Research On Wind Power Forecasting And Control Of Dispersed Wind Farm

Comparingwith the traditional centralized electricity generation, distributed grid of wind power is a new mode with characteristics of multi-point interconnection, local consumption and scattered layout. At present, centralized wind power generation encounters the wind curtailment problem. Hence, it has a significant strategy meaning for developing distributed wind power. However the properties of intermittent and bi-directional current challenge the economic efficiency and stability of distributing network(DN). According to different characteristics of dispersed wind farm(DWF), this dissertation focus on the following four key issues: single wind turbine power forecast under multi-timescale situation, economic operation of distribution network, stability scheduling for dispersed wind power farm, and multi-object voltage fluctuation mitigation based on wind turbine for improving the economy and stability of the grid based on DWF grid-connected operation mode.The feature of time-varying and intermittent restricts energy balance between wind power and load in real-time. Accurate wind power forecasting makes wind power to be a schedulable energy. However, traditional forecast methods cannot completely meet the requirements of DWF. As the uncertainties of wind power impact on power system, a multi-scenario power forecasting method based on optimal fitting wind power curve for single turbine was proposed. Firstly, atmospheric movement model was discretized. As a result, computational fluid dynamics(CFD) speedup matrix was established which reflected wind flow distribution accurately. Secondly, a new variable-weight CFD wind forecast combination method for each wind turbine was put forward, considering rotor thrust coefficient of free flow field and wakes of actuator disc. Finally, a new site-specific optimal power curve fitting method was developed with history data of wind speed and power measured by wind turbine’s anemometers. The contribution rate was used for end judgment of eliminating illegal points by an iterative algorithm. The engineering examples show that proposed method improves the accuracies of wind power curve fitting and forecasting effectively.Dispersed gridding of wind power can improve voltage stability and reduce network losses of DN. Nevertheless, it is a challenge for traditional power grid scheduling mode without reasonable plan. From the perspective of distribution network economic planning, the plan of maximum wind power integrating capacity into DN was studied considering network loss minimum and loadability. A multi-point integrating strategy of DWF was raised based on optimal active power and power factor(PF). At first, a network loss calculation method was presented considering DWF gridding. Secondly, effects of reactive power network losses on total network losses were considered for the first time, and two judgment indices were put forward for characterizing distribution network losses: active and reactive power network losses indexes. The multi-objective optimization function of minimum network losses was established. The effect of DWF on DN voltage stability margin was analyzed. At last, simulation examples of IEEE-33 and IEEE-69 prove the effectiveness of max wind power integrating and minimum network losses of the proposed strategy.In terms of power control of DWF, a new three-layer reactive power coordinated control strategy was put forward based on wind power forecasting including reactive power forecasting, setting and distribution. For reactive power distribution, two new methods of reactive power distribution for point of interconnection(POI) were proposed: reactive power allocation method considering forecasting wind speed layout using the maximum ability of wind turbine reactive power output. A new wind turbine reactive power allocation strategy was put forward considering forecasting trend of each wind turbine. The wind turbines suitable for being adjusted were selected according to dynamic filter coefficient considering reactive power forecasting and wind speed fluctuation to track reactive power commands. A multi-objective economic optimization allocation method was raised considering network losses to achieve the objects of maximum reactive power output capacity and minimum power losses. Calculating examples proved that proposed strategy increases the ability of reactive power output and reduces network losses of wind farm effectively.The uncertainty of distribution network enlarges with distributed wind power grid. The sensitivity of power voltage fluctuation influences the clients, hence it is a challenge for traditional control methods to mitigate voltage fluctuation. In order to solve this problem, a voltage fluctuation analysis and mitigation strategy of DN containing DWF was put forward. In the beginning, influences of different control strategies such as power factor control strategy, voltage control strategy and reactive power dispatch strategy on voltage fluctuation mitigation were analyzed in-depth and mathematical models were analyzed. On the secondary, influences of different wind speed and grid conditions on voltage fluctuation were proposed. Finally, a voltage fluctuation mitigation strategy was put forward as an integrated control scheme to control wind turbines’ exchanging reactive power with distributed network in order to reduce voltage fluctuation due to wind power variation. Simulation examples explain that proposed strategies effectively mitigate voltage fluctuation of POI and distribution feeder.