基于神经网络的分布式光伏电站短期功率预测

发布时间：2018-03-23 19:31

本文选题：神经网络　切入点：预测方法　出处：《安徽理工大学》2017年硕士论文

【摘要】：随着世界经济的高速发展,我们需求的能源越来越多,在化石能源的逐渐短缺和环境保护双重压力下,能重复利用的清洁能源在发电系统中占据的比例越来越大。太阳能以其自身的很多优点受到社会的大力瞩目,光伏发电也成为一种可重复利用的发电形式。由于,光伏功率发电存在着不确定性和间歇性等缺点,给大范围的光伏发电并网运行带来了很高的难度,使其影响发电系统的安全、可靠、经济的运行。因此,精确预测光伏发电功率有其重要的现实意义和指导意义。本文是将光伏发电的短期预测功率作为研究对象的,通过分析影响光伏发电短期功率的因素,建立准确的预测模型,研究了光伏发电系统的短期功率预测。大部分的研究成果都是在一定的条件下完成的,本文首先分析了光伏发电的国内外现状,其次介绍了神经网络的一些理论基础、组成结构、分类情况。在大量阅读国内外文献的基础上,总结了一些典型的预测方法和预测模型。本文分析了人工神经网络预测模型优缺点,提出了基于改进型的BP神经网络分布式光伏发电短期功率的预测方法。在大量历史数据的基础上进行了神经网络模型的训练,模型的预测结果证明论文中提出的光伏发电短期预测功率的预测方法是正确的、可行的。本文的研究为光伏发电系统的大量运行提供了理论依据,可以提高光伏并网发电系统的预测精度,帮助电力部门制定详细的发电和调度计划,以提高电力发电的运行稳定性。
[Abstract]:With the rapid development of the world economy, we need more and more energy. Under the pressure of fossil energy shortage and environmental protection, Renewable clean energy accounts for more and more of the power generation system. Solar energy has attracted great attention of society for its many advantages. Photovoltaic power generation has become a form of renewable power generation. There are some disadvantages in photovoltaic power generation, such as uncertainty and intermittency, which bring high difficulty to the grid-connected operation of large-scale photovoltaic power generation, and make it affect the safe, reliable and economical operation of the power generation system. Accurate prediction of photovoltaic power generation has important practical significance and guiding significance. In this paper, the short-term predictive power of photovoltaic generation is taken as the research object, and an accurate prediction model is established by analyzing the factors affecting the short-term power of photovoltaic generation. The short-term power prediction of photovoltaic power generation system is studied. Most of the research results are completed under certain conditions. Firstly, the present situation of photovoltaic power generation at home and abroad is analyzed, and then some theoretical bases of neural network are introduced. On the basis of reading a lot of literature at home and abroad, some typical prediction methods and models are summarized. The advantages and disadvantages of artificial neural network prediction model are analyzed in this paper. Based on the improved BP neural network, the short-term power prediction method of distributed photovoltaic power generation is proposed, and the neural network model is trained on the basis of a large number of historical data. The prediction results of the model prove that the short-term power prediction method proposed in this paper is correct and feasible. The research in this paper provides a theoretical basis for a large number of photovoltaic power generation systems. It can improve the prediction accuracy of grid-connected photovoltaic power generation system and help the power sector to formulate detailed generation and scheduling plans to improve the operational stability of power generation.
【学位授予单位】：安徽理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP183;TM615

【相似文献】