A cloud model based DNA genetic algorithm for numerical optimization problems

Bio-inspired algorithms for optimization are significant topics in the areas of computational intelligence. Traditional genetic algorithm easily gets stuck at a local optimum, and often has slow convergent speed. To overcome these drawbacks, the Cloud model based genetic algorithm with DNA encoding (CM-DNAGA) is originally proposed in this study. The CM-DNAGA algorithm is based on not only the properties of randomness and stable tendency of the normal cloud model, but also the idea of GA with the bio-inspired coding method, i.e., DNA. In CM-DNAGA, a Y conditional normal cloud generator is used as the genetic crossover operator, and a basic normal cloud generator is used as the mutation operator. The simulation experiments are conducted on 12 numerical optimization functions, which evaluate the performance of the proposed algorithm. The experimental results indicate that the proposed method is a competitive optimizer in comparison with the three state-of-the-art heuristic algorithms, i.e. standard GA, PSO and RNA-GA.