Figure 1.

Developed hybrid optimization algorithm with tolerance adjustment consisting of the features: initialization within the feasible region (A); initial optimization using the SQP (B); interactive hybrid process using SQP (C); STRiN (D); and optimization control algorithm (E). f(Θ*_k): objective function value at the current local minimizer Θ*_k; χ²_UL: upper limit of f(Θ*_k) to invoke STRiN (if fΘ*_k) <χ²_UL; α: parameter scaling constant; Θ₀: initial guess; Θ*: local minimizer from a successful sub-optimization; Θ°: iterate recorded for the smallest function value up to the current optimization trial; $\widehat{\Theta }$: ultimate minimizer.