Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m *...
Transcript of Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m *...
![Page 1: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/1.jpg)
Evaluating energy maximising control
systems for WECs using CFD
Josh Davidson
HyWEC Workshop, Bilbao,
6th April 2017
![Page 2: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/2.jpg)
Outline
• Motivation
• CFD simulation of a controlled WEC
• Illustrative example
• Adaptive control
– System Identification
– Evaluation of adaptive WEC control using CFD
– Example
• Discussion and Conclusions
![Page 3: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/3.jpg)
Motivation
* Weber, Costello and Ringwood, WEC Technology Performance Levels (TPLs) – Metric for
Successful Development of Economic WEC Technology, Proc. 10th EWTEC, Aalborg, 2013
![Page 4: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/4.jpg)
Motivation
* Adapted from: Weber, Costello and Ringwood, WEC Technology Performance Levels (TPLs) – Metric
for Successful Development of Economic WEC Technology, Proc. 10th EWTEC, Aalborg, 2013
![Page 5: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/5.jpg)
Motivation
* Adapted from: Weber, Costello and Ringwood, WEC Technology Performance Levels (TPLs) – Metric
for Successful Development of Economic WEC Technology, Proc. 10th EWTEC, Aalborg, 2013
![Page 6: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/6.jpg)
Motivation
* Adapted from: Weber, Costello and Ringwood, WEC Technology Performance Levels (TPLs) – Metric
for Successful Development of Economic WEC Technology, Proc. 10th EWTEC, Aalborg, 2013
![Page 7: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/7.jpg)
Motivation
* Davidson et al, ch. Evaluation of energy maximising control systems for WECs using OpenFOAM,
11th OpenFOAM Workshop, Springer, 2017
![Page 8: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/8.jpg)
* Yu and Li, Reynolds-Averaged Navier-Stokes simulation of the heave performance of a
two-body floating point-absorber wave energy system, Computers and Fluids, 2013
Motivation
![Page 9: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/9.jpg)
Computational Fluid Dynamics
* Davidson et al , Implementation of an OpenFOAM Numerical Wave Tank
for Wave Energy Experiments, Proc. 11th EWTEC, Nantes, 2015
![Page 10: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/10.jpg)
Wave driven motion
![Page 11: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/11.jpg)
Free decay
* Davidson, Giorgi and Ringwood, Identification of Wave Energy Device Models From Numerical Wave Tank
Data – Part 1: Numerical Wave Tank Identification Tests, IEEE Transactions on Sustainable Energy, 2016
![Page 12: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/12.jpg)
Free decay – Reflection analysis
![Page 13: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/13.jpg)
CFD + PTO = WEC
PTO force on
body
![Page 14: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/14.jpg)
Input PTO force
![Page 15: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/15.jpg)
WEC Operation
![Page 16: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/16.jpg)
CFD + PTO + Controller = Today’s topic
MATLAB
script
Free surface
elevation
PTO force on
body
![Page 17: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/17.jpg)
Example
- No Control
- PI Control
Buoy
PTO - Buoy diameter = 0.1m
- Buoy resonance = 0.61s
- Sea spectrum peak = 1s
PTO strategies
- Linear model (Cummins Equation)
- OpenFOAM (CFD)
Simulations
![Page 18: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/18.jpg)
Example - Tank
![Page 19: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/19.jpg)
Example : Results
![Page 20: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/20.jpg)
Example : Operational space
![Page 21: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/21.jpg)
Example - Relative disp. and vel.
![Page 22: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/22.jpg)
Example - Conclusions
- While linear hydrodynamics works well for a surface following
floating body, the large amplitude motions of a WEC under
controlled conditions challenge the validity of linear models
- Control evaluation requires a modelling approach that includes
all system dynamics, nonlinearities, inefficiencies and energy
dissipating mechanisms
- A CFD based NWT offers a good possibility for control
evaluation
![Page 23: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/23.jpg)
CFD + PTO + Adaptive control =
Current work
• Adaptive control parameters
• Adaptive control models
– Model based control
• CFD provides a realistic simulation model different from the control model
– System Identification
• Tune the parameters of the control model, based on measured WEC responses
• Control model is then representative of controlled conditions, present sea state and other time varying influences
![Page 24: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/24.jpg)
System Identification
![Page 25: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/25.jpg)
2n+1
PARAMETERS
Example - Parametric model structure
State space representation of Cummins Equation
![Page 26: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/26.jpg)
Example - Identification algorithm
![Page 27: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/27.jpg)
Example - Data : BEM vs CFD
![Page 28: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/28.jpg)
Example - Results
* Davidson, Giorgi and Ringwood, Linear parametric hydrodynamic models for ocean wave energy
converters identified from numerical wave tank experiments, Ocean Engineering, 2015
![Page 29: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/29.jpg)
Example - Results
![Page 30: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/30.jpg)
Example - Results
![Page 31: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/31.jpg)
Example - Results
![Page 32: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/32.jpg)
Example - Results
![Page 33: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/33.jpg)
Example - Results
![Page 34: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/34.jpg)
