基于聚类分析的电网综合能效诊断模型及应用

Comprehensive energy efficiency diagnosis model and application of power grid based on cluster analysis

针对电网综合能效的残差估计效果较差和诊断耗时较高的问题,基于聚类分析方法,进行电网综合能效诊断模型及应用研究。充分考虑电网特点,建立电网综合能效诊断指标体系,通过极值化无量纲归一化处理定量指标,获取所确定指标体系内各指标权重;依据灰色变权聚类方法确定各项指标权重确定变权重系数,明确与灰类同属的不同对象位次,根据位次排序结果诊断电网综合能效。实验结果表明,诊断精度高达99.5%以上,绝对误差、正误差以及负误差均较低,残差估计效果得到提高,在样本覆盖范围较大的情况下诊断耗时低于50ms。

Regarding the problems of poor residual estimation and high time-consuming diagnosis of comprehensive energy efficiency of power grid,the diagnosis model and application of comprehensive energy efficiency of power grid are studied based on cluster analysis method.Considering the characteristics of power grid,the comprehensive energy efficiency diagnosis index system of power grid is established,which obtain the weight of each index in the index system by extreme value dimensionless normalization;the weight of each index is determined according to the grey variable weight clustering method,and the variable weight coefficient is determined,and the rank of different objects belonging to the same gray class is determined,and the comprehensive energy efficiency of power grid is diagnosed according to the ranking results.The experiment results show that the diagnostic accuracy is up to 99.5%while keeping low the absolute error,positive error and negative error,and the residual estimation effect is improved,and the diagnosis time is less than 50ms when the sample coverage is large.