An adaptive hybrid atom search optimization with particle swarm optimization and its application to optimal no-load PID design of hydro-turbine governor

One metaheuristic algorithm recently introduced is atom search optimization (ASO), inspired by the physical movement of atoms based on the molecular dynamics in nature. ASO displays a unique search ability by employing the interaction force from the potential energy and the constraint force. Despite some successful applications, it still suffers from a local optima stagnation and a low search efficiency. To alleviate these disadvantages, a new adaptive hybridized optimizer named AASOPSO is proposed. In this study, the individual and group cognitive components in particle swarm optimization (PSO) are integrated into ASO to accelerate the exploitation phase, and the acceleration coefficients are introduced to adaptively achieve a good balance between exploration and exploitation. Meanwhile, to improve the search performance of the algorithm, each individual atom possesses its own force constant, which is effectively and adaptively adjusted based on the feedback of the fitness of the atom in some sequential steps. The performance of AASOPSO is evaluated on two sets of benchmark functions compared to the other population-based optimizers to show its effectiveness. Additionally, AASOPSO is applied to the optimal no-load PID design of the hydro-turbine governor. The simulation results reveal that AASOPSO is more successful than its competitors in searching the global optimal PID parameters.