Parametric identification by minimizing the squared residuals (Application to a photovoltaic cell)

In this study we develop an algorithm of nonlinear programming to identify the structural parameters of a photovoltaic cell. This algorithm adjusts at best the parameters of the cell's electrical model to the experimental measurements. Thus, to achieve this optimization, we minimize a sum of squared residuals by Gauss Newton's Method which presents an interesting rate of convergence but with sensitivity to the initial conditions. To overcome this issue, we apply, beforehand, the Least Squares Method to the two distinct parts (linear and not linear) of the I
_PV
=f(V
_PV
) characteristic. This first phase allows us to improve the convergence of the algorithm but not its rate. Regarding the last issue we suggest a modified version of Gauss Newton's algorithm.