Back propagation neural network method of solution of normal fat dipole and truncated conical grounded monopole and optimization by genetic algorithm

In order to regularize the software by Back Propagation Neural Network (BPNN) two types of dipoles viz. normal fat dipoles as treated in many handbooks and truncated conical dipoles are selected in this paper. The first type is essentially to find out the feasibility of BPNN software to be applied for grounded truncated conical monopole. The second case is a semi-empirical approach that has been developed, where from the optimal dimensions are selected by means of genetic algorithm.