Research And Application Of Sailfish Optimization Algorithm Based On Levy Flight And Cloud Mode

The sailfish optimizer algorithm is a new swarm intelligent optimization algorithm,which is based on the simulation of the alternating attack behavior of sailfish when hunting sardine in nature.There are two populations in the algorithm,in which sailfish is the predator and sardine is the prey.The predator is the candidate solution for optimization,and the prey is used to increase the randomness of the algorithm.The sailfish optimization algorithm has the characteristics of strong robustness,fast convergence speed,high convergence accuracy and simple structure,and the strategy of simulating two populations to find the optimal solution improves the global search ability of the algorithm.However,it also has obvious needs to be improved.For example,in optimization,the algorithm has a weak ability to jump out of local optimization,the sardine is easily affected by the optimal position of sailfish and falls into local optimization.On the other hand,the information exchange between the sailfish and sardine is not sufficient during the algorithm operation,which affects the time and precision of optimization.In view of the above problems,this paper has made relevant improvements to the original algorithm,and mainly completed the following work:1.On the basis of the sailfish optimization algorithm,the levy flight is introduced,and the levy sailfish optimizer algorithm is proposed.During the operation of the algorithm,when the sailfish’s attack power is less than 0.5,levy flight is introduced into the position update of the sardines.The ability of the algorithm to jump out of local optimization is increased through the large step with high probability during the random walk of levy flight,so as to avoid the problem of precocity caused by the premature aggregation of sardines,and thus the convergence accuracy of the algorithm can be improved.The 17 sets of international standard test functions are tested.The experimental results show that the convergence accuracy and convergence speed of the algorithm are better than that of the sailfish optimizer algorithm.2.On the basis of the levy sailfish optimization algorithm,the cloud model is introduced,and the cloud levy sailfish optimizer algorithm is proposed.During the operation of the algorithm,when the position of sailfish and sardines is updated,the position of sailfish is taken as the sample data to generate cloud droplets through the cloud model.Then the fitness values of the sailfish,sardine and cloud droplets are compared respectively,and the best ones are retained to generate a new generation.In this way,the influence of elite sailfish population is increased,the problem of sardine population shaking around the optimal solution is avoided,and the information exchange between the populations is enhanced.The17 sets of international standard test functions are tested.The experimental results show that the algorithm is superior to the sailfish optimizer algorithm and the levy sailfish optimizer algorithm to a certain extent.3.On the basis of the above,the levy sailfish optimizer algorithm and the cloud levy sailfish optimizer algorithm are applied to the optimization of support vector machine parameters for classifying and predicting student grades.The experimental results show that the parameters optimized by the two algorithms have good results when support vector machine is used for student data classification and prediction,which verifies the feasibility of the application and broadens the application range of the sailfish optimization algorithm.In summary,after a comprehensive and in-depth study of sailfish optimization algorithm,the levy sailfish optimizer algorithm and the cloud levy sailfish optimizer algorithm are proposed,and the effectiveness of the improved algorithm is proved by standard test function and correlative application.Finally,the work is summarized and the prospect of future work is put forward.