基于并行NSGA-Ⅱ算法的桥梁网络养护策略优化

制定辖区全体桥梁的养护策略是在役桥梁管养方法研究的前进方向。给出桥梁网络养护策略的优化目标与约束方程,建立网级桥梁策略优化框架。引入带精英策略的非支配排序遗传算法(NSGA-Ⅱ),将其与并行遗传算法(PGA)相结合,提出基于并行NSGA-Ⅱ的网级养护策略优化方法。采用约束联赛法处理单座桥梁的最低性能约束问题,通过修改支配关系的定义改进原有NSGA-Ⅱ算法。应用于工程案例,输出结果为一系列Pareto最优解集,管理方可根据自身实际方便选择。算例分析验证该框架的有效性,以及建议算法的可靠性与高效性。

Optimization of maintenance strategy for bridge networks using parallel NSGA-Ⅱ algorithm

The development of bridge maintenance strategies for the entire management area is an advanced research direction for bridge management methods.The objective function and constraint equations of bridge network maintenance strategy were defined.The network-level bridge strategy optimization framework was established.The non-dominated sorting genetic algorithm with elitist strategy(NSGA-Ⅱ) was introduced,and combined with the parallel genetic algorithm(PGA).The parallel NSGA-Ⅱ method was proposed for the network-level maintenance strategy optimization.The constraint tournament method was employed to deal with the minimum performance constraint of single bridge.The NSGA-Ⅱ was improved by modifying the definition of dominance relations.A case study was analyzed by using the proposed method.The result was a range of Pareto optimal solution set.The management department can easily make the selection according to the actual demand.The example analysis verified the validity of the proposed framework,the reliability and efficiency of the proposed algorithm.