光伏发电功率与气象影响因子关联关系的分析研究

发布时间：2018-10-05 16:04

【摘要】：可再生能源已成为我国应对世界能源危机和经济发展新形势的战略新兴产业。光伏发电作为可再生能源的重要组成部分近年来得到了快速发展，而大规模随机波动光伏发电的并网必将对电网的安全稳定和调度运行产生不利影响。对光伏发电的输出功率进行准确预测，是打破规模化光伏发电并网应用瓶颈的有效措施。光伏发电功率受多元气象因素的影响，其预测模型输入变量选取的是否合理直接影响预测精度。目前，关于光伏发电气象影响因子作用程度的定量研究很少，本文针对光伏发电功率与多元气象影响因子之间的动态关联关系开展研究，为预测模型输入变量的识别优化与合理选取提供科学依据，具有重要的理论意义和应用价值。本文在对比分析光伏发电功率与多元气象影响因子变化规律的基础上，给出了气象影响因子作用程度强弱的科学表示。首先，针对不同的气象因素，通过散点图和相关系数对其与光伏发电功率的相关性进行了分析，并讨论了不同天气类型对相关性的影响。根据相关性的大小，确定辐照度、组件温度、环境温度和风速为光伏发电功率的主气象影响因子。在相关系数的基础上，为了消除不同变量数值差异的影响，并考虑极值信息对关联程度的作用，采用灰色关联分析方法对气象影响因子作用程度进行了趋势分析。计算光伏发电功率与气象影响因子的灰色关联度和因子权重系数用以衡量它们之间的关联程度，并对不同归一化方法的计算结果进行了讨论，指出0~1区间归一化方法更适合。通过不同天气类型下灰色关联度和因子权重系数的对比，分析了气象影响因子作用程度的变化趋势。其次，由于光伏发电功率与气象影响因子之间是多重耦合的非线性关系，利用线性的相关系数和灰色关联度衡量气象影响因子作用程度较难获得满意效果，为此，采用信息熵理论对光伏发电功率与气象影响因子之间的动态关联关系进行量化研究。从信息损失的角度，定义了光伏发电功率与气象影响因子的互信息，选择等间距法近似计算其值，并对不同天气类型下互信息值的大小进行了比较。从信息相对减少的角度，引入统计相关系数的概念，分析了光伏发电功率与气象影响因子的相关性。利用互信息和统计相关系数给出了光伏发电功率与气象影响因子动态关联关系的科学度量，，并根据不同数据源的历史数据，验证了量化研究的结果。最后，通过综合对比，对相关分析、趋势分析和量化研究三种不同方法进行了评价。
[Abstract]:Renewable energy has become a strategic emerging industry in China to deal with the world energy crisis and the new situation of economic development. Photovoltaic power generation as an important part of renewable energy has been rapidly developed in recent years, and large-scale random fluctuations of photovoltaic power grid will inevitably have a negative impact on the security and stability of the grid and dispatching operation. Accurate prediction of the output power of photovoltaic power generation is an effective measure to break the bottleneck of grid-connected application of large-scale photovoltaic power generation. The power of photovoltaic generation is affected by multiple meteorological factors, and whether the input variables of the prediction model is reasonable or not has a direct impact on the prediction accuracy. At present, there are few quantitative studies on the effect of meteorological impact factors on photovoltaic power generation. This paper focuses on the dynamic correlation between photovoltaic power generation and multiple meteorological impact factors. It is of great theoretical significance and practical value to provide scientific basis for the identification, optimization and reasonable selection of input variables of prediction model. On the basis of comparing and analyzing the variation law of photovoltaic power generation power and multivariate meteorological influence factors, the scientific expression of the degree of action of meteorological influence factors is given in this paper. Firstly, according to different meteorological factors, the correlation between PV power and scattered plot and correlation coefficient is analyzed, and the influence of different weather types on the correlation is discussed. According to the correlation, the irradiance, module temperature, ambient temperature and wind speed are the main meteorological factors of photovoltaic power generation. Based on the correlation coefficient, in order to eliminate the influence of different variables and consider the effect of extreme value information on the correlation degree, the grey correlation analysis method is used to analyze the trend of meteorological influence factors. The grey correlation degree and factor weight coefficient of photovoltaic power and meteorological influence factors are calculated to measure the correlation degree between them. The results of different normalization methods are discussed and it is pointed out that the normalization method is more suitable in 0 ~ 1 interval. Based on the comparison of grey correlation degree and factor weight coefficient under different weather types, the change trend of the action degree of meteorological influence factors is analyzed. Secondly, because of the nonlinear relationship between photovoltaic power and meteorological impact factors, it is difficult to obtain satisfactory results by using linear correlation coefficient and grey correlation degree to measure the effect of meteorological impact factors. The information entropy theory is used to quantify the dynamic correlation between photovoltaic power and meteorological factors. From the point of view of information loss, the mutual information between photovoltaic power generation and meteorological influence factors is defined, and the value of mutual information under different weather types is calculated by the equal-distance method. From the point of view of relative reduction of information, the concept of statistical correlation coefficient is introduced to analyze the correlation between photovoltaic power generation and meteorological factors. Based on mutual information and statistical correlation coefficient, a scientific measure of dynamic correlation between photovoltaic power generation and meteorological impact factors is presented, and the results of quantitative research are verified according to historical data from different data sources. Finally, three different methods of correlation analysis, trend analysis and quantitative research are evaluated by comprehensive comparison.
【学位授予单位】：华北电力大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM615

【参考文献】