Univariate versus Multivariate Models for Short-term Electricity Load Forecasting
Univariate versus Multivariate Models for Short-term Electricity Load Forecasting
复制标题
短期电力负荷预测的单变量与多变量模型
DOI：
--
复制
发表时间：
2015
期刊：
影响因子：
--
通讯作者：
H. S. Hippert
中科院分区：
文献类型：
--
作者：
Guilherme Guilhermino Neto;Samuel Belini Defilippo;H. S. Hippert
为了有效地管理电力系统的需求，需要在线的短期负荷预测。为了模拟负荷，发展了单变量和多变量预测方法：第一种方法将负荷视为其时间序列的线性函数，另一种方法考虑了与天气有关的变量(主要是气温)的非线性影响。尽管最近对多变量模型有了广泛的研究，但一些作者认为单变量模型在短期内是足够的，他们认为包括温度变量不一定会增加模型的复杂性，使简约性和稳健性处于危险之中。在这项研究中，我们比较了几个单变量和多变量时间序列以及基于神经网络的负荷曲线模型对实际数据的预测结果。然后，我们使用非参数假设检验来比较每种最佳预报员的日平均误差，从而验证考虑气温是否会在预测中带来任何统计上的显著改善。
Online short-term load forecasts are needed for efficient demand man- agement on power systems. To model the load, univariate and multivariate fore- cast approaches were developed: while the first consider the load as a linear func- tion of its time series, the other also takes in account the nonlinear effects of weather-related variables (mainly the air temperature). Despite the wide recent li terature on multivariate models, some authors state that univariate ones are suf- ficient for short-term purposes, claiming that including temperature variables un- necessarily elevates the model complexity, putting parsimony and robustness at risk. In this study, we compare the forecasts produced, for real data, by several univariate and multivariate time series and neural network-based load curve mod- els. We then use a nonparametric hypothesis test to compare the daily mean errors of the best forecaster of each kind and, so, verify if considering the air tempera- ture leads to any statistically significant improvement in the forecasting.