摘 要随着科学技术的发展，人们经济水平不断提升，国家电力系统也在进行着不断地改革与进步。电力负荷预测是实现合理利用能源、统一输配电的重要步骤，是进行系统规划、设计、调度和配电等重大工作的必备环节。进一步提高预测水平，不仅有助于电能消费的合理与经济，降低发电量，减少煤、石油等能源的消耗，提高资源利用率，保护环境；而且可以帮助合理规划电网建设，提升系统经济效益与社会效益。因此，电力系统负荷预测已逐渐被人们重视起来，成为电力系统运行和管理的主要研究对象。而短期电力负荷预测研究有利于文持系统平稳运行、 帮助电网未来进行合理规划。本文从电力系统负荷预测的基础着手，学习研究了负荷预测的意义；在充分分析已有的研究方法后， 根据实际要求和步骤， 构建具有实际意义的预测模型。最终，在详细比较不同预测方法的优缺点后，确定将现今 BP 神经网络预测模型同遗传算法模型相结合优化，利用组合算法进一步提高短期预测的精度。最后将某工厂5 月份实际负荷数据代入组合模型检测， 确定了该组合模型具有可行性。37754
毕业论文关键字： 负荷预测 遗传算法 BP 神经网络 智能算法
Abstract With the development of science, technology and economicconditions, the national power system is constantly reforming andimproving. Power load forecasting is the important step for the realizationof the reasonable utilization of energy,as well as the unified transmissionand distribution. It is the essential link of the system planning, design,scheduling and power distribution. To further improve the predictionlevel, not only help to electric of rational consumption and economy, butalso reduce generating capacity. It can reduce the consumption of coal toimprove the utilization rate of resources, in favour of protecting theenvironment. We have the reasonable planning of power grid constructionby it .Therefore, the load forecasting of power system has graduallybecome the main research object of power system operation andmanagement.The short-term power load forecasting is more far-reaching for thesystem running and the reasonable planning of the power network. Fromthe significance of load forecasting and research status of power system,this paper is to understand the theoretical knowledge of power loadforecasting. Then , established the forecasting model according to thebasic requirements and steps of load forecasting. After the carefulcomparison of different forecasting methods advantages and disadvantages, the paper determined the current BP neural networkprediction model with the genetic algorithm model combiningoptimization, using the combination algorithm to improve the short-termfurther forecasting accuracy. Finally, making instance load data into thecombination of model checking, the combination model is feasible.
Key words: Load forecasting Genetic algorithmBP nerve network Intelligent algorithm
目录
摘要I
Abstract.II
1绪论1
1.1研究背景及意义.1
1.2国内外研究现状.2
1.3本文主要内容.4
2电力负荷预测研究.5
2.1电力负荷预测的概念及分类.5
2.2电力系统负荷的分类及特点.5
2.3电力系统负荷预测的特点.6
2.4电力系统负荷预测的影响因素.8
2.5负荷预测的基本步骤及误差原因.8
2.5.1负荷预测的基本步骤8
2.5.2产生误差的原因9
2.6总结.10
3BP神经网络电力负荷预测11
3.1人工神经网络概述.11
3.1.1人工神经网络历史和发展11
3.1.2人工神经网络的结构11
3.2BP神经网络及其学习算法13
3.2.1BP神经网络的产生13
3.2.2BP神经网络结构及算法实现.13
3.3BP神经网络在负荷预测中的存在的问题15
3.4总结.16
4基于遗传算法的BP神经网络电力负荷预测17 基于智能算法的电力负荷预测研究:http://www.youerw.com/zidonghua/lunwen_36599.html