利用粒子群优化算法进行桥梁维修管理计划的优化

制定桥梁维修管理计划是一项非常繁杂的工程优化难题,使用常用的优化算法很难取得满意的结果。利用耐荷性和耐久性作为桥梁的健康指数,考虑维修方案和维修费用的问题,用费用最小化和品质最大化2种方案建立了桥梁维修管理的优化模型。探讨利用粒子群优化算法(PSO)求最优桥梁维修管理计划的解的可能性,并与作者开发系统中的遗传算法(SGA)和免疫遗传算法(IA)进行了比较,运用多样度的概念说明了粒子群优化算法(PSO)在解决这类问题的先进性。结果表明,粒子群优化算法(PSO)对于桥梁维修管理计划的优化是一种普适高效的算法;而且,考虑维修的管理期间越长,应用粒子群优化算法求解问题收敛性与其他2种方法相比显得更好,得到准最优解的频率也更高。

Optimization of Maintenance Planning for Existing Bridge Using PSO

How to draw up a plan of bridge maintenance management is a very complicated optimization problem.It is very difficult to achieve satisfactory results using usual optimization algorithm.An optimum model for bridge maintenance planning considering both maintenance cost minimization and quality maximization is established by adopting load-carrying capability and durability as the bridge rating indices.The optimum possibility by using Particle Swarm Optimization(PSO) is discussed and compared with Simple Genetic Algorithms(SGA) and Immune Algorithms(IA).It is illuminated that PSO is an advanced method in solving this problem by using concept of diversity.It is found that PSO is a widely applicable and effective method for bridge maintenance planning.At the same time,relative to the other methods,the longer the maintenance period become,the better the reliability of the solution using PSO is and more easily the optimum solution can be found.