It was used as a cost function to be optimized, the injection barrier height parameters converged to the exact value, while all the other parameters did not. Yet, a good correlation between n_sim (P_exact) and n_sim (P_1) has been noted according to the relative error (≈〖10〗^(-13)) between them. This shows that these two parameters are highly sensitive during the first 30 seconds of polarization. This result is compatible with that found by the Sobol indices method.
When Z_2 (P) was optimized, the exact value of the barrier height of injection and detrapping parameters were easily computed, while all the other parameters were not. Yet, an agreement between n_sim (P_exact) and n_sim (P_2) has been seen, according to the relative error (≈〖10〗^(-9)) between them. This implies the high sensitivity of the detrapping parameters at temperature ∈[30-50] °C or while charging for time greater than thirty minutes, which is also in agreement with the results produced by the Sobol indices method.
When Z_3 (P) was chosen to be optimized, mobility parameters converged to their exact values. On the other hand, recombination parameters were not very sensitive to thisthese kind of experiments, only R_2 converged to the exact value.
Finally, when all the experiments were combined in a single cost function Z(P), the optimization algorithm was able to find all the exact values of all the unknown parameters. This also contributed in finding the exact values of the trapping parameters, which seemed to be impossible by optimizing each cost function individually. By the same token, it also assisted in finding the exact values of R_1 and R_3 which was difficult to be foundfind.
In addition, the standard deviation parameter σ can be easily found using any IEC (i.e. σ could be computed using any cost function). Besides, increasing the temperature more than 50°C did not affect the convergence of the detrapping parameters.