A Novel Hybrid Method: Genetic Algorithm Based on Asymmetrical Cloud Model

Traditional Genetic Algorithm (GA) easily falls into local optimum and its speed of searching global optimum is very slow. Considering the cloud model has the characteristic of randomness and stability, a new hybrid algorithm (ACGA) based on asymmetrical cloud model and GA is proposed. ACGA use the asymmetrical y-conditional cloud model as cross operation, basic normal cloud generator as mutation operation. In order to search the global optimum better and faster, sampling strategy, tightening strategy and extension strategy are also proposed. The experiments of function optimization are conducted to compare ACGA with other algorithm based on GA. Experimental results show that ACGA outperforms NQGA, CAGA, LARES and CGA, and has good convergence performance.