A Project on Swarm Meeting Ant Colony Optimization Based on the Experience Model
|Course||Computer Software and Theory|
|Keywords||Combinatorial Optimization Ant Colony Optimization Max - Min Ant Algorithm Cluster Experience Groups meet Ant Colony Algorithm|
Behavioral characteristics of social insects, scientists discovered that insects very limited ability of each individual in the community's cooperation is essentially a self-organization, in many occasions, although such cooperation may be very simple, but it can solve complex problems . Gregarious behavior of insects for the computer scientists to design distributed control and powerful optimization method. Ant colony algorithm is the use of swarm intelligence to solve combinatorial optimization problems typical example. Ant colony algorithm is a kind used in combinatorial optimization problems heuristic search algorithm, with positive feedback, distributed computing and inspiring search for other characteristics. Since its emergence ant colony algorithm has been successively applied to TSP problem, resource allocation and other classic quadratic optimization problem to get better results, attracting a large number of academic interest. In recent years, the field of study is also extended to the dynamic environment, chaos computing, multi-target areas, based on ant colony optimization algorithm advent of new technologies also went and continuous improvement. This paper studies the typical NP problem - start traveling salesman problem, ant colony optimization algorithm development background, content, implementation, and performance are described in detail, the algorithm itself is conducted in-depth research, put forward their own improvements. In this paper, ant colony optimization algorithm made the following main points: 1. Propose a new algorithm based on ant colony clustering experience, build the new algorithm ant model, and through the program realization of this algorithm to complete a new ant model and maximum - minimum ant algorithm combines, and through experiments to find a new algorithm for the optimal parameter settings; 2 in 5 different sizes of TSP problem, and with the improved algorithm and maximum - minimum ant algorithms are simulation. Through comparative analysis found that experience-based cluster model enables ant colony optimization algorithm already searching speed and optimization capability has been significantly improved, but in a large-scale urban agglomerations stability of the algorithm is not high enough; 3. the original encounter algorithm is proposed to improve the group met idea of ??the algorithm. Then the algorithm to achieve through the program, and through experimental analysis of the improved algorithm What are the advantages; 4. These two algorithms combined into a new algorithm: Based on the experience of the cluster group met ant colony algorithm, experiments show that the new algorithm For ant colony algorithm based on cluster is not stable enough experience disadvantage had a good improvement, while optimizing the algorithm speed and optimization ability did not decline.