Sparse voltage amplitude measurement based fault location in large-scale photovoltaic power plants. (1st February 2018)

• Record Type: Journal Article
• Authors: Jia, Ke; Gu, Chenjie; Li, Lun; Xuan, Zhengwen; Bi, Tianshu; Thomas, David
• Is Part Of: Applied energy. Volume 211(2018)
• Page Start: 568
• Page End: 581
• Publication Date: 2018-02-01
• Subjects: Fault location -- Bayesian compressive sensing -- Large-scale PV power plant -- Sparse voltage measurement; Power (Mechanics) -- Periodicals; Energy conservation -- Periodicals; Energy conversion -- Periodicals; 621.042
• DOI: 10.1016/j.apenergy.2017.11.075
• Languages: English
• ISSNs: 0306-2619

Abstract

Highlights: A sparse measurement based fault location is proposed for large-scale photovoltaic (PV) power plant. The improved compressive sensing (CS) technique is empoyed in the negative sequence network. Data from the world's largest PV power plant is used to prove the proposed method. The improved method offers accurate fault location considering all the possible influence factors. Abstract: Large-scale photovoltaic (PV) power plants contain numerous transmission line branches and laterals inside. When a fault occurs conventional fault location methods face challenges due to the complex system structure and the diversity of PV inverter controls. Most of the published fault location methods cannot be directly used in the PV power plant due to the following issues: (1) Most of the fault location methods consider the PV inverter as a constant voltage source while the actual inverters have varied controls during faults. Without analysis of the unique fault transients of the PV, the fault location will suffer from errors. (2) In a complicated large-scale PV power plant with massive quantity of nodes, the synchronised measurements from all the nodes are almost impossible. A method with sparse un-synchronized measurements is required. Therefore, a new negative-sequence voltage amplitude sparse measurement based fault location method is proposed for unbalanced faults. The improved Bayesian compressive sensing algorithm is used to recover the sparse fault current vector and then determine the faulted node. Both the field testing and the simulation results indicate that the proposed method can locate the faulted nodes accurately and effectively without synchronizing measurement requirements from all the nodes. This method also presents a good performance for various unbalanced fault types, fault resistances, inverter controls and signal noise. All these factors make the propose method feasible for industrial applications.