Flower pollination algorithm is a new meta-heuristic algorithm that simulates the pollination process of plants,this algorithm not only utilizes special Levy flight to achieve cross-pollination,but also designs transition probabilities to adjust the balance between global exploration and local exploitation.Because this algorithm has the advantages of few parameters,good robustness,easy adjustment and easy implementation,it is favored by many scholars and widely used in different fields.However,the algorithm still has problems such as low solution accuracy and easy precocious convergence.In view of the above shortcomings,this paper proposes two improved flower pollination algorithms,and the details are as follows:Aiming at the problem that the flower pollination algorithm is easy to converge to the local optimum when solving the optimization problem,considering the characteristics of the golden sine algorithm jumping out of the local optimum by reducing the search space,a new method is proposed that combines the dynamic convergence factor and the golden sine algorithm.In this algorithm,dynamic convergence factor is introduced in cross-pollination to improve the solution accuracy,and golden sine optimization is carried out in self-pollination to enhance the ability to jump out of local optimum.With the help of simulation experiments and case analysis,it is verified that the improved algorithm can more effectively avoid algorithm prematurity,and at the same time improve the optimization accuracy and convergence speed of the flower pollination algorithm.Traditional flower pollination algorithm uses random initialization of the population,which is easy to reduce the quality of the optimal solution,and the artificial selection of the conversion probability value will lead to a decrease in the solution accuracy of the algorithm.In order to solve the above problems,an adaptive flower pollination algorithm integrating multiple strategies is proposed.Firstly,the population is initialized by Logistic mapping to improve the quality of the optimal solution,and the adaptive transformation probability is used to improve the solution accuracy of the algorithm.Secondly,the step size of the Levy flight mechanism in cross-pollination is changed to speed up the convergence speed.Finally,a double-difference mutation strategy is introduced in self-pollination to enhance the ability to jump out of the local optimum.The simulation experiments of three types of test functions show that the improved algorithm has higher solution accuracy and faster convergence speed.Meanwhile,the parameters of the PID controller are optimized with the help of the improved algorithm,which shows a good optimization effect and verifies the feasibility and applicability of the improved algorithm. |