An Immunity Based Hybrid Evolutionary Algorithm for Engineering Optimization

The immune system has been recognized possesses pattern recognition ability in which the lymphocytes can learn to distinguish selves and match a variety of pathogens. Consequently, sufficient antibodies are generated to eliminate the growth of the foreign antigens. This paper describes the inspiration from the immune system and how to apply immune system principles to develop the global unconstrained and constrained optimization algorithms. The features of the proposed approach contain: the affinity maturation in immune system has been employed as the primary principle, the real number code has been used as genes representation in this development; the modified expression strategy for constraints handling and a diverse multiplication generated in genetic algorithm. Numerical structural engineering optimization problems demonstrate that the proposed immunity based evolutionary approach has the solution consistency; avoiding premature and can achieve a robust final design.