Research On Wind Power Prediction And Wind Farm Energy Storage Capacity

Due to continuing deterioration of environment and climate as well as rising price of non-renewable energy sources such as oil and coal, the governments around the world treat clean and renewable energy with strategic significance. Because of its mature technology, wind power is widely applied and the installed capacity of wind power is growing year by year. Therefore, the impact of wind farms on grid stability is greater and attracts more and more public attention.On account of the randomness and intermittency of wind, it is difficult to predict wind power accurately and dispatch the wind output power from wind farm in real time by power system. For this reason, wind power integration brings many influences on power grid, such as voltage fluctuation and flicker. Thus, to improve the prediction accuracy of wind power and smooth wind output power with storage system are of great theoretical significance and actual value, which can make the wind power dispatching be compatible with power dispatching, promote the development of wind power and increase the ability of accommodating wind power.Under this background, it is proposed in this paper to apply storage system in wind farms, and calculate optimal storage capacity to smooth wind power from different time level, which can enable wind output power to participate grid dispatching. The main work can be summed up as follows:(1) Considering the influences that non-linear relationship between wind speed and wind power has on the wind power prediction accuracy, an improved model to forecast wind speed and wind power in short term is built based on time series and back propagation artificial neural network (BP-ANN). First, the time series model is built to forecast wind speed. Then the BP-ANN model of wind speed-to-power is set up, whose input data is the predicted wind speed. The improved model can effectively improve wind power prediction accuracy, and provide the basis for wind power dispatching.(2) Given the limits of storage capacity and charging and discharging power in storage battery, a mathematical battery model is established, which provides proper battery model for storage capacity optimization.(3) To reduce the impacts of wind power randomly fluctuating component on power grid, an optimization method for determining storage capacity is proposed in the paper to smooth the minute-scale part of wind power, which is based on the statistical analysis of wind power minute-scale characteristics. The storage capacity and maximum charging and discharging power is determined by interval theory of probability and statistics on the basis of wind power minute-scale characteristics The method can smooth output wind power by small capacity storage and reduce the impacts of wind power on power grid.(4) The energy storage system makes randomly fluctuating wind power feasible for pre-determined power dispatching. Taking the adaptability of power dispatching decision and the economy of wind power system operation into consideration, an optimal model for determining energy storage capacity is proposed in the paper. The model is built to minimize the total cost of energy infrastructure and operation including loss of wind energy and storage system on the basis of reference valve of wind farm output power suitable for power dispatching operation program. The research could guarantee the economy of wind-storage system and minimize the probability of power fluctuation of wind farm in the power dispatching operation program, and then reach the stable and effective connection between wind power dispatching and power dispatching.