Abstract This work proposes a Multi Agent based Symbiotic Organisms Search (MASOS) by incorporating multi agent system (MAS) into the Symbiotic Organisms Search (SOS) algorithm. Each organism in MASOS acts… Click to show full abstract
Abstract This work proposes a Multi Agent based Symbiotic Organisms Search (MASOS) by incorporating multi agent system (MAS) into the Symbiotic Organisms Search (SOS) algorithm. Each organism in MASOS acts as an agent participating in local interactions, to search for the optimum solution. In doing so, the conventional operations, namely, mutualism, commensalism and parasitism become localized and thus, decentralized. The decentralization of the SOS algorithm helps in exhaustive exploration of the search space, eliminating the chances of getting trapped into local optima. The algorithm is tested over ten benchmark functions, and the results compared with well known evolutionary and swarm intelligence based algorithms. Further, the proposed algorithm is applied to the tuning of 1-Degree of Freedom (DOF) and 2-DOF Integer Order and Fractional Order Proportional-Integral-Derivative (PID) controllers for magnetic levitation plant. The obtained results validate the accuracy, stability and strong performance of the proposed algorithm over other algorithms.
               
Click one of the above tabs to view related content.