Wind energy conversion chain and MPPT in MATLAB/Simulink

Context

Individual first-year Master's lab in MESI (Renewable Energy module), supervised by Prof. Dominique Knittel. A turbine drives a DC generator that feeds a load. The task was to compare three control strategies to see which one extracts the most power.

What I did

• Direct connection, to recover the Cp(λ) curve and the maximum power point (about 7.5 kW).
• Boost chopper with a PI on the output voltage, switching at 10 kHz.
• Perturb & Observe MPPT, which finds the top of the power curve on its own.
• Static and dynamic study of each case.

What I noticed and learned

• The smarter the control, the more power is extracted, without touching the hardware.
• The Boost can only raise the voltage. A Buck-Boost would remove this limit.

My mistakes

• My first PI was too aggressive and oscillated, so I had to cut the gains sharply.
• I reused a photovoltaic P&O code. With a Boost on a wind turbine, the step signs had to be inverted.
• At very high power, the model diverges, because the components were not sized.

Results

• Cp ≈ 0.44 with the MPPT, close to the maximum of the turbine.
• About 2.4 times more power captured than regulation alone.
• Without regulation, a load step changed the voltage by a factor of 31.

Tech stack