Immune-genetic algorithm and its application to introduction planning of new products
免疫遗传算法及在新产品投入计划中的应用

To solve complex constrained optimization problems, we propose a new immune-genetic algorithm. It randomly produces a lot of antigens for production and training of antibodies. Then, an efficient immune system with the capability to recognize self and non-self antigens is consisted by these trained antibodies. We embed the immune system into genetic algorithm, and use it to identify the illegal and infeasible chromosomes in the genetic iterations. The recommended algorithm is able to improve the performance of GAs for complex constrained optimization problems. It has been applied into the new product introduction problem presented by a semi-infinite programming model. The satisfactory results have been achieved.