One of the recent frontiers of the multiple objectives optimization strategies is the introduction of an artificial neural network in place of a more time-consuming numerical tool or approach to… Click to show full abstract
One of the recent frontiers of the multiple objectives optimization strategies is the introduction of an artificial neural network in place of a more time-consuming numerical tool or approach to compute the cost functions. This paper describes the development of a multiple objectives optimization strategy for an electromagnetic bandgap common mode filter by using an artificial neural network. The network training, that is to say the evaluation of the internal weights and biases, is optimized by a genetic algorithm. The results are discussed and also compared with those stemming from an already developed multiple objective sequential optimization strategy applied to the same common mode filter. In both strategies, the cost functions are the same and the optimization algorithm is a differential evolutionary one.
               
Click one of the above tabs to view related content.