Application Research Of Artificial Fish Swarm Algorithm Of In Conbinatorial Optimization Problems

Combinatorial optimization problems with strong project representativeness are widely applied in real world. However, it is difficult to get optimal solutions; the present solution of combinatorial optimization problems is mainly based on heuristic algorithm. Artificial fish swarm algorithm, a new type of swarm intelligent algorithm, was proposed by Chinese scholar Li Xiaolei in 2002. Because of its simple principle, fast convergence speed and high precision, the algorithm receives extensive attention and application in recent years.Traveling salesman problem is one of the most basic and typical combinatorial optimization problems. Currently, the commonly used treatment methods are ant colony optimization, genetic algorithm and particle swarm optimization. It's easy to fall into local optimum by using the basic genetic algorithm. However, the preying behavior of artificial fish swarm algorithm to the solution of the traveling salesman problem are the basis of global convergence of the algorithm, and it is further strengthened by the rear-end behavior and cluster behavior. In solving traveling salesman problems, the convergence speed of ant colony algorithm is slow and the setting of its parameters will greatly influence the performance of the algorithm, while the artificial fish swarm algorithm has superior convergence speed, which has been proved through examples. Furthermore, the setting parameters of artificial fish swarm algorithm have limited impact on the algorithm performance. A reasonable choice of the number of artificial fish is the key to improve the efficiency of the algorithm. In solving the traveling salesman problems, artificial fish swarm algorithm takes a significant advantage in convergence speed and the high precision is also guaranteed.Job shop scheduling problem is one of the typical combinatorial optimization problems with constraints. To get its encoding has always been one of the main and difficult points of the problem. This thesis will adopt a new encoding in which a constraint will be added. According to the multi-constraints and dynamic nature of the job shop scheduling problem, the thesis will solve the problem by dividing them into four scenarios. Artificial fish swarm algorithm will be used to solve job shop scheduling problem in this thesis. Comparing with existing literature, its adopted calculation examples have higher convergence speed and are much precision. It mainly takes a great advantage in the convergence accuracy.