A Probabilistic Evaluation of Fitness Based Immune Chaotic Algorithm for Constraint Optimization Problems

Multidisciplinary collaborative decision of product design is modeled as a constraint satisfaction problem and a new hybrid algorithm which combined chaotic search with immune algorithm is developed to solve the model. The meticulous searching capability of chaotic algorithm can keep immune algorithm from stunning into local optimum and find global optimal design solution. For the fitness evaluation of antibody, a probabilistic method was adopted to deal with constraints in place of traditional punishment function approach, thus the problem that the punishment coefficient is difficult to determine can be solved effectively. Optimization design result of a rotor system validated correctness and efficiency of hybrid algorithm.