Hybrid Adaptive Improved Sparrow Search Algorithm And Its Application

Sparrow search algorithm is a new swarm intelligence optimization algorithm proposed by Xue Jiankai and others in 2020.The design inspiration of this algorithm is derived from the foraging and anti predatory strategies of sparrow populations in nature.Compared with other swarm intelligent optimization algorithms,sparrow search algorithm has faster convergence speed and stronger optimization ability.However,at the end of the algorithm’s operation,the sparrow search algorithm cannot avoid the decline in population diversity and the problem of being prone to falling into local optimal values.Therefore,in order to improve the shortcomings of the sparrow search algorithm,two improved sparrow search algorithms were proposed,and benchmark test functions were designed to verify the effectiveness and optimization ability of the improved method.Finally,the improved sparrow search algorithm was applied to practical engineering problem solving.The main research contents are as follows:(1)To improve the situation that the search coverage of sparrow search algorithm is not comprehensive enough and the lack of local development ability reduces the optimization accuracy,an adaptive sparrow search algorithm with mixed quadratic interpolation was proposed.Firstly,the dynamic adaptive weight is introduced into the sparrow finder update strategy to adjust the global search and local development ability of the algorithm.Secondly,the quadratic interpolation strategy is introduced to enhance the local development ability and solution accuracy of the algorithm.Through benchmark function tests and engineering design problems,the performance and effectiveness of the improved algorithm are verified.(2)In order to enhance the population diversity and improve the ability of the algorithm to jump out of the local optimal value in the later stage of operation,a chaotic sparrow search algorithm based on hybrid adaptive t-distribution is proposed.Firstly,in the initialization phase of sparrow population,a population is generated using a sin chaotic map to improve the coverage of sparrow in the search space;Then,an adaptive t-distribution variation perturbation is introduced to perturb the position of sparrow participants to improve the search performance of the algorithm.Through experimental verification on several classical test functions and practical application problems,the effectiveness and superiority of the proposed algorithm are proved.Finally,the improved Sparrow Search algorithm is applied to BP neural network optimization to verify the effectiveness and practicality of the improved method in practical applications.