Immune Algorithm Based on Exponential Variation

In this paper, an immune algorithm based on exponential variation (IABEV) is proposed to solve global numerical optimization problems. Variability scale of the algorithm is divided into several levels by the power series with a base of 10: large-scale level is conducive to jump out of local optimal solutions, which achieve global optimization, a small class of high-precision scales in favor of local optimization. It tests the performance of the new algorithm with five benchmark functions and is compared with other optimization algorithms. Experiment results show that IABEV has excellent performance.