Design Research Of Artificial Fish Swarm Algorithm To Solve Function Optimization Problems

Natural selection cause the evolution of animals,Animal’s survival and foraging lifestyle to be formed, Scientists have get inspiration of solving difficult problems in life by studying the nature of the animal population lifestyle. The researchers find that the fish itself does not have a complex mix of judgment and reasoning ability, but they influence each other by the simple act of community, and ultimately they are able to survive and evolve. Artificial fish swarm algorithm (AFSA) is first proposed by Dr. Li Xiaolei in his doctoral thesis in 2002. Inspired by the movement behavior of fish, the algorithm Solve optimization problems by swarm intelligence. As more and more researchers come to understand the artificial fish swarm algorithm, the algorithm has become current research focuses.The original Artificial fish swarm algorithm has many advantages, such as its powerful robustness and excellent global optimizing ability, In addition, it is a random optimization algorithm which is simple、universal and insensitive to the initial value selection. With the deeper research of fish swarm algorithm and it’s wide range applications, some disadvantages of artificial fish swarm algorithm have been found, for instance, slow convergent speed、low solution precision and easy to fall into local optimum in the later stage of the algorithm. The original AFSA cannot satisfy with people request, in order to overcome these shortcomings of the original artificial fish swarm algorithm,some study work has been done in this paper:1.In order to accelerate the convergence rate and improve Solution precision, a novel artificial fish swarm algorithm (AO-AFSA) is proposed. After the deep research of principles of the artificial fish swarm algorithm (AFSA) and particle swarm optimization (PSO), the way of information exchange and update approaches of particles in PSO were imitated to improve update formula of its three acts based on artificial fish’s individual learning ability and social learning ability. Five typical test functions is selected, experimental results demonstrate that the novel artificial fish swarm algorithm can quickly converge to the global optimum solution by analyzing its Solution precision of the algorithm optimization, convergence speed and robustness, besides, it has greater stability and good performance optimization.2.Aiming at overcoming these disadvantages of original artificial fish swarm algorithm.In this paper, gravity operator is added in original artificial fish swarm algorithm (AFSA), an artificial fish swarm algorithm based on gravity (GA-AFSA) is proposed accordingly.The new algorithm can improve information resources sharing of the various pieces of artificial fish.In addition, it combines the global search ability of gravity operator with local search ability of artificial fish swarm algorithm. Six benchmark function tests show that this algorithm has significantly improved both in convergence speed、stability and optimizing ability, compared with the basic artificial fish swarm algorithm (AFSA) and a novel artificial fish swarm algorithm (AO-AFSA).