Sensitivity analysis on the weights in power system restoration decision-making

Power system restoration is a multi-objective problem with limited optional plans. Generally, multi-criteria decision-making (MCDM) method is employed to rank optional restoration plans and provide decision support for dispatchers. Weights set by decision-makers based on their experience may have subjectivity and then influence optional plans ranking result. In order to properly select attribute weights for restoration decision-making and ensure power system safely and rapidly restored after blackout, it is necessary to analyze the sensitivity of ranking result on weights variation. Based on the technique for order preference by similarity to an ideal solution (TOPSIS) method, this paper proposes a weights sensitivity analysis method. The ranking result sensitivity of TOPSIS method was analyzed and the absolute sensitive index of the optimal plan was presented by using Euclidean distance and nonlinear programming method. Simulation results of Shandong power grid show that the method is conducive to help decision makers easily recognize and carefully set critical weights.