Abstract
Details concerning the implementation of a versatile genetic algorithm are presented. Solar cell and Schottky diode model parameters are extracted based on the fitness of experimental data to theoretical curves simulated in the framework of certain physical processes and the use of this genetic algorithm. The method is shown to be a reliable alternative to conventional numerical techniques in fitting experimental data to model calculations and the subsequent determination of model related parameters. It is demonstrated, through two examples in particular, that some of the drawbacks associated with the conventional methods can be circumvented if a genetic algorithm is used instead. For instance, a good initial guess is not a critical requirement for convergence and an initial broad range for each of the fitting parameters is enough to achieve reasonably good fits.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
