采用线性动态聚类的谐波责任区间估计

Harmonic Responsibility Interval Estimation Based on Linear Dynamic Clustering

传统的谐波责任分摊方法需要采用专门设备同步测量谐波电压和多个用户的谐波电流,工程应用受到限制。该文采用谐波监测装置的非同步量测数据,提出一种基于区间型监测数据的多谐波源谐波责任分摊方法。首先,根据现有谐波监测数据构建区间监测样本,通过基于参数化回归的线性动态聚类算法划分谐波源的运行场景,以保证每一场景下的谐波阻抗与背景谐波电压的波动均维持在较小范围内;其后,构建区间型谐波责任量化方程,并利用相关性数据选择方法辨识区间参数;最后,提出一种谐波责任区间指标,定量评估谐波源的谐波责任。通过仿真与算例验证,验证了该文方法的可行性及有效性。

The conventional harmonic responsibility determination method needs the specialized equipment to measure harmonic voltages and currents of users synchronously,therefore its engineering applications are limited.In this paper,a method for the responsibility determination of multiple harmonic sources based on interval-type monitoring data,which is the asynchronous measurement data from harmonic monitoring devices,is proposed.First,interval monitoring samples are constructed from the existing harmonic monitoring data.Then,the scenarios are divided by the linear dynamic clustering algorithm based on the parameterized method to ensure that the harmonic impedance variation and the background harmonic voltage fluctuation are kept in a small range in each scenario.Subsequently,the interval linear equations for quantification of harmonic responsibility are established and the interval parameters are identified by the correlation data selection method.Finally,a harmonic responsibility interval index is proposed to evaluate the harmonic responsibility of harmonic sources quantitatively.The feasibility and effectiveness of the proposed method are verified by simulation and arithmetic cases.