短期风速预测的研究电力系统及其自动化专业论文.docxVIP

摘要 风能是一种不稳定的能源，即风速具有随机性和不可控性，而风电机组的运行特征 随风速的随机波动而变化，所以准确的去预测风电场的风速就显得尤为重要，可以减小 风电并网对电网造成的影响和提高对风电机组的控制，可以提高电网的经济性和稳定性? 本文首先简要介绍了课题研究的背景和意义，对风能的概念?特点及各种风速预测 的方法作了详细的阐述；然后针对目前对风速预测的方法优缺点和各种预测方法的试用 范围做了探讨和比较，得出选用时间序列法和小波分析法综合建模的依据? 针对风速数据原始序列的非平稳性和异变点，本文利用小波分析法的特点对先期的 风速数据进行一个先期的平稳化处理，即利用小波分解理论对先期风速进行小波分解与 小波重构，接着把重构之后的概貌和细节部分都运用时间序列模型进行预测，最后加权 生成出未来的预测风速?通过Matlab仿真可以证明,小波分析和时间序列综合模型预测效 果要远好于单独用时间序列预测的效果，该方法能够提高短期风速预测的精度，使误差 控制在10%以内，而传统的时间序列预测精度是20%左右?同时，用综合模型预测和单 独时间序列模型预测相对比，使结果更加显而易见，本文通过Matlab仿真证明了该方法 的有效性? 最后，本文对在实际的工程应用中利用时间序列及其与小波工具箱相结合的方法对 短期风速进行预测作了简要介绍，探讨了短期风速预测的实用价值? 关键词：风速预测，小波分析，Matlab，时间序列 Abstract Wind energy is an unstable energy, namely wind speed is random and dont control, and wind power unit operation with the speed of the fluctuation characteristics random changes, so accurate to predict the wind speed wind farms are particularly important to minimize wind power grid to power and improve the effects of wind generator to the control, can improve the efficiency and stability of the power grid. This paper firstly introduces the background and significance of the subject research, the winds concept, the characteristic and the forecast method of wind speed of detail; Then according to the method of predicting the advantages and disadvantages of wind speed and all kinds of forecasting methods of trial range to discuss and compare, and concludes that the selection method and time series the wavelet analysis method based on comprehensive modeling. According to the data of the original sequence wind speed of non-stationary and variation point, in this paper the characteristics of wavelet analysis method to the upfront wind speed data of the first a smooth process, namely using wavelet decomposition theory first wind speed of wavelet decomposition and reconstruction of wavelet, then after the reconstruction of the general view and the details are using time series prediction model, and finally a weighted generation predicting the future wind spe