面向低碳理念的桥梁群多目标维护决策优化研究

Multi-objective Maintenance Optimization Strategy for Bridge Networks Oriented to Low-carbon

为解决低碳理念下路网桥梁群维护策略制定难题,提出了基于改进第三代非支配排序遗传算法(NSGA-Ⅲ算法)的桥梁群多目标维护决策计算框架。首先,基于碳排放因子理论提出了路网桥梁群维护阶段碳排放计算方法,该方法涵盖了加固原材料生产、车辆通行以及辅助设施三级碳排放来源。以单个桥梁和节点连通可靠度为基准,建立了融合碳排放量、综合维护费用等指标的桥梁群多目标维护优化模型。其次,通过遗传K均值聚类法产生最佳聚类中心替代原NSGA-Ⅲ的参考点,利用基于遗传参数的动态交叉变异机制增强种群的多样性,提出了改进NSGA-Ⅲ算法;同时,结合组合权重逼近理想解排序法实现考虑决策者偏好的桥梁网络多目标维护优化模型的快速、精准求解。最后,以某地区桥梁网络为背景,分析了考虑决策者偏好的维护策略,并对比了不同目标组合对维护策略的影响。研究结果表明:不同的权重分配和不同的目标组合均会导致桥梁网络维护策略产生差异;相比于侧重经济效益的策略,侧重环境影响的策略费用增加了23.93%,但碳排放量减少了28.74%;在桥梁网络维修决策中,细化决策目标有助于获得更均衡的维护策略。

A calculation framework for multi-objective maintenance strategies based on the improved non-dominated sorting genetic algorithm-Ⅲ(NSGA-Ⅲ) is proposed to address the formulation maintenance strategies for bridge networks oriented toward the low-carbon principle. First, a carbon-emission calculation method for bridge networks in the maintenance stage was developed based on the theory of carbon-emission factors. Three carbon emission sources were considered: production of reinforced raw materials, vehicle traffic, and auxiliary facilities. Based on the connectivity reliability of individual bridges and nodes, a multi-objective maintenance optimization model that integrates carbon emissions and comprehensive maintenance costs was established. Second, the multi-objective maintenance optimization model of the bridge network considering the preferences of decision makers was quickly and accurately solved using an improved NSGA-Ⅲ algorithm and the combined weight technique for the order of preference by similarity to the ideal solution. The improved NSGA-Ⅲ algorithm was developed using optimal cluster centers generated using the Genetic K-means clustering method instead of the original reference points. It also incorporates a dynamic crossover and mutation mechanism based on genetic parameters to enhance population diversity. Finally, a maintenance strategy considering the preferences of decision makers was analyzed, and the influence of different objective combinations on the maintenance strategy was compared through a case study of a regional bridge network. The results show that different weight distributions and objective combinations can lead to different maintenance strategies for bridge networks. Compared to the strategy focusing on economic benefits, the strategy focusing on environmental impact has a 23.93% increase in cost and a 28.74% reduction in carbon emissions. For the maintenance decision-making of bridge networks, the refinement of decision objectives contributes to the comprehensive elucidation of the decision problem and results in the attainment of a more balanced maintenance strategy.