Research On Nonlinear Adaptive Control Of Wind Turbine

The natural environment on which mankind lives is being destroyed by excessive use of coal,oil and other resources.A new sustainable use of energy — wind energy,more and more people's attention.The use of wind energy can effectively alleviate the consumption of non-renewable resources and will not produce harmful gases,in line with the requirements of sustainable development.The turbulence system of the wind turbine is complicated and the random interference caused by wind shear and tower shadow effect makes the system show non-linear characteristics.In the face of complex and variable system,the traditional PD control can not achieve fast optimal Real-time parameter tuning.Therefore,the design of the fan system parameters unknown or subject to external interference control system can adapt to the object and disturbance dynamic characteristics of adaptive robust controller is very important.Aiming at the problem of poor stability in the case of wind turbines under the uncertainties of external factors,the problem of stable tracking of fan power output is studied and verified by adaptive robust control method.Firstly,the existing wind turbines are analyzed and their operating mechanism is expounded in detail.The nonlinear mathematical model with uncertainties is established for each system.The self-adaptive self-adjusting method is used to design the self-Adaptation law and controller.Experiments show that the adaptive controller can make the system quickly track the desired curve and be more accurate than the PD control when the fan system changes or is affected by external disturbance.Aiming at the problem that the aerodynamic torque in the adaptive controller is known,it is difficult to obtain the accurate representation in practical application.A neural network adaptive control algorithm is proposed.The algorithm can make the system no longer rely on structural parameters,quickly track the speed,get the optimal power curve.Simulation results show that the algorithm can meet the requirements.