Study On Dispatching And Control Theory Of Wind Farm Integrating Into Grid

As the most mature and most widely commercialized technical form of renewable power generation, wind power generation is already the most prospective alternative of fossil fuel generation. It plays an important role in reducing carbon emission and improving energy structure. However, the fluctuation of wind power brings dispatch and control challenges in wind farm and connected grid operation, which becomes an obstacle to large-scale development of wind power and even an issue to safety and economic operation of power grid. Thus, how to effectively utilize wind power and develop wind power generation in coordination with the connected grid is the most concerned and focused problems in researches.Currently, reactive voltage regulation and active power dispatch in wind farm and the connected grid are difficult to control coordinately, because fluctuating wind power generation is unable to realize initiative regulation as traditional generation does. However, initiative regulation of wind power system can be realized or partly realized by way of accurately grasping the law in wind power fluctuation, deeply analyzing the power characteristic of wind power generators and rationally formulating coordination control between wind farm and the connected grid. And then under the condition of ensuring security, the economy of wind power can be fully in action. These are theoretically significant and actually valuable for the development of wind power.Consequently, the dissertation tries to grasp and adapt to the influence of wind power fluctuation on wind power system and connected grid and makes use of it as a clue, and regards fully or partly realizing the initiative regulation of wind power system as the core. Based on the above, reactive voltage regulation and active power optimal dispatch between wind power and connected grid are analyzed. The main work can be summed up as follows:(1) An optimization method for determining reactive compensation capacity is proposed in the thesis, based on the statistical analysis of wind output power distribution characteristic by using the estimated power data. Considering the wind power distribution characteristic and the ability of reactive power regulating, the optimization model is built to minimize the total cost of operation and reactive power compensation investment compromisingly. The study can guarantee that the individual and regular information of wind power is used to determine the reactive compensation model. And then it can realize the continuous and seamless reactive voltage regulation for DFIG wind farms by the least lost, which is the basis of wind farm's participation in reactive voltage regulation of connected grid. Meanwhile, the optimal reactive compensation can effectively improve the pertinence of reactive compensation for wind farm and enhance the economy and initiative of reactive voltage regulating for wind farm system.(2) STATCOM is introduced as a reactive compensation unit for wind farms, based on the analysis of DFIG's power characteristic, which is that DFIG's reactive power output changes with active power output variation, and that DFIG needs to absorb reactive power in order to keep the operation characteristic when active power output is close to its rated power. And then, considering the relationship between wind farm, reactive compensation unit and connected grid, a reactive power control strategy for wind power system is proposed in the dissertation The control strategy can effectively restrain the fluctuation of reactive power and stabilize the ability of reactive power regulation. What's more, it can enlarge operation range of wind farm reactive power and solve the problem of drawing reactive power under a certain condition. Thus, the voltage level of point of common coupling is improved, and active power of wind farm is guaranteed.(3) According to the low-voltage ride-through of DFIG, a real time coordinated control strategy of reactive power and voltage for wind farm system is proposed, which effectively coordinates reactive power demand relationship between wind farm, reactive compensation unit and connected grid. The control strategy can coordinate the reactive compensation unit and reactive power output of wind farm during the fault, provide the voltage support for power system and improve the fault ride-through ability of wind farm and connected grid. Considering the control strategy at the normal and fault state, a real time coordinated control system of wind farm reactive power is built, which improves the ability of reactive power regulation of wind farm system at the normal and fault state by the flexible operation mode and rapid reactive power response.(4) The wind power prediction error seriously influence dispatch and control with wind farm scale increasing. Thus, a novel method for classifying forecasted wind power according to its credible level is presented based on accurate analysis of wind power fluctuation. The model classifies forecasted power into three types:base output, secondary output and high-frequency output. Then, the classification can be used as a guide in power grid decision, so that the effects of prediction error can be reduced. What's more, it can correspond with time interval division used in power system dispatch and load prediction, provide basis for power dispatch and control, reduce decision risk due to prediction error and cut down reserve of wind power. Therefore, the model can balance economy and risk of power system dispatch decision.(5) Based on the classification of forecasted wind power, an model for unit commitment optimization and spinning reserve allocation which considers base output and secondary output is built, and the coordinated operating strategy of wind power for regional power grid is also presented. Different output can be integrated into day-ahead dispatch, pre-dispatch and on-line dispatch respectively, so that the risk of day-ahead dispatch due to prediction error can decrease. Meanwhile, ultra-short term power prediction is introduced to evaluate high-frequency output again, then pre-dispatch and on-line dispatch can be re-optimized. The decision risk is further reduced by ultra-short time prediction based on the wind power control strategy at the cycle of revised pre-dispatch and on-line dispatch. The initiative and economy of regulation and control of wind power is improved by matching different type of output with responding time interval.