基于神经网络的电力系统短期负荷预测方法研究电力系统及其自动化专业论文.docxVIP

山东科技大硕士学位论文摘要 山东科技大硕士学位论文 摘要 摘要 电力系统负荷预测就是考虑系统运行特性、自然条件、社会条件和地区经济发展状况等 重要因素，将历史负荷值经过一系列的数学计算，确定未来某特定时刻的负荷值，并保证预 测误差满足一定精度要求的技术。短期负荷预测是指提前1天至1周内的日负荷预测，其核心 问题是预测模型的建模问题。基于神经网络、专家系统、模糊理论等人工智能理论的非数学 模型法的提出，为解决负荷预测中复杂的建模问题提供了一种有效的方法和途径，推动了电 力负荷预测由简单地依靠计算预测模型向着依靠计算、分析、推理等多种接近于人脑逻辑思 维的智能化预测模型的方向发展。 本文采用神经网络对短期负荷预测问题进行研究，并对BP网络与Elma n网络进行了优 化，利用优化后的网络进行建模与预测方法研究。论文在预测模型建立后，利用用菏泽某地 区电网的历史负荷数据，对影响当地负荷特性的因素进行了分析。本文采用MATLAB软件进 行编程，建立网络模型并训练，实现仿真预测，并对不同的网络模式，相同网络不同输入节 点模式的预测误差进行分析比较。同时设计了电力短期负荷预测的用户界面(GUI)，使输入 输出更为直观和方便。通过对典型系统的测试分析，表明本文所设计的电力短期负荷预测模 型是有效可行的。 关键词：短期负荷预测；BP网络；ELMAN 网络；MATLAB；GUI Abstract Power system load forecast ing based on the case of a certain precision, consider the system operating character ist ics, natural conditions, socia l conditions, regiona l economic development and other important factors . It use s the histor ica l load value date which after a series of mathematica l calculations to decide the particular moment load value in the future. Short-term load forecast ing is the daily power load forecast ing which form 1 day or 1 week in advance. The key issues of power Load forecast ing is the technical problems and mathematica l modeling problems of the forecast ing. When the non-mathematica l model method that based on neura l networks, expert systems, artificia l intelligence and fuzzy theor y has been proposed, thus there have an effective method to solve the complex load forecast ing problem of modeling. This thesis studied the issue of short-term load forecast ing by neura l network, and it improved the BP network and the Elma n network forecast ing method based on the origina l. With consider ing the impact of temperature on the load, it analyses the factors on the loca l load character ist ics using histor ica l data of a regiona l power grid in Heze City. Programming by MATLAB software, it establishes networking model and achieve simula tion predict ion, and the forecasting error of different networks mode and input nodes mode were analy