Metering system planning for state estimation via evolutionary algorithm and HΔ matrix

This paper proposes a methodology for metering system planning for state estimation purposes. The proposed methodology is based on Evolutionary Algorithms (EAs) and on the analysis of the called HΔ matrix. By analyzing the structure of this matrix, which is obtained via a triangular factorization of the Jacobian matrix, the proposed methodology can determine reliable metering systems (RMS). In this paper a metering system is considered as reliable if it is observable and has no critical measurements; critical sets neither critical Remote Terminal Units (RTU). An EA is developed to find the best RMS with minimal investment cost. The developed EA uses a fitness function that measures the installation cost of meters and RTUs from a given RMS. One relevant advantage of the proposed methodology is its strategy to obtain RMS. An indirect chromosome encoding representing a preferential order of meters installation combined with properties of the HΔ matrix guarantees the proposed EA generates only feasible solutions, i.e. RMS. The proposed methodology was tested in the IEEE 6, 14, 30 and 118 bus systems, as well as in a real Brazilian system, providing to be reliable. The results obtained by the proposed methodology were compared with those obtained by other two existing methods for metering system planning.