The radiation pattern synthesis of rectangular microstrip patches arranged on three-dimensional arrays using numerical deterministic optimization methods was vulnerable to the local optima problem. Then, the use of probabilistic optimization methods constitutes an attractive alternative. Hence, these methods give the opportunity to deviate from the local optima to global optima. The local aspect of the findings is a frequent problem for the calculus-based methods, where the obtained solution depends on the starting point. It is towards this perspective that we develop in this paper a synthesis based on genetic algorithm which remains a probabilistic optimization method, easy to integrate in computations and very efficient even though it use probabilistic and deterministic rules, particularity which makes the efficiency and robustness of this heuristic algorithm.   

The radiation pattern synthesis of rectangular microstrip patches arranged on three-dimensional arrays using numerical deterministic optimization methods was vulnerable to the local optima problem. Then, the use of probabilistic optimization methods constitutes an attractive alternative. Hence, these methods give the opportunity to deviate from the local optima to global optima. The local aspect of the findings is a frequent problem for the calculus-based methods, where the obtained solution depends on the starting point. It is towards this perspective that we develop in this paper a synthesis based on genetic algorithm which remains a probabilistic optimization method, easy to integrate in computations and very efficient even though it use probabilistic and deterministic rules, particularity which makes the efficiency and robustness of this heuristic algorithm.   

The radiation pattern synthesis of rectangular microstrip patches arranged on three-dimensional arrays using numerical deterministic optimization methods was vulnerable to the local optima problem. Then, the use of probabilistic optimization methods constitutes an attractive alternative. Hence, these methods give the opportunity to deviate from the local optima to global optima. The local aspect of the findings is a frequent problem for the calculus-based methods, where the obtained solution depends on the starting point. It is towards this perspective that we develop in this paper a synthesis based on genetic algorithm which remains a probabilistic optimization method, easy to integrate in computations and very efficient even though it use probabilistic and deterministic rules, particularity which makes the efficiency and robustness of this heuristic algorithm.