Ant Colony Optimization Algorithm Based on Immunity Vaccine and Dynamic Pheromone Updating

To solve the problem of complexity and uncertainty in logistics and distribution system, a kind of Ant Colony Optimization algorithm based on immunity vaccine and dynamic pheromone updating is put forward in this paper. In this new Algorithm, initial antibody is vaccinated first to produce better solutions, and initial parameters are then set depending on these better solutions by the Ant Colony. Mechanism of dynamic adjustment and pheromone updating is also introduced into this algorithm to slow down the increment of the pheromone concentration difference between the paths, avoiding sectional solution of the algorithm. Experiment results show that DACOIV can effectively find optimal path for logistics and distribution system.

I. Introduction

Artificial immune algorithm is a kind of stochastic optimization method based on the function, principles, basic features and related theory of biological immune system, which is put forward to solve the various complicated problerusl‘ [1]–[4] It has fast random global search capability, but can not make full use of the feedback information in the system, and its efficiency is low in some certain range when doing a lot of unnecessary redundant iteration. ACO (Ant Colony Optimization) is a kind of Bionic evolutionary algorithm put forward first by Dorigo [5]–[8]. The inspiration of ACO is from the behavior of Ant Colony'S looking for foods. It can not only solve the problems of static combinatorial optimization, but also those of dynamic combinatorial optimization. Its inherent parallelism, robustness, and other good characteristics make it a kind of effective calculation model which can solve complex optimization problems, but it also has some defects such as slowness, prematurely and stagnation.