《基于支持向量机的电能质量综合评价》由会员分享，可在线阅读，更多相关《基于支持向量机的电能质量综合评价（45页珍藏版）》请在金锄头文库上搜索。
1、 学号： 常 州 大 学 毕业设计（论文）（2012届）题 目 学 生 学 院 专业班级 校内指导教师 专业技术职务 校外指导老师 专业技术职务 二一二年六月基于支持向量机的电能质量综合评价摘 要 ：统计学习理论是在研究小样本统计估计和预测的过程中发展起来的一种新兴的理论，它试图从更本质的层面上来研究机器学习的问题。作为统计学习理论的最新发展和结构风险最小化准则的具体体现，支持向量机（SVM）具有全局最优、结构简单、推广能力强等优点，在解决小样本、非线性及高维模式识别问题中表现出许多特有的优势，并能够推广到函数拟合等其他机器学习问题中。本文研究了SVM在电能质量分析中的应用，提出了一种包含综合
2、集成赋权法和支持向量机的电能质量评价方法。在综合集成赋权法中,层次分析法确定主观权重,改进拉开档次法确定客观权重,基于加法原理将两者结合确定综合权重。由此产生支持向量机的初始训练数据,通过训练支持向量机能够得到电能质量综合评价模型及综合评价结果。算例结果表明,提出的电能质量评价策略保留了支持向量机在处理小样本、非线性问题时的独特优势,能快速得到合理的评价结果,并具有良好的泛化能力。关键词：电能质量；支持向量机(SVM)；综合集成赋权；层次分析法；综合评价 Power Quality Comprehensive Evaluation based on Support Vector Machine
3、 Abstract：Statistical learning theory is newly developed theory for studying the statistical estimation and prediction problem based on small number of samples. It studies the nature of machine learning. As the latest development of statistical learning theory and the embodiment of structural minimi
4、zation criterion, Support Vector Machines(SVM) provide efficient and powerful algorithms that are capable of dealing with high dimensional input features and with theoretical bounds on the generalization error and sparseness of the solution provided by statistical learning theory. SVM has few free p
5、arameters requiring tuning is simple to implement, and are trained through optimization of a convex quadratic cost function, which ensures the global optimization of the SVM solution. Furthermore, SVM-based solutions are sparse in the training date and defined only by the most informative training p
6、oints. SVM presents a lot of advantages for solving the small samples, nonlinear and high dimensional pattern recognition, as well as other machine-learning problems such as function fitting. A novel comprehensive strategy, consisting of an integrated weight method and support vector machine(SVM) me
7、thod, was proposed to evaluate PQ. In the integrated weight method, analysis hierarchy process (AHP) was employed to determine the subjective weights, and improved scatter degree (ISD) was introduced to decide the objective weights, then the comprehensive weights could be determined based on additio
8、n principle. The integrated weight method could produce the training data for SVM, and then the evaluation model of PQ was built. Test results show that the proposed strategy can deal with small data sets and nonlinear problems effectively, and obtain reasonable evaluation results quickly. The propo
9、sed strategy has so good generalization performance that it is fit for large practical samples. Key words: Power quality;Support Vector Machine; integrated weight method; analysishierarchy process; Comprehensive Evaluation. 摘要目次术语表1 绪论11.1 引言11.2 选题背景21.3 电能质量研究现状32 电能质量62.1 电能质量的基本概念62.2 电能质量研究概况72.3 电能质量国标与主要内容82.4 电能质量的其他指标112.5 本章小结123 电能质量指标权重的确定及评价体系的建立133.1 电能质量六大单项指标143.2 电能质量指标等级量化143.3 指标权重的确定153.4 建立电能质量综合评价体系163.5 关于电能质量指标制定的探讨173.6 本章小结184 基于支持向量机的电能质量分析204.1 引言204.2 统计学习理论214.3 支持向量机244.3.1 支持向量机原理244.3.2 支持向量机数学模型254.3.3 优化条件和序列最小优化算法264.4 基于Alphaminer建立评价模型284
