Ant Colony Optimization (ACO) is a meta-heuristic approach for solving hard combinatorial optimization problems, and since it was introduced, it is superior to other approaches at solving combinatorial optimization problems. As a global searching approach, ant colony algorithm has many advantages, such as distributing, positive feedback, robustness and easily being connected with other meta-heuristic approaches to solve problems. Although traditional ant colony algorithm has good capability of searching global solutions, it has some shortcomings, such as long time searching, being stagnant during searching, and being premature when there is a big scale of problem. As for the drawbacks above, the paper introduces immune algorithm to improve basic ant colony algorithm, because immune algorithm has advantages of speediness, randomicity, and global astringency. As a result, it is good to make up drawbacks of basic ant colony algorithm by using advantages of immune algorithm, so that the traditional ant colony algorithm can be improved.Firstly, the paper gives three improved ant colony algorithms based on immune mechanism. The first one is ant colony algorithm with immune ability of side consistency restraining. At the beginning of the algorithm, generate original pheromone distribution by immune algorithm. Then at the end of the algorithm, adjust the pheromone consistency on the route by using side consistency restraining mechanism, so that the diversity of ants can be kept. The second one is ant colony algorithm with immune ability of vaccine. Through picking up and inoculating vaccine, we can improve the quality of solutions of problems. Combine max-min ant colony algorithm to update pheromone. So that the phenomenon of stagnation and premature can be avoided. The third one is ant colony algorithm with immune ability by introducing knowledge. At the beginning of the algorithm, generate original pheromone distribution by immune algorithm. Afterwards, introduce knowledge as follows: if there are crosses on the route, clear up the crosses and save the results. Then adjust pheromone adaptively. So the quality of solutions, global searching ability, and searching speed can be improved.Secondly, the paper applies three improved algorithms above in traveling salesman problem. The result shows that the improved algorithms raised by the paper are much better than basic ant colony algorithm and other optimal algorithms, and they are more efficient and have good convergence speed.Finally, the paper applies colony algorithm with immune ability of side consistency restraining in combinatorial problem of water distribution system. By the example, it is further proved that the improved algorithm given by the paper is better than other traditional algorithms at solving optimal problems of water distribution, which validates the efficiency of the improved algorithm. |