Artificial Fish Swarm Algorithm And Its Application In Coverage Optimization Of WSN

Optimization problems need to be solved in many fields which have great application prospect in the development of the national economy. With increasing of the complexity and scale of the optimization objects, its objective function has the features of non-linear, constrained, multi-objective, multi-model and even non-continuous or non-analytical. So it has become more and more difficult to solve the combinational optimization problems by the classical optimization methods based on the strict mechanism model.The intelligent optimization algorithm provide a very effective new method to solve complicated optimization problems, which can make up a computing system which has the characters of large-scale parallel, Self-organization, adaptive and self-learning by analogue and bionics. Compared with the traditional optimization methods they have the advantages of non-centralized- control, multi-agent mechanism, simple algorithm structure, implicit parallel and easy to understand and implement. So they play more and more important roles in the optimization of production process, improvement of production efficiency and saving resource.Artificial fish swarm algorithm is an optimization algorithm based on the simulation of fish swarm behavior. It has the advantages of parallel processing and global optimization. It has no special requirements for objective function, the initial value and parameters.In this paper, the improvement and application of the artificial fish swarm algorithm are mainly discussed. The contents are as follows:We analyze the nature, characteristics, mechanism and impact to the optimization result of the crowd factor in artificial fish swarm algorithm (AFSA). Then we test it with some trial functions. The simulation results show that the crowd factor has little advantages to overcome local extreme, improve convergence speed and precision. So we can ignore the crowd factor to simplify algorithm in the practical application. We can obtain better convergence speed and optimization precision and keep strong ability to overcome local extreme.Based on the analysis of the optimization mechanism of artificial fish swarm algorithm, four adaptive artificial fish swarm algorithms(AAFSA) are proposed to simplify the parameter settings and improve its convergence speed and optimization precision. These methods make every artificial fish be able to adjust their own visual and step length according to the condition of fish swarm. The simulation results show that the convergence speed, optimization precision and the ability to overcome local extreme of the AAFSA are better than that of AFSA.AFSA is used to coverage optimization of the wireless sensor networks. An optimal distribution algorithm based on AFSA is proposed using the mobility of sensor nodes. The simulation results show that, the AFSA can not only optimize the node distribution in target area but also improve the network's coverage rate with a lower cost.