Improvement Research And Application Of Ant Colony Algorithm And Leapfrog Algorithm

The ant colony algorithm and the leapfrog algorithm are typical intelligent optimization algorithms.The convergence speed is fast,the global optimization ability is strong,and it is easy to implement.However,the local search ability is relatively weak and easy to premature.This paper mainly studies the improvement,mixing and application of ant colony algorithm and leapfrog algorithm.The details are as follows:According to the shortcomings of ant colony algorithm,such as premature aging and poor local search ability,the iterative local search strategy is introduced into ant colony algorithm.The basic idea of the new algorithm is:starting from the initial solution,using the ant colony algorithm for local search,such as falling into local optimum,then generating a perturbation solution as a new initial solution and then performing a local search,and proceeding to the next iteration according to the acceptance rules.The local optimal solution.The improved algorithm is applied to two-dimensional path planning.Numerical experiments show that this improved algorithm has better local convergence than the basic ant colony algorithm,and can obtain better path than the basic ant colony algorithm.Aiming at the defect of population diversity in the late evolution of the frog leaping algorithm and being trapped in the local optimal solution,an adaptive mutation frog leaping algorithm is proposed.The basic idea is to establish an adaptive mutation selection mechanism according to the rate of change of function;when the rate of change of function is large,Gaussian mutation is used to improve the local convergence ability of the algorithm;when the rate of change of function is small,the algorithm may fall into local convergence.Using the Cauchy mutation causes the algorithm to jump out of the local optimum.Numerical experiments show that this adaptive mutation selection mechanism not only improves the local convergence of the leapfrog algorithm,but also avoids the premature phenomenon to a large extent.According to the characteristics of ant colony algorithm and leapfrog algorithm,a hybrid algorithm of ant colony and leapfrog is presented.The basic idea is to apply the ant colony algorithm to find the optimal solution of the stage,and then use it as the initial frog group.Adapt to the mutated leapfrog algorithm to continue to optimize it.The hybrid algorithm is applied to the capacity constrained vehicle routing problem,and the effectiveness of the new algorithm is verified.Figure[18]Table[4]Reference[47]...