Water Column (OWC) that “behaves” as a point absorber: due to some ballast tanks on the bottom it can also be tuned for different weather conditions.

A consistent part of the research for this kind of controls has been based on linear models as well as linear waves (using the linear wave theory). But characterizing thoroughly any kind of wave energy device must take nonlinearities into account, as no linear approximation can be done in high seas and severe/extreme conditions, to be accurate enough.

A considerable part of this research will deal with it and will finally test the WaveRAM device in the so called “Real Time Hybrid Testing” as well. Once the control algorithm is written, thanks to the established facility in Trinity College, Dublin (already in use for several years by now, and that has already allowed a WRAM emulator to be built) thanks to a bidirectional air flow from the air chamber of the device that is directed through a controllable bidirectional turbine, it can be tested and the results compared to numerical simulations.

So, an expected outcome for this research will be to contribute to the scientific community with the focus on control strategies, nonlinear behaviours and the optimization of the WaveRAM device, by SGL.

 

 

The scientific content of the PhD project

  1. Background

Amongst the more widely researched solutions on Wave Energy Converters (WECs), Oscillating Water Column has attracted considerable attention as these make one of the most effective means of absorbing wave energy. SGL has developed a single body device, consisting of a surface float that is rigidly linked to a substantial mass of trapped sea water. Operation of this kind of device is quite different though, as it is the heaving buoy responding to the wave excitation, and not the water column. To maximize the power output, innovative control algorithms were already being formulated for the device, but much more needs to be investigated. The use of Real Hybrid Testing is a relatively fast and cost-efficient way to prove the effectiveness of these algorithms and their features. Two types of control algorithms are proposed, and they will be tested on a base of a day-by-day week-by-week change of the reference mass, based on weather conditions. This will lead in including one of the most important part of the research that is the tuning the device for severe/extreme weather conditions, building a robust pneumatic air pressure non-linear model controller.

 

  1. Introduction and State of Art

The first type of controller it has been decided to deal with is a kind of Model Predictive Control, since it has achieved considerable success in the process industries, thanks to its ability to deal with linear and nonlinear models, while observing the constraints and future behaviour. The WEC energy maximization however, requires significant modification of the traditional MPC objective function. So, in terms of this research refer to Section 9 for the state of the art. Applying this type control to the WaveRAM device will be what is a consistent part of the research. After that it will be including then all the nonlinearities and tuning it for different weather conditions. Eventually, the Real Hybrid Testing of the algorithm will be processed as well.

A second, and maybe third type of controller will be fully studied as well with all the sub steps cited later. A second type of control algorithm is a linear time-varying system with a forward Riccati formulation in wavelet domain.  Many others are still under investigation and the literature review will be part of the research as well.

 

  1. Project Objectives

The project objectives will be to: develop and compare control algorithms for the WaveRAM device, tuneable in different weather conditions, especially high seas and extreme states, including nonlinearities of the model and testing these algorithms using the Real Time Hybrid Testing rig available and ready to use in Trinity College, Dublin, with the WRAM emulator and the bidirectional controllable air turbine. This will be pursued in distinct stages, focusing first on one type of controller only, and adding constraints and more and more details along the way.

 

  1. Key methods and substudies

The whole process will be pursued in several sub-stages. After a literature review of the control strategy to be looked at, its strengths and weaknesses, issues to be addressed, some are chosen and applied to the model of the device. To initially simplify the analysis, the model studied belongs to a general point absorber, and its data are taken from previous researches and publications. The chosen algorithms are then coded, tested and together evaluated. The ones that are then believed to be better performing for the WaveRAM device are selected, and they are applied to the WaveRAM model with the same procedure. With the progression of the analysis focus on different areas of the control is expected (wave prediction, application of the control force). Eventually the control algorithm is tested on the rig, and tuned. This is a key part of the research. Similar step by step procedure will be applied for other controls.

 

  1. Expectations

9 months: WaveRAM model is expected to be used and several Model Predictive Control configurations reviewed. A deeper study on wave prediction with the use of historical data and Kalman filtering is expected to be done. Non-linearities are attempted to be included at least partially in the model created. Results on the WaveRAM model are expected. Submission of the conference paper to ICOE 2018 (Normandy) and OTO’18 Oceans is expected on the 30th of November 2017 and 1st of December. A Journal paper will be submitted in January.

 

  1. Time schedule of the project

9 months: (see Figure 1)

 

 

Collaboration Agreements

 

The cooperation between the student and the supervisor tries to be on a weekly basis availability of the both parties permitting. The student summarizes his work through weekly reports. Depending on availability the supervisor agrees in giving the student guidance, explanations or sort of clarifications via e-mail, telephone or Skype on a regular basis. The student must show initiative and producing new ideas.

Plan for PhD Courses

 

CE7J06 – Wave and Hydro Energy: Trinity College Dublin , 5 ETCS, To Be Attended

ME5B09 – Control Engineering II: Trinity College Dublin , 5 ETCS, To Be Attended

 

Plan for fulfilment of knowledge dissemination

  • Conference Paper @ ICOE (International Conference of Ocean Energy) June 2018, Normandy OR
  • Conference Paper @ OTO’18 Oceans Japan, 28-31 May, Kobe
  • Poster presentation @ CORE 2018 International Conference, 26th-28th August, Glasgow
  • Conference in 2019 (?)

 

External Collaboration

 

As required by the ICONN program, 3 months in Aalborg University (or elsewhere), and 15 months in SWIRL GENERATORS LIMITED (SGL) also located in Trinity College Dublin.

 

List of References

[1] Ted K.A.. Brekken. On Model Predictive Control for a Point Absorber Wave Energy Converter. Member IEEE, 2011.

[2] M. Patel. Dynamics of Off-shore structures. Butterworth-Heinemann, 1989.

[3] Z.Yu, J.Falnes. State-space modelling of a vertical cylinder in heave Applied Ocean Research, vol. 17, pp.265-275, Oct.1995

[4] E.Rusch. Catching a wave, powering an electrical grid? Smithsonian Magazine, July 2009

[5] J. Falnes. Ocean waves and oscillating systems Cambridge University Press, 2002

[6] G. Li, M. R. Belmont. Model Predictive Control of sea wave energy converters – Part I: A convex approach for the case of a single device Renewable Energy, Volume 69, September 2014, Pages 453-463.

[7] C.Signorelli, B. Basu. Control Algorithm Development for WRAM using a Real-Time Hybrid Test Platform 12th EWTEC, Cork, Ireland

[8] U.A.Korde, J.V. Ringwood. Hydrodynamic Control of Wave Energy Devices, 2016, Cambridge University Press.