Optimal micro-PMU placement and virtualization for distribution network changing topologies. (September 2021)

• Record Type: Journal Article
• Authors: Ibarra, L.; Avilés, J.; Guillen, D.; Mayo-Maldonado, J.C.; Valdez-Resendiz, J.E.; Ponce, P.
• Is Part Of: Sustainable energy, grids and networks. Volume 27(2021)
• Publication Date: 2021-09
• Subjects: Distribution network -- Grid reconfiguration -- Optimal placement -- Monitoring -- Genetic algorithms; Renewable energy sources -- Periodicals; Smart power grids -- Periodicals; Electric power systems -- Periodicals; 333.794
• DOI: 10.1016/j.segan.2021.100510
• Languages: English
• ISSNs: 2352-4677

Abstract

Abstract: Distribution networks are currently the main focus of modernization in electric power systems. For instance, to deal with emerging and increasingly challenging scenarios (e.g., high penetration of renewables, distributed generation, electric vehicle proliferation, etc.), new smart grid technologies are being deployed over distribution networks. This is the case of micro phasor measurement units ( μ PMUs), which can be seen as a modern high-precision version of the legacy transmission system PMU. Motivated by the benefits, as well as the costs associated with μ PMU deployment, in this work we introduce a new methodology to determine optimal μ PMU placement, considering grid topology changes. Firstly, a technique to minimize power losses across the grid by changing its topology is presented. Afterward, a technique to estimate the grid states – full numerical observability – with the minimum number of μ PMUs is disclosed, introducing the concept of bus "virtualization" through pseudo-measurements. This work is based on a multi-objective approach, acting on genetic algorithms, and validated over a distribution network with a variant structure, which is optimally instrumented. It is shown that the proposed methodology can maintain low approximation errors despite topology variations. Also, the virtualization approach enables full numerical observability even when using fewer measurement units, i.e., the observability constraint is eliminated and the number of μ PMUs can be solely related to the estimation error.