Research On Characteristics Of Wind Power And Relevant Application

With the growing scale of wind power connected to our grid, on the one hand, it provides us clean energy, on the other hand, it brings a series of adverse effects to power system control, operation as well as planning. Wind power has its internal stochastic volatility can be the essential reason for its disadvantages as energy source. At present, fluctuation of wind power has become a hot area of wind power research topic. For regional grid connected to a couple of wind farms, similarities and differences in fluctuations between wind farm cluster and a single wind farm is a significant problem to be solved. In this dissertation, mehanism of production and influencing factor of wind power fluctuations are illustrated. Then several indexs to measure wind power fluctuations are proposed. Both of them are foundation and necessary condition for the following studies. Afterwards, time-space effects of wind power fluctuations is analyzed. For time effect, probabilities of different amplitudes of fluctuation concerning different time scales are disscussed. Space effect indicates that power variation from a single wind farm to wind farm clusters focuses on the space scale effect, that is, smooth effect. Wind power tends to fluctuate weakly with expansion of its scale is founded to be a normal phenomenon. Analysis of smooth effect mainly include its definition, mechanism, influencing factors and probable grid application. Making reference to power network dispatching adjustment is a significant application of wind power fluctuation research. The purpose for discussion in this paper is to decrease wind energy curtailment as much as possible with respect to power network safe operation considering wind generation limitation. Based on correlation analysis, we put forward a priority queuing algorithm with optimal sequence of wind farms according to cross correlation coefficient when there is a limit to regional wind power output. Then, optimum power allocation among local wind farms referring to the queuing algorithm is presented, and its superiority is verified through comparision to other possible schemes from the point of view of economy and volatility. In the end, method and process to produce forecast states of wind energy fluctuations based on Markov chain are presented, referring to Markov theory. In the method, state switching thought is the core point, and the main point is to divide states and generate initial probability vector as well as transition probability matrix. By the means of comparing forecast results to real results, validity of the forecasting method based upon Markov chain is proved. And for this forecast approach, the result of multistep prediction and one-step prediction is quite resemble.On the basis of fluctuation state forecast, taking conventional energy resources as spare capacity to smooth power unbalance caused by wind power fluctuation tends to be a potential meaningful research field in the following study on fluctuations of intermittent energy.