基于SARIMA-GRNN-SVM的短期商业电力负荷组合预测方法

针对短期商业电力负荷预测准确性与周期难以满足现有电力现货市场的问题,提出了一种基于SARIMAGRNN-SVM(seasonal autoregressive integrated moving average-generalized regression neural network-support vector machine)的商业电力负荷组合预测模型。首先,对商业电力负荷变化的周期规律与随机因素的复杂影响进行了分析;然后,结合以上分析,选用SARIMA和GRNN为单一预测模型对商业电力负荷进行预测,并利用SVM进行组合,实现日前商业电力负荷预测;最后,通过某商业综合体的电力负荷数据进行验证。所提组合预测模型较单一预测模型拥有更优的预测精度与鲁棒性,可以为短期商业电力负荷预测提供借鉴。

Short-term Combined Commercial Load Forecasting Method Based on SARIMA-GRNN-SVM

Considering that the accuracy and cycles of short-term commercial load forecasting are difficult to meet the requirements from an electricity spot market,a combined commercial load forecasting model based on seasonal autore?gressive integrated moving average-generalized regression neural network-support vector machine(SARIMA-GRNNSVM)is proposed.First,the periodicity of commercial load changes and the complex effects of random factors are ana?lyzed.Then,based on the above analysis,SARIMA and GRNN are selected as single predictive models,which are fur?ther combined with SVM to realize the commercial load forecasting.Finally,through the verification of the load data from one commercial complex,it is indicated that the proposed combined predictive model has higher prediction accura?cy and robustness than single predictive models,which can provide reference for short-term commercial load forecast?ing.