Sensitivity analysis for optimization problems solved by stochastic methods
Ricardo H. C. Takahashi, Jaime Arturo Ramírez, J.A. Vasconcelos, R.R. Saldanha · IEEE Transactions on Magnetics · 2001
The problem of obtaining sensitivity analysis information from optimization stochastic methods is addressed in this work, The sensitivity analysis calculation is obtained using the evaluations of the objective function as samples of the behavior of the objective function in the vicinity of the optimum, A linear optimization problem is formulated for the approximation of a contour surface of the objective function in the vicinity of the optimum, The formulation is applied to TEAM benchmark problems 25 and 22. The results, which are in good agreement with the literature, show the applicability of the formulation in practical optimization problems.