Research On Power Optimization Problem Of Megawatt Direct-drive Permanent Magnet Wind Turbine

With the proportion of wind power in grid system increasing, it puts forward higherrequirements on wind power which is that the wind power can provide enough energy tosupport grid system in dynamic-state and the output power is smooth enough insteady-state. Therefore, output power control of wind power system seems more and moreimportant. The optimization control of power has become a hot spot of large wind turbineresearch. The direct-drive permanent magnet wind turbine is taken as the research object inthis dissertation based on the project supported by National Natural Science Foundation ofconsidering nonlinear dynamic load uncertainties of wind turbines control strategy research(project number:51207095) to further research the control strategy of wind turbine poweroutput. The goal of the research work is to improve energy conversion efficiency under therated wind speed, to stabilize output power and inhibit power fluctuation above the ratedwind speed. The main content of the dissertation is as following:First of all, the direct-drive permanent magnet wind turbine power optimization underrated wind speed was studied. Based on the theory of optimal tip speed ratio wind energyconversion, the paper analyzed optimization problems of the direct-drive permanentmagnet wind turbine under the rated wind speed. Following the analysis, robust speedcontroller of the L2norm optimization was designed which applied the theory frameworkof dissipative system which adopted the dynamic error equation to build system model anddisturbance restraint and tracking requirements for wind speed are summed up L2designproblems. In order to solve the control law, the Lyapunov function is applied to constructthe storage function and requirements of performance index for system are changed intodissipative stability problem and the virtual function back stepping method is adopted tosolve the controller. In order to overcome the bad influence of wind speed fluctuation tothe controller, the paper used the perturbation theory and dynamic observations, andpneumatic torque changes is accurately feed forward compensated, which greatly improvesthe anti-interference ability of the system and realizes the maximum wind energy captureof wind turbines. Meanwhile，feasibility of the design method is verified by virtue ofcomputer simulation．Secondly, the direct-drive permanent magnet wind turbine power optimization aboverated wind speed was researched. Based on the rotor aerodynamics and linear controltheory, this thesis deeply analyzed constant power control problem of direct-drivepermanent magnet wind turbine above the rated wind speed. Then, On the basis of problemanalysis, nonlinear predictive control strategy is applied to pitch control to solve the bigrotor moment of inertia time delay and the wind speed disturbance issues of feed-forward compensation. In order to further improve the prediction model for adaptive parametervariable, this thesis applied model free adaptive algorithm to optimize predictive modelparameter, which improve accuracy and robustness of predictive model, and greatly reducethe workload of the design of the controller parameters. Further, the thesis Derive controllaw of nonlinear predictive control strategy based on Generic model adaptive algorithm.Meanwhile, the simulation experiment verifies the correctness of the control strategy.Thirdly, the direct-drive permanent magnet wind turbine power fluctuation aboverated wind speed was researched. Based on aerodynamics and pitch control theory, thethesis deeply analyzes power fluctuation problem of direct-drive permanent magnet windturbine above the rated wind speed. On the basis of the analysis, this thesis appliesmodelfree model theory and rotor aerodynamics to study wind load calculation model andthe power fluctuation change trend and deduce the power fluctuation trend model anddesign a special individual pitch control strategy. It decouple into unified pitch control anddeviation pitch control. On the one hand, unified pitch control adopts nonlinear predictivecontrol based on modelfree model of the fourth chapter, which maintain wind turbineoutput power rating unchanged; on the other hand, deviation pitch control adopts the bladeroot edgewise load changed into pitch Angle feedback control strategy based on modelfreemodel. The control strategy through the real-time detection of each blade root edgewiseload, the deviation of measured value with the given value to reflect rotor unbalancededgewise load caused by wind speed change and dynamic model is established by thetheory of modelfree and the blade root edgewise load deviation value transformation intothe pitch Angle feedback adjustment volume and single blade is individually controlled.Control’s main goal is to real-time tracking response to the change of the wind speed,through pitch Angle feedback adjustment, reduce unbalanced edgewise load caused bywind shear and tower shadow effects, thus inhibiting output power fluctuation of windturbine.At last, Experiment research of wind turbine power optimization was done. In order toverify the above power optimization control strategy, the dissertation construct simulationexperiment platform based on hardware in the loop, for the above control methods ofpower optimization, do the computer simulation and experiment based on the platform ofthe hardware-in-the-loop. On this basis, combining with typical the wind of wind farmssimulation, it further verify the application of the proposed control algorithm and strategyin the smoothing output power and improving power quality and the result is feasible andeffective.