摘要
针对配电网空间负荷预测方法进行了研究,将总量负荷预测与空间负荷预测相结合。在总量负荷预测上,采用基于人工神经网络的组合预测方法,尽可能消除不确定性因素的影响。在空间负荷预测上,采用基于模糊贴近度的负荷密度法,建立了相应的负荷密度选取指标。根据总量负荷预测结果对空间负荷预测结果进行校核和修正,为规划区配电网的网架规划、变电布点、容量和建设时间提供了依据。
This paper focuses on the researeh of load forecast- ing in distribution network planning. One comprehensive approach based on artificial neural network is applied to forecast load by us- ing the forecasting results of various forecasting methods, in which the experience of the planners are fully drawn and the actual situa- tion of planning zone is frilly taken into account, so as to eliminate the influence of uncertainty factors on load forecasting results. The forecasting model and the zone partition method for spatial load folcasting, and the load characteristics of the regional power grid are analyzed in detail; the average classification load density approach and the based on fuzzy similarity- dgree load density approach is used respectively to build the corresponding load density selection index system aiming to old and new planning areas; the cheek and correction methods for the forecasting results of spatial load are dis- eussed. The results provide a planning foundation for determining high-voltage substation placement and capacity, planning distribu- tion network layout and timing distribution equipment commission.
出处
《电力需求侧管理》
2016年第5期21-24,共4页
Power Demand Side Management
关键词
配电网
负荷预测
空间负荷预测
人工神经网络
模糊贴近度
distribution network： load forecast： spatial loadforecast： artificial neural network
fuzzy similarity degree
