蚁群算法中挥发系数与迭代次数和最优路径长度关系研究

The Study on Correlation of Volatilization Coefficient and Iterations,Length of Optimal Path in Ant Colony Algorithm

蚁群算法中挥发系数ρ对算法的收敛速度和全局最优路径有很大的影响。以TSP(Traveling Salesman Problem)问题为例,深入探讨了挥发系数与收敛所需迭代次数、全局最优路径长度之间的关系。首先通过对相关文献提供的研究进行归纳和总结,提出了挥发系数与迭代次数、最优路径长度之间的关系假设。然后编程读取不同的挥发系数值对应的迭代次数与最优路径的长度,并对所读取的数据进行分析拟合。最后验证了所提出的关系假设的正确性。

In ant colony algorithm,volatilization coefficient has a great influence on convergence rate and global optimal path.The correlation between volatilization coefficient and iterations required for convergence,length of global optimal path,taking TSP as an example are explored.After summarizing the correlative study,the hypothesis of the correlation between volatilization coefficient and iterations,length of optimal path are proposed.Then different values of volatilization coefficient by programming are read the iterations and length of optimal path corresponding.Lastly the validity of the proposed hypothesis by analyzing and fitting the obtained data is verified.