The application research of improved cloud genetic annealing algorithm

Abstract

Improved cloud genetic algorithm (ICGA) was proposed in this paper. ICGA combined the characteristics of the powerful global search capability of genetic algorithm (GA) and the powerful local search capability of simulated annealing (SA). The initial solution was generated by GA, the crossover probability (Pc) and the mutation probability (Pm) were generated by the characteristics of randomness and stable tendency of the droplets in the cloud models. Adopting the metropolis sampling process of the SA in the process of crossover and mutation operation, the obtained solution became the initial population of the genetic operations for further evolution. This structure effectively avoided the GA premature and defects of weak local search ability, which improved the search ability of the system. The simulation results further showed the effectiveness of the algorithm.