Skip to Main Content
The purposes of our work was to design an improved particle swarm optimization (PSO) framework that is capable of wider search area and better fitness values by diversity-guided bilateral objective function. To address the problems of premature convergence of existing PSO techniques, an improved PSO framework have been proposed, consisting of bilateral objective function (BOF). The BOF was developed based on both smaller cost value (minimized) and bigger distribution (diversity-guided) of individuals. Instead of using only cost value down for the objective function as in the case of the conventional optimization techniques, the proposed PSO framework employs both minimized cost value and promoted solution search ability by exploring more wider search space. The global optimum was obtained easier after the combination of bilateral objective function, especially for complex test functions. The proposed algorithm has been demonstrated with a significantly improvement for cost value in couples of test functions.