Multi-strategy Heuristic Moth-flame Optimization Algorithm And Its Application

Moth-flame optimization algorithm(MFO)is a new meta-heuristic algorithm that simulates the lateral navigation and positioning mechanism of moths in nature at night.This algorithm is mainly composed of moth search mechanism and flame adaptive reduction mechanism.It has the advantages of simple structure and easy implementation,less parameters,and fast search speed.It has been widely used in various optimization problems and and has achieved great success.However,the researchers found that there are some defects in the moth-flame optimization algorithm with the deepening of research,such as the lack of diversity of the population,the lack of global search ability of the algorithm,easy to fall into local optimization or premature convergence.In view of the shortcomings of the mothflame optimization algorithm,this paper analyzes and improves the quality of population generation and the mining ability of the algorithm.The improved algorithm is applied to cluster analysis and multi-threshold image segmentation to improve the performance of the moth-flame optimization algorithm and expand its application range.The main research contents of this paper are as follows:(1)Aiming at the shortcomings of traditional K-means clustering algorithm,such as dependence on the initial cluster center and easy to fall into local optimum,a quantum-inspired moth-flame clustering optimizer with enhanced local search strategy(QLSMFO)is proposed.The rotation angle of the quantum rotary gate(QRG)is updated by the differential evolution algorithm,which adaptively guides the moth to the direction of the better solution.The addition of leapfrog algorithm enhances the local search strategy and improves the accuracy of the algorithm.Finally,Levy flight strategy is used to accelerate the convergence speed of the algorithm.QLSMFO is tested on 10 famous UCI benchmark data sets.The simulation results show that the proposed algorithm is superior to other swarm intelligence optimization algorithms in terms of convergence speed,accuracy and robustness.(2)Aiming at the problem of multi-threshold image segmentation,the accuracy of segmentation will be affected with the increase of threshold,and the computational complexity will increase exponentially,an improved moth-flame optimization algorithm(IMFO)based on Kapur entropy is proposed.In the initialization phase of the algorithm,the opposite learning strategy(OBL)is introduced to generate high-quality moth population,and the boundary constraint processing mechanism is introduced to bring back the moths beyond the boundary of the search space at the time of update to the boundary of the search space to obtain better candiate solutions.The simulation results show that the IMFO is superior to other algorithms in terms of fitness value,PSNR,SSIM,FSIM,and has a good effect on multi-threshold image segmentation.