Artificial fish swarm algorithm is a kind of swarm intelligence algorithm that imitates the behavior of individuals and information interactions among them during feeding process in real environment. This paper proposes an improved artificial fish swarm algorithm. First of all, it adopts dynamic adjustments for the view and step length of artificial fish, which let it keep a large value during the early stage of the algorithm. The value will change from big to small and thus it realizes the transformation from rough search to refined search. Secondly, let the artificial fish move several steps towards a better position during the foraging, bunching and rear-end behaviors and avoid repeatedly searching the same regions for the feeding behavior by introducing the concepts of decay factor as well as information uncertainty of searching regions, which will speed up the convergence. Finally, improve the update strategy of the call-board by comparing the food density in historical best position of each individual artificial fish with that of the partner's in the group or nearby regions and pick out the best to replace each other in order to make the group develop towards a better direction. Simulation result shows that the convergence performance of the improved algorithm is much better than that of the original one's.
[1]
Li Xiao,et al.
An Optimizing Method Based on Autonomous Animats: Fish-swarm Algorithm
,
2002
.
[2]
Shi Qiu-hong.
Parameters Analysis of Artificial Fish Swarm Algorithm
,
2010
.
[3]
Xie Jia.
Improved artificial fish swarm algorithm based on polar coordinate coding
,
2010
.
[4]
Wang Yan-jiao.
Niche artificial bee colony algorithm for multi-peak function optimization
,
2011
.
[5]
Ramachandra Raghavendra,et al.
Particle swarm optimization based fusion of near infrared and visible images for improved face verification
,
2011,
Pattern Recognit..
[6]
Wang Lian-guo,et al.
An Improved Artificial Fish Swarm Algorithm
,
2019,
International Journal for Engineering Modelling.