In order to solve the problem of power consumption management for special transformer users,a piecewise linear model was proposed to forecast the monthly average electricity consumption ratio data of special transformer users and to realize the goal of saving the energy consumption. With the N point length data segmentation, the piecewise linear model was updated and generated, and the average electricity consumption ratio data for the next month were estimated. When the segmentation point length N was from 4 to 8,the forecast accuracy difference of piecewise linear model was analyzed and discussed. The actual power consumption data for a special transformer user from 2015 onwards were forecasted and analyzed, and the error between the forecast values and actual data was compared and verified. The results show that when the optimal modeling point length N is 6,the proposed model has the minimum average prediction error.%针对专变用户数量激增引起的用电配给管理问题,提出一种分段线性模型来预测专变用户月平均用电占比数据,实现节能配给的目标.采用N点长数据分段更新生成分段线性模型,从而估计得到下一个月的平均用电占比数值.讨论分析了分段点长N分别取4~8时,分段线性模型预测精度的差异.针对某专变用户2015年以来的实际用电数据进行预测分析,比较验证了预测数值和实际数据之间的误差.结果表明,当最优建模点长N=6时,模型的平均预测误差最小.
展开▼
