A novel hybrid immune algorithm and its convergence

Abstract

In this paper we propose a hybrid algorithm that can overcome the typical drawback of an artificial immune algorithm, namely, the propensity to runs slowly and experience a slower speed of convergence is than a genetic algorithm. Our hybrid algorithm combines the steepest descent algorithm with an artificial immune adaptive algorithm based on Euclidean distance. The hybrid algorithm fully displays global search ability and the global convergence of the immune algorithm. At the same time, the hybrid algorithm inserts a steepest descent operator to strengthen the local search ability. Experimental results show that the hybrid algorithm successfully improves the operational speed and convergence performance. In addition, this paper proves the convergence of the hybrid algorithm with a quasi-descent method.