Strategic observation of power grids for reliable monitoring

In this paper, strategic observation of power grids refers to the optimized placement of measurements necessary to obtain the most likely operating condition of the grid, achieved in real-time monitoring via the state estimation (SE) functionality. The measuring system designed for SE considers the types, placement at various grid buses, and the total available measurements, which is crucial for reliable monitoring. The objective to be reached is to find the best places to install additional measuring units, in a limited number divided into pre-specified lots, to reduce data criticalities-whose importance is related to impending unobservability and incapacity of SE residual analysis to handle bad data. The process of finding an optimized solution for the meter placement problem is typically combinatorial and of considerable size, which makes metaheuristics the realistic option to solve the issue at hand. Metaheuristics are approximate optimization techniques with a pervasive success story in tackling various complex, real-world combinatorial optimization problems. This paper proposes the combined implementation of greedy randomized adaptive search procedure (GRASP), variable neighborhood descent (VND), and variable neighborhood search (VNS) optimization techniques. Numerical results performed on benchmark power grids (IEEE 30- and 118-bus) illustrate the application of the proposed approach and evince that high-quality solutions are achieved (validated by the exhaustive search results) in a computing time adequate for power grid operation planning studies.

Automated summary

This paper proposes an optimized placement method for measurements in power grids to achieve strategic observation, using metaheuristics to find high-quality solutions. The results show that the proposed approach achieves high-quality solutions in a short computing time for power grid operation planning studies.