Research On On-line Parameter Identification Of Direct-driven Permanent Magnet Synchronous Generators For Wind Energy Conversion

The control of permanent magnetic synchronous generator depends on accrurateparameters of generator. The PM flux linkage could fluctuate and even demagnetize in theprocess of running generator. Error of position feedback signal could also lead to inaccuraterotor flux-oriented. On-line parameter identification has significant implications for generatorcontrol. Direct-driven PMSG is choosen as the object of thesis to discuss online parameteridentification methods.Mathematical model of Direct-driven PMSG is derived. Morecver, Maximum powerpoint tracking (MPPT) control scheme based on optimal electromagnetic torque reference isadopted. Additionally, the closed-loop current control system is designed based on rotor fluxoriented control strategy. Simulation model of vector control system for WECS with positionand speed sensor is built up in Matlab/Simulink module. Simulation results show MPPTcontrol could be realized effectively.The permanent magnet flux linkage could fluctuate and even demagnetize in the processof running generator. Error of position feedback signal could also lead to inaccurate rotorflux-oriented. With position sensor, simulation identification of parameter amplitude suddendrop under accrurate and faulty rotor flux-oriented is completed with EKF algorithm.Simulation results indicate that PM flux linkage could be precisely estimated by using thismethod. Meantime, system has good steady and dynamic performance. The combination ofPM flux linkage parameter and rotor position state identification has been respectivelycompleted based on EKF and STF algorithm. Simulation results indicate that the estimatedrotor position and speed based STF algorithm could be more precise and have steady values.Besides, estimated flux linkage has faster convergent speed and less steady errors.Simulation model of PMSG based on Ansoft is built up to simulate partial demagnetization faults, and gap flux density and back EMF are observed. Simulation spectralanalysis results show, when it is running in the state of no-load or the resistance andinductance load, certain harmonic signal which has specific frequency is found in parameterslike back EMF and stator flux linkage. Then, feactures of demagnetization faults frequencyare obtained. Generator stator flux linkage and rotor flux linkage are estimated based on STFalgorithm, which could provide references for detection research of demagnetization faults inpermanent magnetic synchronous generator.The design of two types of converter hardware system, uncontrolled rectifier andfull-controlled rectifier, have been proposed. What’s more, DSP program of EKFidentification algorithm for flux linkage with position sensor is compiled. Lastly,experimental results show the proposed EKF identification algorithm can accurately estimatePM flux linkage. Meantime, MPPT control scheme is effective with estimated flux linkageparameter.