Example - Results
![Page 35: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/35.jpg)
Example - Conclusions
- Local linearisation- Nonlinear resistive energy dissipation captured into
the linear radiation term
- Amplitude dependent
- Possibility for model scheduling depending on
operating conditions
- Optimisation problem for the given model
structure was non-convex, required genetic
algorithm to solve- Leads to investigation of discrete time models
![Page 36: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/36.jpg)
Discrete time models
• Ringwood, Davidson and Giorgi, ch. Identifying models using recorded data,
Numerical Modelling of WECs : State-of-the-art for single devices and arrays, Elsevier, 2016
![Page 37: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/37.jpg)
Discrete time models
ARX
ARX - Polynomial (KGP)
![Page 38: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/38.jpg)
Discrete time models
ANN
![Page 39: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/39.jpg)
Discrete time models - ExampleTraining
Validation
![Page 40: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/40.jpg)
Identifying subcomponents
Example – Heaving cone
0.2m
0.8m
* Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time
hydrodynamic models , 1st Int. Conference on Renewable Energies Offshore (RENEW 2014), 2014
![Page 41: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/41.jpg)
Example – Nonlinear static block
![Page 42: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/42.jpg)
Example – Nonlinear static block
![Page 43: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/43.jpg)
Example – Small amplitude results
![Page 44: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/44.jpg)
Example – Large amplitude results
![Page 45: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/45.jpg)
Example – Large amplitude results
![Page 46: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/46.jpg)
Discrete time models - Conclusions
- Discrete time models better suit the inherent
discrete nature of sampled data- Least squares parameter identification
- Disadvantage in that they are black-box- Parameters have no direct physical interpretation
- However, can be given a ‘shade of grey’ by identifying
subcomponents that have physical meaning
- Range of linear and nonlinear model structures
available
- CFD offers a range of testing and measurement
possibilities for obtaining useful data
![Page 47: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/47.jpg)
Adaptive Control
* Davidson, Genest and Ringwood , Adaptive control of a wave energy converter
simulated in a numerical wave tank, Proc. 12th EWTEC, Cork, 2017
![Page 48: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/48.jpg)
Optimal control (1/2)
• Based on a receding horizon PS control algorithm
• The state and control variables are approximated by their truncated series
on a given set of orthogonal functions on a fixed control horizon I=[t,t+TH]
• The basis function chosen for the optimal control are the half-range
Chebyshev Fourier functions (see next slide)
• The performance function maximised by the control algorithm
corresponds to the absorbed energy
`
• Since all the basis function are orthogonal, the cost function is directly
proportional to:
![Page 49: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/49.jpg)
HRCF functions
![Page 50: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/50.jpg)
Optimal control (2/2)
• Whilst maximising the cost function, J, the control algorithm needs to
ensure that the dynamical equations describing the system are satisfied :
• Expressed in terms of residuals, and replacing each state and control
variable by the truncated series:
Where, D is the differentiation matrix, R is the radiation matrix
corresponding to the radiation force generated by the velocity over the
control horizon and Fr(t) is the radiation force generated by past velocities
![Page 51: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/51.jpg)
Model initialization- Choosing 7 basis functions in the RHPS control model leads to a 15x30 matrix of parameters
M
N
![Page 52: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/52.jpg)
Adaptive control - Results
![Page 53: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/53.jpg)
Adaptive control - Parameter adaption
![Page 54: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/54.jpg)
Adaptive control - Parameter adaption
![Page 55: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/55.jpg)
Adaptive control - Backstepping
![Page 56: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/56.jpg)
Adaptive control - Results
![Page 57: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/57.jpg)
Future work
- Implement PTO and mooring models in NWT
- Extend PTO control evaluation, to holistic control evaluation
- Adaptable control of physical properties of the WEC
- Geometry. Inertia, submergence etc
- Implement controllers outside of MATLAB
- … make comparison between CFD and fully nonlinear potential
flow NWT for WEC under controlled conditions
![Page 58: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/58.jpg)
Overall Conclusions
- Numerical Wave Tanks are a valuable tool for WEC analysis and
development
- Modern computational power, makes CFD feasible for some, but
perhaps not all, applications (yet?)
- SPH could also prove useful, particularly for cases where CFD has
difficulties in handling large amplitude mesh motion
- Identifying nonlinear parametric models from CFD experiments,
allows long duration simulation to be run quickly
- Evaluating model based controllers should be performed with a
simulation model different from the control model
![Page 59: Evaluating energy maximising control systems for WECs ......Example –Heaving cone 0.2m 0.8m * Davidson, Giorgi and Ringwood, Numerical wave tank identification of nonlinear discrete-time](https://reader035.fdocuments.in/reader035/viewer/2022071416/6112d420a01c9d506a02ef24/html5/thumbnails/59.jpg)
Discussion / questions???
Acknowledgement: This presentation is based upon work supported
by Science Foundation Ireland under Grant No. 13/IA/1886.