基于原子稀疏分解的风电功率实时预测研究

发布时间：2018-05-30 23:13

本文选题：风电功率 + 超短期　；参考：《东北电力大学》2017年硕士论文

【摘要】：风能是至关重要的低碳能源,有实现可持续能源供应的潜力,风力发电已成为各国重点发展的绿色能源之一。风电发展迅速,装机容量逐年增加,预计到2020年,全球风力发电装机容量将达到12亿千瓦,能够满足世界电力总量12%的需求。近几年我国风电年装机容量成倍增长,至2014年底,中国累计风电装机容量114609兆瓦,我国已成为世界装机容量最大的国家。根据能源局在2011年发布的文件《风电厂功率预测预报管理暂行办法》可知,实时预测是指自上报时刻起未来15分至4小时的预测预报,时间分辨率为15分钟。故本课题研究的实时风电功率预测是以时间间隔为15分钟的风电功率时间序列为主要研究对象,并对其进行滚动预测16步的超短期风电功率预测。以此得到的预测结果,可以服务于风电场机组实时有功出力的调整,对提高风能的利用率有重要意义。本课题从风电功率波动特性着手,首先阅读国内外文献,找到或定义刻画风电功率波动特性的指标,分析风电功率波动的概率分布,分析风电功率波动的原因;阅读国内外关于风电功率波动特性和风电功率预测方面的文献,了解风电功率预测的研究进展,分析风电功率预测误差的成因,介绍刻画风电功率预测误差的指标;研究国内外关于原子稀疏分解理论方面的文献,将原子稀疏分解理论应用于风电功率时间序列的前期分解;在现有风电功率预测模型的基础上,将原子稀疏分解理论组合现有预测模型应用于风电功率的超短期实时预测,并且分析新的组合预测模型对风电功率实时预测精度的影响;进行风电功率实时预测误差分析,验证新的组合预测模型的有效性;最后搭建基于VB编程语言的风电功率预测平台。
[Abstract]:Wind energy is a very important low-carbon energy, and has the potential to achieve sustainable energy supply. Wind power generation has become one of the key green energy. Wind power is developing rapidly and its installed capacity is increasing year by year. It is estimated that by 2020, the installed capacity of global wind power generation will reach 1.2 billion kilowatts, which can meet the demand of 12 percent of the world's total electricity. In recent years, the annual installed capacity of wind power in China has increased exponentially. By the end of 2014, the total installed capacity of wind power in China was 114609 MW, and China has become the largest country in the world. According to the document issued by the Energy Bureau in 2011, "interim measures for power forecasting and forecasting of wind power plants", real-time prediction refers to the forecast for the next 15 to 4 hours from the reporting moment, with a time resolution of 15 minutes. Therefore, the real time wind power prediction in this research is based on the wind power time series with a time interval of 15 minutes, and the ultra short term wind power prediction with 16 steps rolling prediction is carried out. The predicted results can be used to adjust the real time active power output of wind farm units, and it is of great significance to improve the utilization rate of wind energy. This topic starts with the characteristic of wind power fluctuation, first reads the domestic and foreign literature, finds out or defines the index to describe the characteristic of wind power fluctuation, analyzes the probability distribution of wind power fluctuation, and analyzes the reason of wind power fluctuation. This paper reads the literatures on wind power fluctuation characteristics and wind power prediction at home and abroad, understands the research progress of wind power prediction, analyzes the causes of wind power prediction errors, and introduces the indicators of wind power prediction errors. This paper studies the theory of atomic sparse decomposition at home and abroad, applies the theory of atomic sparse decomposition to the pre-decomposition of wind power time series, and based on the existing wind power prediction model, In this paper, the atomic sparse decomposition theory is applied to the ultra-short-term real-time wind power prediction, and the influence of the new combined forecasting model on the wind power real-time prediction accuracy is analyzed, and the error analysis of wind power real-time prediction is carried out. The validity of the new combined forecasting model is verified. Finally, the wind power prediction platform based on VB programming language is built.
【学位授予单位】：东北电力大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM614

【参考文献】