【作者】 汪新秀；

【导师】 吴耀武；

【作者基本信息】 华中科技大学 ， 电力系统及其自动化， 2004， 硕士

【摘要】 电力系统短期负荷预测是指以周、天、小时为单位的负荷预测,通常预测未来一天24 小时的负荷,它是制定发电计划和输电方案的主要依据,对合理安排机组启停、确定燃料供应计划、进行能量交易等具有重要意义,其预测精度的高低直接影响到电力系统运行的安全性、经济性。随着电力系统市场化的不断深入,短期负荷预测在电力系统中显得更加重要。 本文首先概述了电力系统短期负荷预测的原理、研究现状及发展趋势,对电力系统短期负荷预测的各种传统及现代方法进行了综述,并着重分析了各种方法的特点及适用范围。在对人工神经网络、模糊集理论和小波分析理论进行重点研究的基础上,针对电力系统短期负荷预测的特点,本文提出了基于小波网络和模糊神经网络的模糊小波网络负荷预测模型,该模型综合了小波变换良好的时频局域化性质、模糊推理和神经网络的学习能力,大大提高了网络的泛化能力。根据本文提出的负荷预测模型,编制了相应的软件,对两个不同规模的实际电力系统进行了负荷预测。结果表明该模型应用于电力系统短期负荷预测是可行而实用的,与人工神经网络预测模型相比,具有更快的收敛速度,更高的预测精度,能更好地预测波动较大的电网负荷。本文最后分析了电力系统短期负荷预测研究中存在的问题,并对进一步研究作了展望。更多还原

【Abstract】 The short-term load forecasting of electric power system is to predict electric load for aperiod of hours, days, or weeks, and especially twenty-four hours, which is the primary gistfor making the plan of power generation and the scheme of power transmission. It isimportant for economic arrangement of generating capacity, scheduling of fuel purchases andplanning for energy transaction. Its precision will greatly influence the economy and secureoperation of power system. Furthermore, with the establishment of power market, loadforecasting will play a more important role in the future. The principle, current status and development of the electric power system short-termload forecasting are generalized in this thesis. Varieties of traditional and modern predictiontechniques for load forecasting are summarized, and the differences and features of thesemethods are also emphasized. Based on the research of artificial neural network (ANN),fuzzy set theory, and wavelet analysis theory, Owning to the traits of electric power systemshort-term load forecasting, a fuzzy wavelet network (FWN) model in load forecasting isproposed, which combines the time-frequency localization ability of wavelet, fuzzy inferringand the education character of ANN together, and improves the generalization capability. Thecorresponding program of this load forecasting model is developed and applied to two kindsof practical power systems. The results show it is credible and practicable in load forecasting.Compared with the forecasting result of artificial neural network model, this model can get afaster convergence and more precise result, especially in forecasting greatly fluctuant load. Atthe end of this paper, the main problems in the research of short-term load forecasting areanalyzed and the further work is pointed out as well.更多还原