基于神经网络的电力系统短期负荷预测研究-毕业论文.docx

毕 业 设 计(论文) 基于神经网络的电力系统短期负荷预测研究 基于神经网络的电力系统短期负荷预测研究 摘 要 我国经济发展迅速，当然电力行业也是蓬勃发展，作为一名电气专业的学生，我很骄傲能占有这行业的一席之地，在未来的日子也可以显出自己的绵薄之力。纵观电力事业的发展，在电网管理的方面变得更加智能化，同时电力系统短期负荷预测研究也逐渐成为电力研究方面的重要内容，对于一个电力系统而言，最重要的是安全性，现在我们家家户户都离不开电，所以安全性，稳定性和可靠性是研究电力系统的重要把握之处，也是负荷研究的关注点之一。大学四年，我们去花庄变电站参观过，也去陡河电厂实习过，在电力系统调度运营过程中，主要应用的措施为短期负荷预测，预测结果的精准性对整个电力系统的正常运行起到至关重要的作用。 本文首先介绍了负荷预测的方法，从古至今，从国外到国内。紧接着再对电力负荷的概念、特性特点的介绍，同时也分析了负荷特性。 本文主要应用BP神经网络，Elman神经网络，RBF神经网络以及小波神经网络进行分析，主要对影响负荷预测模型中的各种因素进行了分析。 对于工作日的负荷模型，根据电力负荷的特点，在考虑天气温度、日类型、实际历史负荷等因素对预测负荷影响，通过MATLAB编程语言来实现。 学习蚁群算法，粒子群算法，遗传算法的优化的理论以及步骤，了解在神经网络基础上，还可以进行优化，得到更加精确的结论。 关键词：短期负荷预测；神经网络；蚁群算法；粒子群算法；遗传算法  RESEARCH ON SHORT-TERM LOAD FORECASTING OF POWER SYSTEM BASED ON NEURAL NRTWORK Abstract Chinas economic development is rapid, of course, the power industry is also flourishing. As an electrical student, I am proud to have a place in this industry, and in the future, I can also show my modest strength. Throughout the development of power industry, the management of power grid has become more intelligent, and the research of short-term load forecasting of power system has gradually become an important part of power research. For a power system, the most important thing is security. Now our families are inseparable from electricity, so security, stability and reliability are the important assurance of the study of power system. It is also one of the focuses of load research. In the four years of university, we visited Huazhuang Substation and also interned in Douhe Power Plant. In the process of dispatching and operation of power system, the main measure used is short-term load forecasting. The accuracy of forecasting results plays a vital role in the normal operation of the whole power system. (1) This paper first introduces the methods of load forecasting, from ancient times to present, from abroad to home. Next, the concept and characteristics of power load are introduced, and the load characteristics are also analyzed.  (2) In