地区电网负荷特性分析及短期负荷预测研究

发布时间：2018-04-05 16:13

本文选题：负荷特性　切入点：灰色关联度　出处：《华南理工大学》2015年硕士论文

【摘要】：伴随社会经济的发展电力系统负荷特性也发生明显变化。区域电网的负荷特性分析是地区电网的调峰、市场营销以及负荷预测工作正常开展的重要保证。对广东省某市电网进行调研,采集连续几年负荷数据,利用负荷数据的二维特性设计单一数据缺失以及连续数据缺失的修补方法,并对原始负荷数据进行预处理,为地区电网负荷特性分析及预测提供数据支持。电力系统负荷特性分析时,需要借助负荷特性指标体系进行。建立了负荷特性分析指标体系,利用经过预处理的可靠负荷数据,通过指标法、曲线法和比较法按照日、月、年开展地区电网的负荷特性分析,并对节假日连休期间的负荷特性进行单独重点分析。电力系统负荷特性变化是多因素相互影响的结果,从经济、社会、政策和气象等方面定性分析对负荷特性的影响,并结合当地实际统计数据利用改进灰色关联分析法开展地区电网负荷特性影响因素定量分析,为地区电网调度、规划和营销提供参考。由于节假日负荷特性表现出较强的特殊性,对节假日负荷特性影响因素展开专门研究。地区电网负荷特性分析有利于提高地区电网负荷预测精度。按照正常日和节假日分别进行地区电网短期负荷预测研究。正常日负荷预测采用支持向量机预测模型,并根据地区负荷特性特点构造训练样本集;节假日负荷预测中,提出基于指数平滑关联修正的预测方法,引入气象、日期以及连休类型关联修正,改善节假日负荷预测贫数据影响。通过实例验证,文中预测方法精度较高。
[Abstract]:With the development of social economy, the load characteristics of power system also change obviously.The load characteristic analysis of the regional power grid is an important guarantee for the normal development of the peak-shaving, marketing and load forecasting work of the regional power grid.This paper investigates the power network of a certain city in Guangdong province, collects the load data for several consecutive years, designs the repair method of the single data missing and the continuous data missing by using the two-dimensional characteristic of the load data, and preprocesses the original load data.Provide data support for load characteristic analysis and forecast of regional power network.Power system load characteristic analysis, need to rely on load characteristic index system.The index system of load characteristic analysis is established, and the load characteristic analysis of regional power network is carried out by index method, curve method and comparison method according to day, month and year by using pretreated reliable load data.The load characteristics during the continuous holiday period are analyzed separately.The change of power system load characteristics is the result of the interaction of many factors. The impact of power system load characteristics is analyzed qualitatively from the aspects of economy, society, policy and meteorology.Combined with the local actual statistical data, the improved grey relational analysis method is used to carry out the quantitative analysis of the factors affecting the load characteristics of the regional power grid, which provides a reference for the regional power grid dispatching, planning and marketing.Because of the strong particularity of the holiday load characteristics, the factors affecting the holiday load characteristics are studied.The analysis of load characteristics of regional power network is helpful to improve the precision of load forecasting.According to the normal days and holidays, the research on the short-term load forecasting of regional power grid is carried out.The normal daily load forecasting adopts the support vector machine forecasting model, and constructs the training sample set according to the characteristic of local load, in the holiday load forecasting, the forecasting method based on exponential smoothing correlation correction is put forward, and the weather is introduced.Date and continuous rest type correlation correction to improve the impact of holiday load forecasting poor data.It is proved that the accuracy of the prediction method is high.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TM714;TM715

【参考文献】