CAJ | 학술논문

在我国新型城镇化和美丽乡村宏观政策的推动下，城乡一体化进程加快，带动农网用电需求快速增长。分析新型城镇化和美丽乡村背景下的农网用电需求，对建立和完善乡镇电网发展模式和建设标准具有重要现实意义。结合新型城镇化和美丽乡村背景下的农网用电需求特点，建立了城镇用电需求评估指标体系，提出一种基于全样本空间的类比预测法，将用电需求预测从单一维度扩展到多维空间，适用于目前城镇化过程中电力数据和经济社会发展信息交汇的大数据环境。在此基础上，综合回归分析法、灰色模型、人均用电量法等3种经典预测方法，设计了基于拟合优度赋权的组合预测算法，实现了权重的自动优化调整，算例结果表明组合预测算法提高了预测的精度和可靠性。
재아국신형성진화화미려향촌굉관정책적추동하，성향일체화진정가쾌，대동농망용전수구쾌속증장。분석신형성진화화미려향촌배경하적농망용전수구，대건립화완선향진전망발전모식화건설표준구유중요현실의의。결합신형성진화화미려향촌배경하적농망용전수구특점，건립료성진용전수구평고지표체계，제출일충기우전양본공간적류비예측법，장용전수구예측종단일유도확전도다유공간，괄용우목전성진화과정중전력수거화경제사회발전신식교회적대수거배경。재차기출상，종합회귀분석법、회색모형、인균용전량법등3충경전예측방법，설계료기우의합우도부권적조합예측산법，실현료권중적자동우화조정，산례결과표명조합예측산법제고료예측적정도화가고성。
With the promotion of the new urbanization and beautiful countryside macro policy in China, the process of the urban￣rural integration accelerates, which drives the fast increase in the demand of the rural power network. In the background of the new urbanization and beautiful countryside, the analysis of the demand of the rural power network has the important practical significance to the establishment and improvement of the development model and construction standard of rural power network. Combining the power demand characteristics of rural power network in the background of the new urbanization and beautiful countryside, the power demand evaluation indexes for town are established and the analogy forecasting method is proposed based on full sample space, which could extend the load forecasting from a single dimension to hyperspace. It is applicable to the big data environment where the power data intersected the economic and social development in the process of urbanization. On this basis, this paper synthesizes the three classic forecasting methods, including the regression analysis method, the grey model and the per capita consumption method, and proposes a combination forecasting method based on the goodness of fit empowerment, which could realize the automatic optimization adjustment of weight. The numerical example results show that the combination forecasting method can improve the accuracy and reliability of the forecasting results.