PQ decoupled 3-phase numerical observability analysis and critical data identification for distribution systems

Emerging active distribution systems are operated under more complicated and uncertain conditions. Because of frequent topology changes or temporary malfunctions of data acquisition, the network may lose observability and in turn may lead to the failure of distribution system state estimation. Without distribution system state estimation, the distribution system operator may not be able to make secure decisions, and the system may face with risk of serious damages. In this paper, a new real/reactive (PQ) decoupled 3-phase numerical observability analysis method and a new critical data identification method for distribution systems are proposed. The methods can efficiently identify all unobservable branches and critical data in a non-iterative manner. In addition to the 3-phase decoupling, the normal equation is also PQ decoupled, and the computation burden is greatly reduced. Furthermore, all entries in the Jacobian matrix are non-negative integers and the proposed methods present good numerical performance. Strict theoretical proofs on the performance of the proposed methods are provided and numerical results on different scales of test systems illustrate their validities.