基于改进蚁群算法的渡槽结构优化设计

Structural optimization design of aqueduct by improved Ant Colony Algorithm

渡槽结构优化设计属于混合离散变量优化问题,针对利用一般连续变量方法进行离散变量优化设计的不足,对基本蚁群优化算法进行了改进:引入蚁群更新、沿途搜索等策略,在搜索过程中对设计变量进行工程化处理,蚂蚁按处理后的变量进行离散搜索。开发了基于改进的蚁群优化算法的混合离散变量渡槽优化设计的程序。实例研究表明,该算法对优化设计问题的特性无特殊要求,而且程序运行可靠,全局收敛能力强。

Structural optimization design of aqueduct belongs to hybrid discrete variable optimization. In order to avoid deficiency of discrete variable optimization design using continuous variable method, the Basic Ant Colony Algorithm （ACA） is improved by introducing rebirth of ant colony and strategy of searching on the way. In the searching process, design variables are treated for engineering practice, and ants search discretely according to the treated variables. The improved ACA program for structural optimization design of aqueduct with hybrid discrete variables is developed. The results show that this algorithm has no special requirements on the characteristics of optimal design, and the program - running is reliable with strong ability of overall convergence