【作者】 舒菲；

【导师】 余健明；

【作者基本信息】 西安理工大学 ， 电力系统及其自动化， 2008， 硕士

【摘要】 短期负荷预测是电力系统安全调度、经济运行的重要依据,负荷预测的精度直接影响到电力系统运行的可靠性、经济性和供电质量。因此,寻求合适的负荷预测方法最大限度的提高预测精度具有重要的应用价值。根据电力负荷特性的变化规律,考虑了日期类型、温度、天气状况等影响负荷预测的因素,本文提出了一种将径向基函数(Radial Basis Function,RBF)神经网络与模糊理论相结合的短期负荷预测的方法。首先,考虑负荷的季节性变化,对春、夏、秋、冬四季分别建立预测模型,采用模糊聚类分析的方法对负荷预测相关数据进行聚类,选用同类特征数据作为RBF神经网络的输入,对神经网络进行训练,从而实现电力系统短期负荷预测;其次,采用在线自调整因子的模糊控制对预测误差进行在线智能修正使预测模型适应负荷的实时变化;最后,在未来电力市场的环境下,电价因素也是一个必须在负荷预测模型中加以考虑的因素,为了克服神经网络在电力市场下进行负荷预测时存在的不足,本文利用RBF神经网络的非线性逼近能力对不考虑电价因素的预测日负荷进行预测,然后根据近期实时电价的变化,应用模糊控制对RBF神经网络的负荷预测结果进行修正。实际算例表明,本文的预测方法收敛性好、方便实用、有较高的预测精度和较快的训练速度。更多还原

【Abstract】 Short-term load forecasting is important basis of safely assigning and economically running. The forecasting precision will directly affect the reliability, economy running and supplying power quality of power system. So finding an appropriate load forecasting method to improve the accuracy of precision has important application value.According to the rule of change of load characteristic, after calculating the factors such as date type, temperature, weather status etc which influencing the load forecasting a forecasting method based on radial basis function neural networks and fuzzy theory. At first, considering load seasonal variation, forecasting models are established to forecast each season including spring, summer, autumn and winter respectively, using fuzzy clustering analysis method, the clustering analysis of the related data of load forecasting are carried out, the data of the same property are used as the input of neural network, train RBF neural network to predict short-term load. Secondly, in order to eliminate forecast error, on-line self-tuning fuzzy control is used, make prediction model adapt to real-time change. Finally, Electricity price is also factor which must be considered in load forecasting model in future power market environment. in order to overcome the defect of the RBF network in power market environment, the article first draws on the nonlinear approaching capacity of the RBF network to forecast the load on the prediction day which takes no account of the factor of electric price, and then, based on the recent changes of real-time price, uses the fuzzy control system to modify the results of load forecasting obtained by using the RBF network.Practical examples indicate that the forecasting method is convenient and practical, possesses a good convergence, greater forecasting accuracy and faster training speed.更多还原