Modified branch‐and‐bound algorithm for unravelling optimal PMU placement problem for power grid observability: A comparative analysis. Issue 4 (14th April 2021)

• Record Type: Journal Article
• Authors: Babu, Rohit; Raj, Saurav; Dey, Bishwajit; Bhattacharyya, Biplab
• Is Part Of: CAAI transactions on intelligence technology. Volume 6:Issue 4(2021)
• Page Start: 450
• Page End: 470
• Publication Date: 2021-04-14
• Subjects: integer programming -- tree searching -- phasor measurement -- graph theory -- linear programming -- approximation theory -- power grids -- simulated annealing -- search problems -- recursive estimation -- power engineering computing -- computerised instrumentation -- Matlab; Artificial intelligence -- Periodicals; Computer science -- Periodicals; Artificial intelligence; Computer science; Electronic journals; Periodicals; 006.305
• DOI: 10.1049/cit2.12038
• Languages: English
• ISSNs: 2468-6557

Abstract

Abstract: Safe operation of the power grid requires a complete and robust control network to ensure full observability. However, redundancy measurements can create problems in expense, management, and control. The optimal phasor measurement unit (PMU) positioning problem (OPPP) is proposed to limit the number of PMUs deployed in the power grid and ensure the whole grid's observability in the meantime. A modified branch‐and‐bound algorithm (MBBA) to unravel the OPPP is presented. Original BBA, which uses a single search order to create a binary tree, gives only one solution to the OPPP, although more than one optimal solution exists. The proposed MBBA method consists of two different stages: the vertexes in the search tree are investigated by depth‐first search (DFS) in stage 1, and the search route continues as the breadth‐first search. In stage 1, the LP relaxing problems are solved by dual simplex, and in stage 2, the basic viable solution from stage 1 is used to configure the primary simplex until the optimum solution is found. OPPP is formulated as a binary decision variable MBBA model, minimizing linear objective function subject to linear matrix observability constraints. The MBBA model is unravelled using a linear integer‐based external approximation scheme. IEEE test systems are used to check the feasibility of the proposed approach. Matlab software performs simulation based on a number of graph theory‐based methods such as DFS, graph‐theoretical method, simulated annealing, and recursive N‐algorithms. These algorithms are compared to the algorithmic perspective of the proposed MBBA method. IEEE test network results confirm the validity of the proposed methodology.