基于AHP-GCA及多元线性回归模型的电压合格率预测

为实现基于小样本数据的电压合格率准确预测,对电压合格率的众多影响因素进行分析并按静态影响因素与动态影响因素进行分类。鉴于静态影响因素对电压合格率的决定作用,采用一种基于层次分析法加权改进的灰关联分析方法,对影响电压合格率的静态影响因素进行排序,遴选出了对电压合格率的主要影响因素。由此建立电压合格率与主要影响因素之间的多元线性回归预测模型,进行电压合格率的预测。算例结果表明,提出的预测算法具有较高的预测精度,回归效果显著,对配电网的规划具有较大的指导意义。

The Prediction Method of Voltage Qualification Rate Based on AHP-GCA and Multiple Linear Regression

To achieve accurate prediction of voltage qualification rate. based on small sample of data the influencing factors of voltage qualification rate were analyzed and classified according to static and dynamic factors. Considered the static factors play a decisive role for voltage qualification rate, an improved gray correlation analysis method based on AHP was used to sort static factors that affect the voltage qualification rate, the main factors of voltage qualification rate were selected. Thus, a multiple linear regression model was established between voltage qualification rate and the main factors, and the voltage qualification rate prediction was carried out. Numerical re- suits show that the prediction algorithm proposed has higher prediction accuracy, and the regression result is obvious. The method provides some guidance to actual distribution network planning.