A competitive-cooperative co-evolutionary optimization algorithm based on cloud model

For the problem of lacking efficient regulation mechanism between local search and global search in an evolution algorithm based on cloud model (CEBA), a competitive-cooperative co-evolutionary optimization algorithm based on cloud model (CECD) is proposed in this paper. It introduces the concept of co-evolution in biological systems into intelligent computing. The proposed algorithm makes cooperation evolution between the parent cloud droplets during the process of evolution to enhance the viability of the individual partners and the exploratory capability of the algorithm. It effectively ensures the diversity of the cloud droplets in the evolution. It also employs the competition evolution between sub-population to ensure the global convergence and stability of the proposed algorithm. Theory analysis and experiments on 12 widely used benchmark problems show that when compared with the CEBA which only uses individual fitness to control the biological evolution, CECD algorithm is capable of improving search ability, convergence speed, solution quality and algorithm stability.