基于环境特性网的负荷组合预测

Load combination forecasting based on power grid with environmental characteristics

在分析煤矿井下电力负荷特性的基础上选取单项预测模型,提出一种基于组合系数诱导IOWA算子的组合预测方法。该方法应用Dempster合成法则筛选精度较优的单项模型预测值替代历史数据,克服实际预测期数据未知的缺陷;基于发展相关性指标和精度指标构造递阶层次结构,运用熵权法求取两项模型评价指标的组合权重,求出组合系数;在预测过程中引入等维信息概念,使预测结果能够更合理地反映负荷发展规律。预测结果表明,基于组合系数诱导IOWA算子的模型适用于预测井下负荷,并能提高预测精度。

The characteristics of underground power load in coal mine were analyzed.Based on these characteristics,the appropriate individual forecasting models were selected.It was proposed a combination forecasting method based on combination index induced IOWA operator.In this method,forecasting values of individual model with higher accuracy were substituted for historical data by Dempster synthesize rule.The defect that data are unknown during actual forecasting period was overcome.The concept of equal dimension information was introduced in this model.The forecasting results can be reflected the development law of load by this concept reasonably.Combination forecasting model based on combination index induced IOWA operator is suitable for forecasting underground power load in coal mine.And the forecasting accuracy can be improved.