Novel sensorless fault-tolerant pitch control of a horizontal axis wind turbine with a new hybrid approach for effective wind velocity estimation

In this research, the fault-tolerant pitch angle control of a horizontal axis wind turbine in region 3 (where the wind velocity is greater than rated wind speed) is investigated. The effective wind velocity (EWV) is one of the necessary information for each control system. Wind speed is measured by the anemometers on the top of the nacelle however, the measurement is not precise and is only applicable for one point in the rotor. To address this issue, we have developed a novel hybrid approach. The approach is based on a sliding mode observer to estimate the aerodynamic torque and an adaptive neuro-fuzzy inference system (ANFIS) is introduced for obtaining the EWV. The estimated aerodynamic torque, pitch angle, and rotor speed are the inputs to this system and EWV is the output. New super twisting sliding mode control is introduced to overcome the faulty actuator case and control optimally the output power. Pitch sensitivity is further necessary information that is determined by another ANFIS system. The estimated EWV, pitch angle, and rotor speed are inputs of this system. Thus, the only information that is needed for this control approach is the pitch angle and rotor speed. The full aeroelastic simulations demonstrate excellent performance in comparison to the gain schedule PI control approach in tracking the output power and can reduce both the fore-aft vibration of the tower and the flapwise vibration of the blade significantly.