A New Hierarchical Multi Group Particle Swarm Optimization with Different Task Allocations Inspired by Holonic Multi Agent Systems

•Hierarchical multi group PSO is proposed inspired by holonic organization in multi agent systems.•Proposed structure provides a lot of facilities for improving the performance of PSO.•For creating a suitable balance between exploration and exploitation, different tasks are assigned to different groups.•Particles use different parameter settings, dynamic neighborhood topologies and learning strategies based on their group's task.•Experimental results indicate that HPSO-DTA surpasses other algorithms.