MPC Design for Wind Turbine
Budget: £250 – £750 GBP
I need a control-systems specialist to build a Model Predictive Controller for a wind turbine that I have already represented as a Linear Parameter-Varying (LPV) model in MATLAB/Simulink. The controller must juggle three objectives at once—maximising electrical power output, keeping mechanical stress to a minimum, and sharpening pitch-angle accuracy—while remaining robust to real-world disturbances such as wind-speed fluctuations, grid faults, and the occasional mechanical component hiccup.
The core work happens in MATLAB/Simulink: setting up the prediction model, formulating the multi-objective cost function, and implementing hard or soft constraints that reflect turbine limits and grid-code requirements. I will provide the LPV model and any necessary operating data; you design, tune, and validate the MPC.
Deliverables
• Fully annotated Simulink/Stateflow implementation of the MPC, ready to drop into my existing LPV plant model
• MATLAB scripts or functions for prediction-horizon selection, constraint handling, and gain scheduling (if used)
• A brief report or notebook explaining the objective weighting, disturbance-rejection strategy, and key plots (power output, rotor speed, pitch angle, constraint trajectories) under representative wind profiles, grid events, and simulated component failures
• Guidance on integrating the controller into real-time targets (e.g., Simulink Real-Time) so I can move swiftly to hardware-in-the-loop tests
Acceptance criteria
1. Closed-loop simulations show at least a 5 % improvement in average power capture versus my current baseline control while keeping tower-top bending moments below the existing peak values.
2. Under IEC-class extreme gusts, pitch activity stays within actuator limits with no constraint violations.
3. The controller recovers stable operation within two seconds of a simulated grid fault or component outage.
If you have a proven track record with LPV-based MPC in MATLAB/Simulink and can demonstrate comparable projects, let’s talk.
The core work happens in MATLAB/Simulink: setting up the prediction model, formulating the multi-objective cost function, and implementing hard or soft constraints that reflect turbine limits and grid-code requirements. I will provide the LPV model and any necessary operating data; you design, tune, and validate the MPC.
Deliverables
• Fully annotated Simulink/Stateflow implementation of the MPC, ready to drop into my existing LPV plant model
• MATLAB scripts or functions for prediction-horizon selection, constraint handling, and gain scheduling (if used)
• A brief report or notebook explaining the objective weighting, disturbance-rejection strategy, and key plots (power output, rotor speed, pitch angle, constraint trajectories) under representative wind profiles, grid events, and simulated component failures
• Guidance on integrating the controller into real-time targets (e.g., Simulink Real-Time) so I can move swiftly to hardware-in-the-loop tests
Acceptance criteria
1. Closed-loop simulations show at least a 5 % improvement in average power capture versus my current baseline control while keeping tower-top bending moments below the existing peak values.
2. Under IEC-class extreme gusts, pitch activity stays within actuator limits with no constraint violations.
3. The controller recovers stable operation within two seconds of a simulated grid fault or component outage.
If you have a proven track record with LPV-based MPC in MATLAB/Simulink and can demonstrate comparable projects, let’s talk.