应用分形两种迭代算法作短期负荷预测

摘要

电力负荷受很多因素的影响，故电力负荷的预测也存在一定的随机性，而分形理论是非线性系统理论中十分活跃的一个分支，其中分形插值是构造分形曲线的方法。本文在分形拼贴原理和分形插值方法的基础上，结合遗传算法对垂直比例因子随机寻优的特点，分别利用确定性迭代算法和随机迭代算法来求取电力负荷预测日的迭代函数吸引子，从而构造两种预测模型，最后通过实例对两种模型下的预测结果进行比较。%Electric power load is affected by a lot of factors,so the electric power load forecasting has certain randomness.However fractal theory is a nonlinear system,and the theory is a very active branch,of which fractal interpolation is tectonic fractal method of curve.Based on the fractal collage principle and fractal interpolation method,and basedon the genetic algorithm combining with the vertical scaling factor of random optimization,the characteristics of the certainty iterative algorithm respectively and random iterative algorithm calculate the iterated function attractor of power load forecasting,and finally compare the predicted results of the two models with the examples.