DOC《毕业设计（论文）》

毕业设计(论文) 题目基于支持向量机的 电力短期负荷预测 基于支持向量机的短期电力负荷预测 摘要短期负荷预测是电力系统安全调度、经济运行的重要依据,负荷预测的精度直接影响到电力系统运行的可靠性、经济性和供电质量.

因此,寻求合适的负荷预测方法最大限度的提高预测精度具有重要的应用价值. 论文首先阐述了负荷预测的应用研究现状,概括了负荷预测的特点及其影响因素,归纳了短期负荷预测的常用方法,并分析了各种方法的优劣;

接着介绍了作为支持向量机(SVM)理论基础的统计学习理论和SVM的原理,推导了SVM回归模型;

本文采用最小二乘支持向量机(LSSVM)模型,根据浙江台州某地区的历史负荷数据和气象数据,分析影响预测的各种因素,总结了负荷变化的规律性,对历史负荷数据中的 异常数据 进行修正,对负荷预测中要考虑的相关因素进行了归一化处理.LSSVM中的两个参数对模型有很大影响,而目前依然是基于经验的办法解决.对此,本文采用粒子群优化算法对模型参数进行寻优,以测试集误差作为判决依据,实现模型参数的优化选择,使得预测精度有所提高.实际算例表明,本文的预测方法收敛性好、有较高的预测精度和较快的训练速度. 关键词:短期负荷预测,支持向量机,最小二乘支持向量机,粒子群优化,参数选取 Abstract Short-term load forecasting is important basis of safely assigning and economically running. The forecasting precision will directly affect the reliability, economy running and supplying power quality of power system. So finding an appropriate load forecasting method to improve the accuracy of precision has important application value. Paper first expounds the recent application research of load forecasting, summarized the characteristics of load forecasting and influencing factors, summed up common methods of short-term load forecasting, and analyzed the advantages and disadvantages of each method;

then introduced statistical learning theory and the principle of SVM as the basis of support vector machine (SVM ) theory, SVM regression model is derived;

this paper adopted least squares support vector machine (LSSVM) model, according to the historical load data and meteorological data of a certain area of Zhejiang Taizhou, Analysised the various factors affecting the forecast, summed up the regularity of load change , amended outliers in the historical load data,the load forecasting factors to be considered were normalized. The two parameters of LSSVM have a significant impact on the model, but it is still soluted based on the experience currently. So, this paper adopted particle swarm optimization algorithm to optimized the model parameters, make the test set error as the judgments, realized the optimization of model parameters, maked prediction accuracy improved. Practical examples show that convergence of prediction method was pefect, had a higher prediction accuracy and fast training speed. Key words:Short-term Load Forecasting, Support Vector Machines, Least Square Support Vector Machines, Panicle Swarm Optimization, Parameter Selection 目录第1章 绪论

1 1.1 负荷预测研究的背景和意义

1 1.2 国内外研究和应用现状

2 1.2.1 短期负荷预测的国内外研究现状

2 1.2.2 负荷预测的常用方法

2 1.3支持向量机在短期负荷预测中的应用情况以及存在的问题

4 1.4 本文的主要工作

6 第2章 支持向量机和改进粒子群参数优化