Self-adaptive particle swarm optimization method solving traveling salesman problem

Abstract

The invention discloses a self-adaptive particle swarm optimization method solving a traveling salesman problem. On the basis of existing standard particle swarm optimization, backward learning is adopted for initializing particle population, and an inertia weight w and a learning factor c of particle swarm optimization are regulated adaptively along with increase of number of iteration, so that the optimization capability of particle swarm optimization is improved; and in the later period of particle swarm optimization, chaotic local search is introduced so as to prevent particle swarm optimization from falling into local optimum. The improved algorithm is tested through a standard function, an obtained result is obviously superior to that of the standard PSO, and improved particle swarmoptimization is applied to optimization solution of the traveling salesman problem, so as to obtain a better optimization result.

Description

technical field [0001] The invention belongs to the field of problem optimization, and relates to an adaptive particle swarm optimization method for specifically solving the traveling salesman problem, and relates to an adaptive particle swarm optimization method based on reverse learning and chaotic local search for solving the traveling salesman problem. Background technique [0002] In the process of the development of human society, optimization problems generally exist in the aspects of people's life, study and work. For example, in life, people always hope to find a fast and convenient route from one place to another; in production, they hope to have the lowest energy consumption, etc. [0003] Reynolds proposed the Boid model by simulating the flight behavior of birds in the mid-1980s, and simulated the running track of birds in the computer. Each individual in the model has the ability to perceive, and can perceive changes in itself and the surrounding environment ,...

AI Technical Summary

Problems solved by technology

[0008] In the standard PSO algorithm, the generation of initial particles is generally to randomly generate particles of a certain size within a given range, which will lead to the loss of diversity of initial particles, and make the initial solution of particles poor, which will affect the convergence effect of the algorithm. ;The inertia weight w and the acceleration factor c are both constant, which is not conducive to the evolution of the group; in the later stage of the algorithm, the particles tend to mature prematurely and fall into local extremum

Images