短期负荷预测毕业设计论文.docVIP

【】Support Vector Machine, SVM）算法已经被认为是一种预测负荷数据更好的方法，SVM 是基于结构风险最小化原则的新型机器学习方法，有着比传统的基于经验风险最小化原则的人工神经网络更优越的性能。因此，本文将SVM 引入短期负荷预测，以期开发出一种性能更加优越的短期负荷预测方法。 为了进行有效的预测，提高预测精度，本文提出将最小二乘支持向量机LS-SVM（Least Square Support Vector Machine）对负荷进行预测。在建立模型的过程中，通过对训练对象的不断改进以及调整，提高对负荷数据预测的精确性。得到更加优化的预测模型，使其更好的利用到负荷预测的领域中。 关键词 【Abstract】 Power system short term load forecasting is the premise of safe and economic operation, with the way the promotion of electricity price and electricity market reform, we should strive to timely and accurately grasp the load change information on the importance of load forecasting and urgency referred to an unprecedented level, but also on the accuracy of load forecast a higher demand. During the recent period, support vector machine (Support Vector Machine, SVM) algorithm has been Bei Ren Wei is a better forecast load data of method, SVM is based on structural risk minimization principle of the new machine learning method, You Zhaobis based on the traditional empirical risk minimization of the artificial neural network better performance. Therefore, this article will introduce short-term load forecasting SVM in order to develop a performance more superior short-term load forecasting. In order to effectively predict, improve forecasting accuracy, this paper will support vector machine LS-SVM (Least Square Support Vector Machine) on the load forecast. In the modeling process,through the continuous improvement of training objects and changes to improve the accuracy of the forecast load data. Be more optimal forecasting model, to make better use of the field to load forecast. Keywords: Short term load forecasting; Adjust the training sample; LSSVM 目录 1 引言 1 1.1 负荷预测的特点及影响因素 1 1.2 负荷预测的要求和步骤 3 2 短期电力负荷预测方法综述 4 2.1 短期电力系统负荷预测的传统方法 4 2.1.1 回归分析法 4 2.1.2 趋势外推法 5 2.1.3 时间序列法 5 2.1.4 人工神经网络法 6 2.2 专家系统法 9 2.2.1 专家系统的特点及组成 9 2.2.2 专家系统在短期电力负荷预测中的应用 9 3 LS-SVM原理 10 3.1 统计学理论 10 3.1.1 VC维（Vapnik-Chervonenkis Dimension） 10 3.1.2 推广性的界 11 3.1.3 结构风险最小化原则（SRM） 12