基于VMD-NCWOA-LSSVM的短期电力负荷预测方法

准确的电力系统负荷预测对电力系统安全稳定运行具有重要意义,为提高负荷预测精度,采用变分模态分解(variational mode decomposition,VMD)预处理数据,将原始日负荷曲线分解为不同频率的子序列,降低数据不规律性对负荷预测带来的干扰。使用Piecewise模糊映射策略进行改进,解决鲸鱼优化算法(whale optimization algorithm,WOA)受初值影响容易陷入局部最优的问题。使用非线性收敛因子代替线性收敛因子,进一步提升WOA的全局寻优能力和局部探索能力,得到非线性收敛因子的混沌鲸鱼优化算法(nonlinear convergence factor of the chaotic whale optimization algorithm,NCWOA)优化最小二乘支持向量机(least square support vector machine,LSSVM)的组合预测模型(VMD-NCWOA-LSSVM)。测试结果表明本文所提模型可以降低预测值的最大相对误差和平均绝对百分误差,有效提高短期电力负荷预测的精度。

A short-term power load forecasting methodbased on VMD-NCWOA-LSSVM

Accurate power system load forecasting is of great significance to its safe and stable opera-tion. In order to improve the accuracy of load forecasting, V ariational mode decomposition( VMD) topreprocess the data was used , the original daily load curve was decomposed into the sub-sequence ofdifferent frequencies , and the interference of data irregularity on load forecasting was reduced. By useof Piecewise fuzzy mapping strategy, the problem that whale optimization algorithm ( WOA) was apt to fall into local optima under the influence of initial value was solved. Linear convergence factors werereplaced by nonlinear convergence factors to further improve WOA ''s global optimization capabilitiesand local search capabilities and obtain the combined prediction model ( VMD-NCWOA-LSSVM) of the nonlinear convergence factor of the chaotic whale optimization algorithm ( NCWOA) optimizing thefactor of the least square support vector machine ( LSSVM). The test results show that the proposedmodel can reduce the maximumrelative error and average absolute percentage error of predicted valueand can improve the accuracy of short-term power load forecasting.