Research On The Low-Voltage-Ride-Through Control Of Doubly Fed Induction Generator System Under Serious Grid Voltage Sag

In China, many wind farms are located in the place with poor conditions where the grids are comparatively weak and may drop sharply. The large-scale disconnecting of wind turbines will seriously threaten the operation of power system. Thus, the research on the LVRT of the wind turbine has important theoretical and practical meanings.The main control technology of DFIG under grid faults is the vector control. The key of the vector control is the stator flux estimation. Under normal grid conditions, the voltage-model-based stator flux estimation method is used widely. Although the drift problem of the integrator in the voltage-model-based flux observer can be solved by using filters instead of pure integral, it introduced the amplitude and phase errors. The synchronous angular frequency is often used in order to eliminate these errors. However, the traditional synchronous angular frequency estimation methods are with bad performance. Thus, this dissertation researched new synchronous angular frequency estimation methods.Under grid-fault conditions, there will be dc component(and negative sequence component for an asymmetrical grid fault) in stator flux, which will influence the performance of the traditional stator flux estimation method and even the LVRT control strategy. In order to get the accurate decoupling of the DFIG system, this dissertation also researched the new stator flux estimation method.Under grid-fault conditions, the negative and transient dc components in stator flux can cause rotor overcurrent. In order to avoid this problem and make sure of the safe and stable operation of the DFIG, it is essential to suppress the negative and transient dc components in stator flux. In this case, the negative and transient dc components of the stator flux need to be extracted. Therefore, this dissertation further researched the new method for extracting the components of the stator flux.The primary contents of this dissertation are as follows:A new synchronous angular frequency estimation method based on second order generalized integrator-frequency locked loop(SOGI-FLL) was proposed. The speed of the dynamical response of SOGI-FLL was theoretically analyzed. An induction motor speed control system was built based on the vector control and controlled by TMS320F2812 DSP. Simulation and experimental results proved the effectiveness of the proposed method.A new synchronous angular frequency estimation method based on dual second order generalized integrator-frequency locked loop(DSOGI-FLL) was proposed. According to the simulation and theoretical analysis, this new method can avoid the influence of the change in stator back-EMF and eliminate the AC fluctuant component. The synchronous angular frequency estimated through the new method can improve the accuracy of the conventional flux observer.A new flux estimation method based on Dual Second-Order Generalized Integrator Frequency-Locked Loop(DSOGI-FLL) was proposed. Compared with the conventional low-pass-filter-based method, the DSOGI-FLL does not need compensation and can effectively attenuate the high-order harmonics in the back Electro Motive Force(EMF). The properties and performance of the DSOGI-FLL-based flux estimation method are analyzed, demonstrating that it has excellent static and dynamic response.A new method for extracting the positive and negative components of the stator flux based on DSOGI-FLL was proposed. The principle and the dynamical response of proposed method was theoretically analyzed. Simulations of the DFIG system performed in SIMULINK prove the efficiency of the proposed algorithm.