Search

With the rapid development of renewable energy resources, harmonics in modern distribution systems have become more serious owing to the wide adoption of power electronic devices. Today, distribution-level phasor measurement units (D-PMUs), which can provide synchronized harmonic data with high-time resolution, are being rapidly developed. To utilize the D-PMUs, a hierarchical harmonic contribution (HC) determination method is proposed in this paper for systems with multiple harmonic sources. The HCs were determined using a widely linear complex partial least squares (WL-CPLS) method, which is both accurate and efficient. In an actual power system, the D-PMUs are not fully installed at all buses and the harmonics monitored by each D-PMU are actually the aggregate effects of multiple harmonic sources in the subarea. Based on the above consideration, a hierarchical HC strategy is proposed to locate the main harmonic contributors, layer by layer. The proposed method can be applied to a radial distribution system with any number of D-PMU placements. The methodologies are explained and verified using the IEEE 37-bus system. The real case study results based on the demonstration project are also analyzed in this paper.

Due to the access of a large number of distributed energy sources and power electronic equipment, the power quality problems caused by harmonics have attracted widespread attention. Under the premise of grasping power quality monitoring data, it is urgent to determine the source of harmonics, in order to take targeted measures at the source and improve the power quality problems that occur in the system. This paper proposes a method to determine the harmonic contribution of each harmonic source in a multi-harmonic source system based on data selection method considering regression coefficient fluctuation. Firstly, an index factor for evaluating the harmonic contribution based on partial least squares of each harmonic source is established. Then the data selection method based on coefficient fluctuations in the regression model is used to select valid data segments. Finally, by analyzing the harmonic voltage and harmonic current data, the quantitative magnitude of the harmonic responsibility of each harmonic source is determined. The proposed method is convenient in calculation and strong in engineering practice. Simulation analysis and actual case verifies the effectiveness of the proposed method.