Optimization using civilized swarm: Solution to economic dispatch with multiple minima

This paper presents a new optimization strategy, civilized swarm optimization (CSO), by integrating society-civilization algorithm (SCA) with particle swarm optimization (PSO). In SCA, individuals are grouped in small societies (clusters) with better performing individuals of each cluster as society leaders (SL). All such societies constitute the civilization with the best society leader as the civilization leader (CL). To perform optimization, the society members follow their SL; the society leaders follow the CL. Whereas in PSO, particles modify their positions according to their best experiences and that of the swarm. SCA differs with PSO in the fact that the individuals of SCA follow only their leaders neglecting self-experiences. The proposed CSO considers the swarm to be a civilization with societies. The particles of a society are made to search within the society with the help of both the SL and their own experiences; therefore, they can exploit a “promising area”. All the society leaders are allowed to explore the search space for new promising areas through the guidance of both their own experiences and that of the swarm leader. The efficiency of CSO is tested for a set of multi-minima economic dispatch problems and superior results are obtained.