摘要
中期电力负荷受到温度因素影响,其波动性与随机性强,导致负荷概率密度预测精度低。为此,提出了基于EMD-QRF的温度因素下中期电力负荷概率密度预测方法。采用EMD分解温度因素下中期电力负荷概率密度特征,提取波形平均周期和振幅。构建QRF概率密度预测模型,结合神经网络,避免预测过程陷入过度拟合。通过对参数离散化处理,构建中期电力负荷概率密度预测函数,实现不同波形周期和振幅下的中期电力负荷概率密度预测。实验结果可知,该方法在夏季概率密度值与实际值存在0.01 kW/km^(2)的误差,在冬季密度值与实际值一致,能够有效提高中期电力负荷概率密度预测精度。
The mid-term power load is influenced by temperature factors,which have strong volatility and randomness,leading to low accuracy in load probability density prediction.Therefore,a probability density prediction method for mid-term power load under temperature factors based on EMD-QRF is proposed.Using EMD to decompose the probability density characteristics of mid-term power load under temperature factors,extract the average period and amplitude of the waveform.Construct a QRF probability density prediction model,combined with neural networks,to avoid overfitting in the prediction process.By discretization of parameters,a med-term power load probability density prediction function is constructed to realize probability density prediction of medium term power load under different waveform periods and amplitudes.The experimental results show that this method has an error of 0.01 kW/km^(2) between the probability density value in summer and the actual value,and the density value in winter is consistent with the actual value,which can effectively improve the accuracy of medium term power load probability density prediction.
作者
马冬冬
施佳峰
徐鹤勇
MA Dongdong;SHI Jiafeng;XU Heyong(State Grid Ningxia Electric Power Co.,Ltd.,Yinchuan 750001,China)
出处
《电子设计工程》
2024年第15期90-94,共5页
Electronic Design Engineering
基金
福建省教育厅科技项目(JA13233)
厦门市重大科技项目(3502Z20111008)。
