The Comparing Between Genetic Algorithm and Neural Network to Compute of Three-Basic Solar Cell Parameters with Wide Range of Measured Temperature

Solar cell efficiency considers an important part of the PV system, the parameters (Io, IL, n, Rs, and Rsh) of solar cell is the main part that effected on efficiency. The Matlab simulation program was used to estimate the three parameters' optimization values and evaluated by the Fminsearch method, they calculated for solar cells measured from ۰oC to ۱۰۰oC for seven temperatures, then make comparing for the results between the Genetic Algorithm method with Neural Network Algorithm. This paper establishes the results are frequently in GA was better than NNA, with the Io being ۳.۰۹۹۲ e-۷ and IL being ۳.۸۰۵۹ found by GA. GA is good if they have the same population size and number of iterations. The value of the objective function (fval) in GA is ۰.۰۰۲۸۵۶ but in NNA is ۰.۰۰۵۵۱۸. And also second objective function (fvaltemp) in GA is ۰.۱۰۳۵ with a ۰.۱۰۶۹ value in NNA. From the side, the execution time considers in the Fminsearch method is less than NNA and GA that being ۶۴.۹ s, ۷۸۱ s, and ۲۸۹ s respectively.

Keywords:

genetic algorithm , neural network , Solar cell