Fitting and forecasting of power load are difficult problems involving multiple aspects and factors about sources, grids and loads in power systems. Due to their key roles in power generation planning and distributing, fitting and forecasting are the essential topics in electrical engineering. In this study, fractal concepts and theories are reviewed briefly around fractal interpolation for function fitting; then fractal characteristics of power load are discussed, based on which their fitting and forecasting are improved. More precisely, a fractal interpolation approach with modified vertical scaling factors is developed for power load fitting that can accommodate for more time‐varying details. Also, short‐term power load forecasting under composite modified fractal interpolation with more parametric degrees of freedom for optimization are suggested and validated, whose performance and efficiency are compared with respect to some existing methods via numerical examples.

中文翻译：

---

基于复合修正分形插值方法的短期电力负荷拟合/预测

电力负荷的拟合和预测是棘手的问题，涉及电力系统中源，电网和负荷的多个方面和因素。由于它们在发电计划和配电中的关键作用，因此拟合和预测是电气工程中的基本主题。在这项研究中，围绕分形插值以进行函数拟合简要介绍了分形的概念和理论。然后讨论了电力负荷的分形特征，并在此基础上改进了它们的拟合和预测。更准确地说，针对功率负载拟合，开发了一种具有改进的垂直比例因子的分形插值方法，该方法可以适应更多随时间变化的细节。也，