节点文献

母线负荷预测系统的研究与实现

Study and Realization of Bus Load Forecasting System

【作者】 王健

【导师】 罗滇生; 姚建刚;

【作者基本信息】 湖南大学 , 电力系统及其自动化, 2009, 硕士

【摘要】 “十一五”期间,国民经济继续保持平稳快速发展,面对电力负荷增长压力和电力体制改革的新形势,保证电网安全、稳定和经济运行成为电力系统的首要任务。负荷预测为合理安排电源和电网的建设进度提供宏观决策的依据,而母线负荷预测可用于电网动态状态估计、安全稳定分析、无功优化、厂站局部控制和电力系统运行方式预测等方面,为电网日前安全校核与节能发电调度工作提供了有力的技术支持,提高了电网调度的精细化、智能化水平。论文在对负荷预测基本概念介绍的基础上,对母线负荷的定义、构成、特点和预测方法等做了全面深入的阐述,并针对母线负荷的自身主要特点,即负荷变化的趋势不明显和负荷变化的模式易变,提出了一种基于粒子群优化BP神经网络的预测方法。该方法用粒子群优化算法来优化神经网络的初始权值,克服了BP神经网络易陷入局部极小值、收敛速度慢和泛化性差等缺点。应用该方法对湖南某变电站的母线负荷进行预测,结果表明该方法能提高母线负荷预测精度和改善预测的泛化性。论文以现已完成的省地一体化负荷预测与管理平台为背景,在对母线负荷预测理论的综合叙述和预测方法详细阐明的基础上,构筑了一套基于J2EE的母线负荷预测系统的设计方案,详细介绍了系统的核心模块功能和实现方法,解决了当前母线负荷预测应用软件功能单一、可操作性不强、平台封闭、计算性能低和缺乏规范化的管理功能等问题。该系统的成功开发,对实现母线负荷预测具有重要的实用价值,将成为调度部门进行相关工作的强有力的辅助工具。

【Abstract】 In the "11th Five-Year" period, the national economy continues to develop stably and rapidly. With the pressure of electrical load growth and new situation of the power system reform, it is top priority of power system to ensure power grid operates securely, stably and economically. Load forecasting provides a basis for macroeconomic policy-making in arranging source reasonably and the progress of the power grid construction. Bus load forecasting can be used for state estimation, static security analysis, reactive optimization, local controlling of power plant, and prediction of power system operation, etc. It provides a strong technical support for day-ahead safety proofing and energy-saving power generation and improves the fine and intelligence level of grid dispatching.This article describes definitions, composition, characteristics and prediction methods of bus load on the basis of the concept of load forecasting comprehensively and deeply. According to major features of bus load that changes in the trend of the load is not obvious and the pattern is volatile, it provides a forecasting method based on PSO to optimize BP ANN. The method uses PSO algorithm to optimize connected right value of ANN. It can overcome some disadvantages in BP ANN such as running into local minimum value easily, slow convergence rate and so on. This paper applies the method to forecast bus load of a substation in Hunan, the results indicate that this method can improve the accuracy of the bus load forecasting and generalization quality.This paper is set in the load forecasting and management platform of province and local area all-in-one. On the basis of comprehensive description of bus load forecasting and elaboration of forecasting methods, it builds a set of bus load forecasting design based on J2EE. This paper introduces core module and implementation method of the system in detail, solves some problems of current bus load forecasting such as single function, weak operability, closed platform, low capacity and lack of standardization of the management functions and so on. With the successful implementation of this system, it has important practical value to realize bus load forecasting, and will become a powerful tool for dispatching departments to carry though correlation work.

【关键词】 电力系统母线负荷粒子群算法神经网络J2EE
【Key words】 Power SystemBus LoadPSOANNJ2EE
  • 【网络出版投稿人】 湖南大学
  • 【网络出版年期】2010年 01期
节点文献中: 

本文链接的文献网络图示:

本文的引文网络