Research And Application Of Adaptive Multi-swarm Artificial Fish Swarm Algorithm With Reproductive Characteristics

Artificial fish swarm algorithm is a new intelligent optimization algorithm,which is based on the simulation of the behavior of fish swarm searching for food in nature.The classical artificial fish swarm algorithm has the advantages of simple structure,easy implementation,fast convergence speed,strong robustness,and it has good global search ability.However,the classical artificial fish swarm algorithm also has obvious shortcomings,such as the insufficient information exchange of individual fish during the operation of the algorithm,the algorithm is easy to fall into the local optimum at the later stage,and the algorithm's ability to jump out of the local optimum is weak,which leads to the low convergence accuracy and slow convergence speed of the algorithm.In view of the shortcomings of the classical artificial fish swarm algorithm,this paper improves it and mainly completes the following work:1.On the basis of classical artificial fish swarm algorithm,this paper introduces adaptive mechanism and multi swarm mechanism,and proposes adaptive multi population artificial fish swarm algorithm.During the operation of the algorithm,the individual fish determines which stage of the algorithm is in according to its own state according to the pre-designed rules,and adaptively adjusts the three important parameters of crowding degree,step size and field of vision.It can avoid the premature problem caused by too large crowding,and avoid the problem of individual fish shaking around the optimal solution caused by too large step size and field of vision.At the same time,through the use of multi population mechanism,the information exchange between individual fish is enhanced,and then the convergence speed and accuracy of the algorithm are improved.Through testing on the selected 10 international standard test functions,the experimental results show that the AM-AFSA algorithm is superior to the AFSA algorithm in terms of convergence speed,convergence accuracy,and the ability to jump out of local optimal solutions.2.On the basis of adaptive multi population artificial fish swarm algorithm,the breeding behavior is introduced,and the artificial fish swarm algorithm with reproductive ability is proposed.Through breeding behavior,individual fish break the limit of fixed number of optimization in the process of optimization,improve the local optimization ability and the ability to jump out of the local optimal solution,and then improve the optimization ability of the population.The 10 sets of international standard test functions are tested.Through testing on the selected 10 international standard test functions,the experimental results show that the RA-AFSA algorithm is superior to the AM-AFSA algorithm and the AFSA algorithm in terms of convergence speed,convergence accuracy and the ability to jump out of local optimal solutions.3.On the basis of the previous two steps,this study further applies AM-AFSA algorithm and RA-AFSA algorithm to solve TSP.The experimental results show that the two algorithms proposed in this paper perform well in both the optimization time and the final path when solving the TSP problem.This research has successfully broadened the application range of the artificial fish school algorithm.In short,this research puts forward the AM-AFSA algorithm and the RA-AFSA algorithm after a thorough discussion and analysis of the artificial fish swarm algorithm,and proves the effectiveness of the algorithm through simulation experiments and practical applications.Finally,the work done is summarized and the prospect of future work is put forward.