Distribution network fault diagnosis method for the deep integration of cyber-physics

Abstract

• A distribution network fault location method for deep integration of cyber-physics is proposed. • The complexity of fault diagnosis is reduced by building a library of faulty suspect components. • The characteristic data collected by the D-PMUs is used and improves the speed of fault diagnosis. • Petri net model is used to accurately describe the process of fault occurrence and improves the accuracy. The deep integration of cyber-physics provides multi-source information data for the fault analysis of distribution networks, but with conventional methods, it is difficult to select and use these data reasonably. This paper proposes a fault diagnosis method for a distribution network based on a distributed phasor measurement unit (D-PMU) and a Petri net. The method introduces a fault suspect component library formed with coupling pre-judgment and a fault diagnosis model, which simplifies the process of network analysis and improves the speed of the fault diagnosis. The method makes full use of the time constraint relationship of the alarm information, correctly selects the action time point, and achieves an improvement of the fault diagnosis accuracy. In addition, considering the problem that the measurement equipment coverage is not comprehensive in the actual distribution network, the proposed method inputs the characteristic information collected by the D-PMU at the typical measurement point into the Petri net model, analyzes the fault of the distribution network, and avoids the problem of the scheduling being risky due to the result of the probabilistic form. Finally, this paper describes the setup of some experimental analyses to verify the effectiveness of the proposed method.