新疆风电场短期发电量预算方法研究

项目名称： 新疆风电场短期发电量预算方法研究

项目编号： No.50867004

项目类型： 地区科学基金项目

立项/批准年度： 2009

项目学科： 金属学与金属工艺

项目作者： 王维庆

作者单位： 新疆大学

项目金额： 23万元

中文摘要： 随着风电利用的增加，对风电场的发电量短期预算显得越来越重要，但是目前国内还没有更准确的分析方法；本项目的研究对增加风能资源利用、未来风电场的建立、合理进行电力调度和含风电电力系统运行安全性和稳定性具有理论和现实意义；同时会有较好的经济效益和社会效益。 本课题对目前已建的新疆某风电场的风力发电机机型与技术、风电场风机分布阵列及并网运行进行了分析，对所研究风电场主要机型的实际风功率曲线进行了描述，并根据此对风力发电机单机输出功率进行了计算和分析；完成了新疆风电场实际风况数据采集与数据分析，主要采集了测风塔资料及历年风电场实际运行情况的数据和部分气象数据与云图资料；基于调研采集的气象数据、风电场当前机组运行监控数据、风电场测风塔数据与风电场单机计算功率，对所研究风电场的风速和风电场的输出功率进行了分析和预测；分别采用基于时间序列、动态神经网和紧致小波神经网模型、支持向量机的方法对所研究风电场的风速和功率进行了预测，编写了预测分析软件，并进行了预测误差分析。

中文关键词： 风力发电机；风电场；风速和功率预测；神经网；支持向量机

英文摘要： As the increasing of the wind power utilization, it has become more and more important to predict the short time output power of the wind farm. But there is not better and more accurate analysis method in our country. This project has theory and practical significance for the increasing of wind energy resource utility, the building of future wind farm, the rational management of electric power, the security and steady property of the power system with on grid wind power. And the result of this project can has efficiency of both economy and society. In this project , to the studied wind farm, we have analized the wind turbine generator(WTG) types and the used technologies, the WTGs distribution array and the on grid method of wind farm. The real power curves of the main WTGs used in the studied wind farm has been discribed and according the power curves, the output power of the WTGs were calculated and analized. The work of real wind condition dada collection and analysis has been complete. The data from the wind masurement tower and from the history accumulation real runing condition resources of the wind farm, some meteorological data and cloud graphs materials were mainly gethered. Basid on the investigation and study, the collected meteological data, the present real WTG runing data from the wind farm monitor control part, the data from the wind masurement tower and the calculated unit output power of the WTG, the wind speed and output power of the studied wind farm have been analysed and forcasted. The method of analysis of time series, dynamic neural network, compact wavelet neural network and the support vector machine were studied and they were used to predict the wind speed and output power of the studied wind farm. We have also written some prediction and analysis software.And the error of predict has been analized.

英文关键词： wind turbine generator; wind farm; prediction of wind speed and wind power; neural network; support vector machine