摘要
电力系统日负荷曲线是一类非线性曲线,受天气、突发事件等敏感因素的影响,同一天的不同时刻,其曲线波动情况不相同,仅具有部分自仿射结构或不具有明显的自仿射结构。分形插值方法在解决非线性问题上具有很大的优势。考虑到天气因素对电力负荷的影响,先采用加权的灰色关联度方法选择相似日,从历史数据中找出与预测日具有相似日特征向量的负荷,然后针对电力系统日负荷曲线的特点,对日负荷曲线进行分段,最后采用分时段分形插值的方法对预测日的电力负荷曲线进行拟合。通过与整体分形插值进行对比发现,分时段分形插值方法更加准确有效。
The electric power daily load curve is a kind of nonlinear curve, and is influenced by weather and some unexpected events, the curve will be different in different time even in the same day, so it is only partly self-affine or is not obvious self-affine. The method of fractal interpolation has great advantages in solving the nonlinear issue. For the wether influence on power road, A method of weighted grey relational degree is adopted to select similar days and daily eigenvector similar to the forecasting day load are selected from the historical data. The daily load curve is divided according to the curve＇s characteristics. The time-segment fractal interpolation is used to fit the load curve. Compared with the holistic fractal interpolation, the time-segment fractal interpolation is more effective.
出处
《现代电力》
2009年第2期37-41,共5页
Modern Electric Power
关键词
负荷预测
分形插值
分时段
灰色关联度
相似日
load forecasting
fractal interpolation
time segment
grey relational degree
similar days
