Research On Wind Speed Forecasting Considering Effects Of Meteorological Factors In Wind Farm

Nowadays global energy crisis has become a thorny issues of common concern of all countries in the world, it is imminent to solve the energy crisis. In this situation, the development and utilization of new energy has been listed on the agenda of the inational energy development. As the most potential substitute for traditional power generation technology,wind energy has been widely used for it’s advantages.Economy, green and renewable wind resource characteristics inherent provides us with a broad application prospect. At the same time,strong fluctuation wind resource for large-scale wind power connected and proposed the new task, how to control the output of wind power well is of great significance to the stability of the power system.So the research of the wind speed forecasting of wind speed forecast should be imperative, accurate results can make us formulate more reasonable schedule planning and other related measures, reduce the power output fluctuation, enhance the stability of the system.There are many factors affect the wind speed such as topography, meteorological factors. Among these factors, the changes of temperature, humidity, pressure have the most effect on the wind speed,which made the disciplinarian of wind speed becomes very poor, the accuracy is not high. Based on the above issues, this thesis has done the following work mainly:(1) It is a prerequisite to know the distribution characteristics of wind resources for the study of wind speed forecasting, based on the development of the introduction of domestic and foreign wind power industry, wind speed forecasting research, the wind resource distribution and characteristics in China are analyzed, and clarifies the impact of wind power integration on power system.(2) The correlation analysis between the temperature, humidity, air pressure and wind speed. Due to the wind has a strong randomness and easily affected by many factors, especially may be affected by the local meteorological conditions, and the affecting factors are not single, is a relatively complex. The degree of correlation coefficient of correlation of mathematical analysis methods used in this paper on the basis of the description of the intuitive analysis on the main factors and wind speed, the correlation coefficient was used to measure the basis, determine the degree of correlation, and the historical wind speed with the largest effected factors as the input of forecast model, we can test the forecast accuracy of model.(3) The improvement of prediction method and predictive model plays a very important role in improving the accuracy of wind speed forecasting. This thesis mainly adopts the combination of BP neural network and hybrid clustering method to form a prediction model for wind speed prediction. The hybrid clustering can be effectively classified training samples, and the BP neural network has a characteristic is higher the similarity of training samples and the prediction precision is high, the advantages of both complement each other, so as to improve the accuracy of wind speed forecasting.