针对电能计量设备中高速电力线载波通信(high speed power line carrier communication,HPLC)单元现场的运行质量评估需求,提出一种基于核独立分量分析-部件关键度(kernel independent component analysis-component importance measure,KICA-CIM)多模型集成的评估方法.首先,整合分析典型客户侧计量设备本地物联应用场景、通信技术的主要性能影响因素,综合构建物联场景性能评价的普适性指标库;其次,面向运行场景中的多源、异构、高维数据环境,一方面利用KICA处理非线性特征并求解主成分,计算各指标的个体权重,另一方面利用CIM模型区别衡量各指标评价结果对整体评估效果的不同影响程度,对指标进行重要性赋权,确定每个指标的作用权重.通过组合赋权模型实施整合,实现运行质量的全面评估;最后,利用某地区电能计量设备数据,验证所提出方法的可行性和有效性,有利于改善评估结果的准确性与合理性.
Abstract
To address the operational quality assessment requirements of the high speed power line carrier communication(HPLC)unit in energy metering equipment,a multi-model integration-based assessment method utilizing KICA-CIM is proposed.Firstly,the main performance influencing factors of local internet of things(IoT)application scenarios and communication technologies of typical customer-side metering equipment are integrated and analyzed,and a universally applicable index library for performance evaluation of IoT scenarios is comprehensively constructed.Next,in the operational scenario with multi-source,heterogeneous,and high-dimensional data environment,on the one hand,kernel independent component analysis(KICA)is used to process nonlinear features and solve principal components,calculating individual weights of each indicator.On the other hand,component importance measure(CIM)model is utilized to distinguish and measure the different impact levels of each indicator evaluation result on the overall evaluation effect,assigning importance weights to the indicators to determine the functional weight of each indicator.Through the implementation of integrated weighting models,comprehensive assessment of operational quality is achieved.Finally,the feasibility and effectiveness of the proposed method are verified using data from energy metering equipment in a certain region,which contributes to improving the accuracy and rationality of the assessment results.
万方数据