风电功率短期预测方法应用研究.docVIP

13届 　 分 类 号: 　　　　　　　　　　　　　　　　　 单位代码:10452 毕业论文（设计） 风电功率短期预测方法研究 　　　　　　 　2013年3月20日摘 要 随着风电接入规模的增加, 风电功率预测日益重要，为了有效减轻风能波动对电网的影响，提高风电在电力市场中的竞争力，风电功率预测研究具有重要意义.该文首先根据时间序列预测模型、灰度预测模型、分段多项式拟合模型三种模型对风电功率进行预测，再对三种模型的预测结果进行准确率和合格率分析.通过分析比较三种模型下的误差，得出了分段多项式拟合模型预测准确率相对较高，但当风电功率占电网负荷的比例超过一定数值之后,可能会严重影响电网的电能质量和电力系统的运行,风电功率随着风速变化而随机波动的特性也给电力系统造成难以承受的冲击.因此必须进一步提高风电功率实时预测的精度.而采用神经网络模型，利用BP网络完成对风电功率的预测，可以进一步提高风电功率的实时预测精度. 关键词：风电功率预测；时间序列预测模型；分段多项式拟合模型；神经网络模型；误差分析 Abstract With the increase of wind electricity access scale of wind power prediction increasingly important, in order to effectively reduce the impact of wind power fluctuations on the power grid, improve the competitive force of wind power in the electricity market, wind electric power prediction research has significance In this paper, first of all based on time series forecastmodel, gray forecasting model, segmented polynomial fitting model three models to predict the wind power, and the predicted results of the three models, the accuracy and the pass rate obtained by analyzing the three model errorpiecewise polynomial fitting model prediction accuracy is relatively high, but when the wind power accounted for the proportion of the grid load exceeds a certain value, may seriously affect the operation of the grid power quality and power system, wind power with the change of wind speed and randomthe the fluctuation characteristics also cause unacceptable impact to the power system. therefore must further improve the accuracy of real-time prediction of wind power and using neural network model, complete the prediction of wind power with the BP network, can further improve the real-time prediction accuracy of the wind electric power. Keywords: wind power prediction; time series prediction model; piecewise polynomial fitting model; neural network model; Error Analysis 目 录 1 引言 1 1.1 问题的提出 1 1.2问题分析 2 1.2.1模型的假设约定 2 2.2.2 符号的说明 2 2 模型的建立与求解 4