具有二重趋势性的季节型电力负荷预测组合优化灰色神经网络模型

COMBINED OPTIMUM GRAY NEURAL NETWORK MODEL OF THE SEASONAL POWER LOAD FORECASTING WITH THE DOUBLE TRENDS

电力负荷预测是电力系统的一项重要工作 ,季节型电力负荷预测是一个难点 ,缺少相应的数量预测方法。对于具有增长和波动二重趋势性的季节型电力负荷 ,首次提出了季节型负荷预测的组合优化灰色神经网络模型 ,研究了同时考虑两种非线性趋势的复杂季节型负荷预测问题 ,说明了此优基金项目 :国家自然科学基金资助项目 ( 5 0 0 770 0 7) ;国家电力公司重点学科基金资助项目 (A98B0 3)。ProjectSupportedbyNationalNaturalScienceFoundationofChina( 5 0 0 770 0 7) .化模型分别优于两种单一发展趋势负荷预测的模型。给出了电力负荷预测的应用实例 ,通过对河北电网季节最大负荷与销售电量的分析 ,建立了对应的组合优化灰色神经网络模型 ,与其它算法进行了比较 ,计算结果表明 ,该方法较大提高了季节型负荷预测的精度 ,为季节型电力负荷预测提供了一种新的、有效的方法 ,编制了季节型负荷预测的软件 。

The power load forecasting is an important work in power system.The seasonal power load forecasting is difficult for corresponding quantity forecasting methods are absent.For the seasonal power load forecasting with the double trends of increasing and fluctuating,it is proposed first for the combined optimum gray neural network model of seasonal load forecasting.We study the problem of complex seasonal load forecasting with double nonlinear trends.The optimum model is better than the tow load forecasting models with one development trend.An application case of the power load forecasting is given.Through the analysis to the seasonal greatest load and selling electric capacity in Hebei power system,the corresponding combined optimum gray neural network model is built.It is compared with other algorithms.The calculation results prove that this method raises the accuracy of the seasonal load forecasting greatly. We provide a new and effective method for the seasonal power load forecasting.The software of a seasonal load forecasting is made and it is practical and general.