Design And Application Of Meta-heuristic Algorithms Guided By Collective Wisdom

In real life,many problems cannot be solved by traditional optimization algorithms.The emergence of meta-heuristic algorithms overcome the shortcomings of traditional optimization algorithms to some extent.In recent years,due to their simple structure,easy to implement,independent of gradient information,and able to avoid local optimal to a large extent,meta-heuristic algorithms have become a hot topic for many scholars due to these advantages and their wide application in engineering design problems.In order to further improve the local exploitation ability or global exploration ability of the algorithms,many scholars put forward various strategies applicable to the meta-heuristic algorithm.In order to further improve the local exploitation ability of meta-heuristic algorithm and better balance the global performance of the algorithms,this paper makes an in-depth study on the beneficial information of population individuals in the process of solving the algorithm,and designs operators that can improve the performance of meta-heuristic algorithms,In this paper,the backtracking search algorithm(BSA)and grey predictive evolutionary algorithm(GPEAe)are taken as research objects,and two improved algorithms(MBSAgC and TOGPEAe)are designed.The main research work of this paper are as follows:(1)Proposed the concept of "collective wisdomf" Inspired by the winner-tendency and clusterity-tendency behavior of humans and animals,this paper proposes a new idea:"collective wisdom".In this paper,the guidance information of the best individual and the mean value of the individual population are mined to balance the overall performance of the algorithm.According to the winner-tendency of collective wisdom,a new winner-tendency topological opposition-based learning and a linear combination guided by winner-tendency are proposed to enhance the local exploitation ability of the algorithm,and an improved mutation operator is designed,which is composed of the linear combination guided by the winner-tendency and a learning strategy guided by the clusterity-tendency,So as to improve the overall performance of the algorithm.(2)Proposed an deeply mining backtracking search optimization algorithm guided by collective wisdom(MBSAgC)This paper introduced the principle and algorithm flow of BSA systematically,and analyzed the advantages and disadvantages of the algorithm,then summarizes the improvement and application of BSA.In view of the slow convergence speed of BSA in the later stage,the designedwinner-tendency topological opposition-based learning operator is introduced after the initialization of the algorithm,and the improved mutation operator is designed,which is composed of the linear combination guided by the winner-tendency and a learning strategy guided by the clusterity-tendency,so as to improve the overall performance of the algorithm.(3)Proposed a topological grey prediction evolutionary algorithm guided by collective wisdom(TOGPEAe)In order to further improve the overall performance of gpeae,this paper introduces the designed winner-tendency topological opposition-based learning operator guided by the winner-tendency before the selection operation of the originalalgorithm,so as to improve the convergence speed of the algorithmTo verify the effectiveness of the proposed algorithm MBSAgC and TOGPEAe,in this paper,the algorithm is tested on two benchmark function test sets(CEC2005 and CEC2014)and on a test set composed of engineering design problems,the experimental results show that compared with the original algorithm,the new algorithm presented in this paper has a great improvement in both solution accuracy and convergence speed;and compared with other state-of-the-art algorithms,the overall performance of the new algorithm is also very competitive.At the same time,it also provides ideas for the combination of operators designed by“collective wisdom" guidance and famous meta-heuristic algorithms such as DE PSO.