Abstract: Ant Colony Optimization for continuous domains（ACOR） is easy to fall into local optimum when dealing with high dimensional problems, and its convergence rate is slow. So ACO for continuous domains algorithm is put forward to solve these problems. In the proposed algorithm, the solution is divided into two parts, the optimal solution and the inferior solution, and the number of the optimal solution and the inferior solution is adjusted dynamically in the iterative process. For the optimal solution, the global search strategy is used to pre process, which can improve the convergence speed and convergence accuracy of the algorithm. For inferior solution, using a random search strategy for pretreatment, this can expand the search scope, and enhance search ability. The test results show that the DPHACO algorithm can effectively improve the convergence speed and the quality of the solution. Compared with continuous ant colony algorithm and other intelligent optimization algorithms, the proposed algorithm is more effective and better than the global search capability.

Key words: Ant Colony Optimization（ACO） algorithm, dynamic partition, global search, random search, pretreatment

摘要： 连续域蚁群优化算法在处理高维问题时易陷入局部最优，而且收敛速度较慢。针对这些问题，提出了一种改进的连续域蚁群优化算法。该算法将解划分为优解和劣解两部分，并在迭代过程中动态调整优解和劣解的数目。对于优解，利用全局搜索策略进行预处理，这样能提高算法的收敛速度和收敛精度。对于劣解，则利用随机搜索策略进行预处理，这样能扩大搜索范围，增强搜索能力。通过标准测试函数对所提算法进行测试，结果表明改进策略能够有效提高连续域蚁群优化算法的收敛速度并改善解的质量。

关键词: 蚁群优化算法, 动态划分, 全局搜索, 随机搜索, 预处理