Abstract:
This project aims to develop the maximum power point algorithm for the system. The system is
tested for both traditional and the artificial intelligence algorithm. The traditional algorithms are
implemented in the the MATLAB function block and is tested in the Simulink of MATLAB. Two
types of traditional algorithm are implemented in the system that is Perturb and observe method
and the incremental conductance algorithms.
Their performance evaluation in terms of efficiency is done. Then three
algorithm Lavenberg Marquardt, Bayesian Regularisation and scaled conjugate is implemented in
the system Their performance characteristics in terms of mean square error, regression , efficiency
is done. The parameter to be optimized is number of hidden layer. Other parameters like type of
transfer function and the type of network can be optimized in the nn tool of MATLAB. The Neural
Network fitting app is used in the system. And the algorithms is used in the system using the
application. And the results for the algorithm is concluded at the end of the report
