Impact of Measurements Placement on Fault Location Using Compressive Sensing

Fault location method based on compressive sensing (CS) is quite suitable to solve the difficulty of sparse measurement in distribution network. This paper analyzes the influence of the measurement units' placement based on the foundation of constructing impedance matrix by adding branches, combined with restricted isometry property (RIP) rule in compressive sensing. The fundamental methodology of fault section location by solving the sparse virtual injection zero sequence current vector is illustrated. The theory of CS is introduced as a recover algorithm. In addition, the measurement units' placements are discussed. The simulation results are implemented on a 12.66 kV, 8 buses distribution network, which demonstrates the validity of fault location by the means of CS. It also proves that when there are measurement units at the beginning and end of the grid, the measurement matrix obeys the RIP, and the fault can be located exactly.