The following corrections were made to the manuscript: Fig. 11 was changed to show the correct markers for the optimal points in the Pareto fronts.
The reference to the markers in the subsequent paragraphs is also corrected: “Since a dynamical model will need to map both steady-state behavior and frequency response behavior, the most dynamically accurate 4R-PRBDM ‘(square marker, □)’ is the one closest to the origin. The closeness to the origin is with respect to the ℓ2 norm of both objectives, as represented in Eq. (27).”
“The panel obtained from this optimization is illustrated in Fig. 12. For comparison, the optimal model parameters for the 5R-PRBDM using the same objective (Eq. (27)) give the solution denoted by the square marker (□), while the star (☆) and upright triangle (△) markers show the optimal model parameters for the 4R-PRBDM and 3R-PRBDM, respectively, in Fig. 12. Although the solution...