| With the continuous development of biointelligence group behavior,swarm intelligence optimization algorithm has become a key technology to solve global optimization problems in multiple types of fields with its powerful search ability and adaptability.In recent years,its application range has become more and more extensive,and it has become a hot topic in the field of optimization.With the development of technology,the swarm intelligence optimization algorithm has successfully overcome the problem of large scale and complexity of the traditional optimization algorithm,which can quickly and accurately search for the global optimal solution according to the social behavior of intelligent individuals and other related factors,and can adjust its own parameters in time according to the characteristics of each intelligent population,so as to better guide the activities of the group.With efficient computing performance and excellent adaptability,swarm intelligence optimization algorithms are applied to various practical scenarios.In recent years,as a novel swarm intelligence optimization algorithm,the sparrow search algorithm has obvious advantages: simple principle,strong search ability,and few structural parameters,but there are also some shortcomings,such as the inability to maintain the diversity of the population,easy to fall into local optimization,and low efficiency in solving complex optimization problems.Since the algorithm was proposed,scholars in various fields have tried to improve different strategies,but the problem of lack of population diversity still exists in the whole process of search,and decreases linearly with the increase of the number of iterations,making it difficult to improve the overall optimization ability of the algorithm.In this regard,after combining domestic and foreign research,the sparrow search algorithm is deeply studied and improved based on the goal of improving population diversity for various types of optimization problems,combined with a variety of effective improvement measures,and the actual performance of the proposed algorithm is verified under the scenarios of multiple different types of engineering optimization problems.The specific research work of this paper is as follows:(1)The correlation group intelligent optimization algorithm and sparrow search algorithm are elaborated,and the basic principles and algorithm flow of classical optimization algorithm and sparrow search algorithm are analyzed,which lays a theoretical foundation for the subsequent work of this paper.(2)A sparrow search algorithm with mixed multi-strategy improvement is proposed.In order to solve the challenges of sparrow search algorithms in finding the optimal solution of the objective function,such as premature convergence,insufficient population diversity,and low solution accuracy in high-dimensional environments.Using elite reverse learning to improve the quality of the initial population,accelerate the convergence speed,and design a phased control step formula,balance the search and development of the algorithm,and introduce the chaotic cosine change factor to act on the follower position update,so as to achieve the purpose of maintaining the diversity of the population and improving the search ability of the algorithm,Lévy flight using the adaptive selection mechanism can greatly improve the flexibility of the algorithm,so that it still has good diversification characteristics in different periods.By setting multiple test functions of single peak,multimodal and fixed dimensions,the performance is simulated and compared with the standard sparrow search algorithm and other algorithms,and the performance is verified in the design problems of welded beams and three-pole trusses.Finally,the simulation results show that the proposed algorithm can maintain high population diversity in the actual problem,improve the accuracy of solving the target problem,and have certain stability.(3)An improved sparrow search algorithm based on Piecewise mapping is proposed.In the face of the problems of poor initial population diversity,uneven distribution and insufficient search accuracy in the face of complex problems,the diversity is difficult to maintain and insufficient search accuracy of the sparrow search algorithm,the diversity of the population is enhanced based on Piecewise chaotic mapping,the uniformity of the spatial distribution is improved,and the finder search adopts a step size adaptive strategy to achieve the balance between global search and local development,improve the search accuracy,introduce Gaussian variation disturbance,provide support for the increase of population diversity in the whole process,and strengthen the ability of the algorithm to jump out of the local optimal.Finally,the transboundary treatment method based on upper and lower bounds is used to adjust the new solution beyond the search range,and further expand the diversity distribution of the population.In this paper,the proposed algorithm is compared with other classical algorithms and improved algorithms with better effects in 14 benchmark test functions,and the improved algorithm is optimized for probabilistic neural networks,which has good experimental effects in transformer fault diagnosis,and the simulation results show that the proposed algorithm has significant advantages in terms of comprehensive performance and ability to maintain diversity.With the increasing synthesis and complexity of real problems,this paper proposes two improved sparrow search algorithms to help the algorithms maintain sufficient diversity and achieve better solution results in the whole process of searching.Whether it is from the comparison of simulation experiments or the comparison of actual engineering application effects,the proposed algorithm has obvious advantages and can be used as an effective means to solve practical problems. |
