| Optimization problems need to be solved in many fields and the fine solutions to the problems may lead to great economic benefit. With the increasing complexity and scale of the optimization problems, classical optimization methods which are based on the strict modeling become difficult to use.The Artificial Fish-Swarm Algorithm (AFSA) is an evolutionary computation technique based on intelligence bionic optimization algorithm. This was developed by Li Xiaolei in 2002 and inspired by the social behavior of a group of fish. The AFSA has a stronger robustness; the fine distributed computing and easy to union with other methods. At present, this algorithm research has already improved many other applications and has developed from a one-dimensional static state optimization solution to a multi-dimensional dynamic combination optimization solution. The AFSA has already become extremely active from research question in the interdisciplinary studies.The main contributions of this paper are as follows:1) To look at the background of intelligence bionic theory. Summarize the theory and characteristics of intelligence bionic algorithms. And the commonness and individuality of intelligence bionic algorithms.2) Description of the AFSA which is a new bionics algorithm. It includes an introduction of the basic theory, the algorithmic description, the parameters and flowchart of AFSA.3) It proposes four improvement methods on the basis of the research on AFSA during the past 6 years. It also introduces several typical improvement models.4) It proposes two improved AFSA, the IM-AFSA and the IV-AFSA when embedded in the information processing mechanism of the immune system. The IM-AFSA is based on immune memory and mediatory mechanism; the IV-AFSA is based on vaccination. Compared with original AFSA, IM-AFSA has a better effect in the aspect of avoiding the premature convergence and the convergent speed. IV-AFSA not only has the same effect, but also has a better value. |
PDF Full Text Request