基于缓冲算子和时间响应函数优化灰色模型的中长期负荷预测

摘要

针对中长期电力负荷预测,对传统灰色预测模型进行改进.一方面,对原始数据序列进行平均弱化缓冲算子作用,有效地消除了冲击扰动系统的干扰;另一方面,提出了时间响应函数的优化方法,即利用原始序列与其模拟值的欧氏距离最小为约束条件,来确定常数C,使原始信息得到更好的利用.将改进的模型应用于某地区年负荷预测中,算例结果表明,改进的模型与传统模型相比具有较高的预测精度,更适用于长期负荷预测.%This paper proposes an improved grey forecasting model for long and medium term power load forecasting. On the one hand, the original data sequences arc processed by average weakening buffer operator, which can effectively eliminate the interference of shock disturbed system. On the other hand, a method of time response function optimization is proposed, namely using constraint condition of minimum Euclidean distance of original sequence and simulation value to determine the constant C and then make better use of the original message. The improved model is applied to forecast the annual power load in some areas. The results show that improved model has higher prediction accuracy than the traditional model and is more suited for long-term load forecast.