Research Of Coordination Control Of Reactive Power And Voltage Of Wind Farms Connected To The Grid

Under the background of today’s deteriorating environment, growing energy shortages and increasing electricity consumption, wind resources are developed and utilized on a large scale. Seeing the wind power situation from domestic and foreign, the installed capacity of wind turbines shows increasing trend in recent years. However, a lot of challenges are brought to the grid because of the integration of wind power. One of the challenges is reactive power and voltage control. The fluctuation of the wind power that brings about the voltage fluctuation is the fundamental cause. Response feature of different time scale regulation devices in the wind farms. Fluctuation and randomness of the wind combined increase the difficulty of the reactive power & voltage control. The thesis makes a study of reactive power & voltage control of wind farms in the above background.Firstly, wind speed model and aerodynamic model was established. Three kinds of wind turbines are studied, including doubly fed induction generator (DFIG), fixed speed induction generator (FSIG) and permanent magnet synchronous generator (PMSG). According to which the relation of reactive power, active power terminal and voltage is deduced.equivalent circuit of FSIG is established; before reactive power limit is deduced, vector control of DFIG is studied. The influence to the grid voltage of three cases is studied, consisting of FSIG as the wind turbine type, DFIG as the wind turbine type of constant power factor and constant terminal voltage. Besides, the features of some reactive power and voltage regulation devices are studied:consisting of shunting capacitors, (OLTC), SVC, STATCOM and on load tap changer.Secondly, Multi objective optimization problem and Multi Objective Particle Swarm Optimizer were studied. So, we proposed a multi objective PSO based on adaptive grid and fan (G-F MOPSO). The algorithm principle and algorithm flow is given. The propose algorithem has the following characteristic:adaptive crowing grid and fan was used to define the diversity of particle in the external archive to keep the size of the external archive, Learning samples was assigned by the information of density and crowing. And updating formula was improved. Finally, G-F MOPSO was tested by multi-objective test functions. The algorithm shows good performance in comparison with ACG MOPSO.Finally, According to the basis of study, the reactive power and voltage control mode of wind farms integrating into the grid is constructed. For the wind farm level control, a control method of two-tier and multi-stage of wind farm cluster was proposed. The model considered the response features of different time scale devices and the uncertainty of the wind power. G-F MOPSO was selected to solve the model. After the solution, analytic hierarchy process was used to select a best solution. At last, the proposed model was tested by a typical topology of multiple wind farms integrating to the grid, which can effectively restrain the voltage fluctuation caused by the wind speed fluctuation and reduce the switching times of discrete devices, achieving the economic objective.