Research On Composite Control For Direct-Drive Wind Energy Conversion System Based On Observer Technology With Model Prediction Method

As a renewable natural non-polluting resource,wind energy has large potential of development.It can replace the increasingly rare fossil fuel to solve the problem of energy shortage in today's world.With the development of the economy,people are demanding higher and higher requirements for wind energy conversion system,but permanent magnet synchronous generator is a gearless system,which can work at low speed without brush,and can operate reliably with low rotor loss.It has great research significance to apply permanent magnet synchronous generators to wind power systems.Control technology is one of core technologies for permanent magnet synchronous wind energy conversion system.Due to nonlinearity,strong coupling and multi-variabilities of the wind energy conversion system,coupled with the randomness,time-varying and uncontrollable characteristics of wind energy,it is difficult to obtain the system model precisely.In addition,because the harsh operating environment of the system,various disturbances caused by environmental factors also make it difficult to control the wind energy conversion system.In order to solve the problem of disturbance and improve the efficiency of the wind energy conversion system.The composite model predictive control strategy based on the extended state observer is used in this paper.The Model Predictive Controller(MPC)is employed as the main controller and the Extended State Observer(ESO)is used to help MPC control the system.The Buck circuit is used as the control object to verify.Under the premise of demonstrating the effectiveness of the strategy,the composite model predictive strategy is applied to the wind energy conversion system to improve ability of the anti-disturbance and enhance the stability and robustness,which achieve the purpose of capturing the maximum wind energy.In the case of large system inertia,fast parameter perturbation,and large external disturbances,ESO may have large deviations,which may cause the overall control performance of the system to decline.Aiming at solving this problem,the design of model compensation controller based on inertia identification observer is proposed.The total disturbance of the system can be partially predicted based on some known model information,and the model compensation control scheme based on inertia identification is designed to further expand the control bandwidth,reducing the workload of ESO,avoiding the integral saturation of the current loop PI controller.The simulation results show that the model compensation controller based on inertia identification can effectively reduce the ESO disturbance estimation burden and increase the response speed while reducing the unnecessary control energy loss.In order to realize the real-time regulation of the permanent magnet synchronous generator speed,it is necessary to obtain the rotational speed of the permanent magnet synchronous generator as a feedback signal for the speed loop control.Compared with the generator position sensor,the detection accuracy of the speed sensor is much more expensive.Aiming at solving this problem,this paper introduces speed sensorless technology and designs speed observer based on flux observer,and combining with composite model predictive method.The simulation results show that the speed observer based on the flux observer can effectively track the real-time speed of the generator.