Artificial immune algorithm and its applications
人工免疫算法及其应用研究

In order to solve ill-posed constrained optimization efficiently, an artificial immune algorithm (AIA) imitating biological immune systems and its approach were presented; several artificial immune operators producing new antibodies were constituted, and the global convergence of the algorithm was studied. AIA succeeds to the mechanism of the survival of the fittest, but the methods producing new antibodies are more varied than ones producing new individuals in genetic algorithm (GA). If the best of all antibodies is made sure into next antibody colony as selecting antibodies, AIA is global convergent. A simulation text of 100-city TSP was done with AIA, and its computational process is compared with that of genetic algorithm. The results show that AIA is more efficient than GA in global optimization and computational process of AIA is faster than that of GA.