Managers of civil infrastructure systems are facing the challenge to effectively and efficiently manage the ageing and deteriorating asset portfolio while accommodating ever increasing demand resulting from socioeconomic development and climate change. In this article, a novel approach using robust optimisation is proposed for risk-based portfolio management of civil infrastructure assets under deep uncertainties associated with climate change. Robust optimisations based on minimax loss and minimax regret criteria are employed to yield intervention plans that can effectively reduce the long-term risk under different realisations of future scenarios. Through sensitivity analysis, it is found that the obtained solutions from robust optimisation are less sensitive to the factors that require long-term predictions (e.g. change in future hazard and deterioration conditions) than those that do not (e.g. the preference of the decision-maker on user cost and the efficiency of intervention actions). Therefore, the proposed approach is especially suitable for management under deep uncertainty. The proposed approach is applied to plan intervention actions for bridges over the Lehigh River, Pennsylvania under deterioration and flood-induced scour, both of which are affected by climate change.

土木基础设施系统的管理者面临着如何有效，有效地管理老化和不断恶化的资产组合，同时适应因社会经济发展和气候变化而不断增长的需求的挑战。在本文中，提出了一种使用鲁棒优化的新颖方法，用于在与气候变化相关的严重不确定性下基于风险的民用基础设施资产投资组合管理。基于最小最大损失和最小最大后悔标准的稳健优化被用于制定干预计划，该计划可有效降低在未来情况的不同实现下的长期风险。通过敏感性分析，发现通过鲁棒优化获得的解决方案对需要长期预测的因素不敏感（例如，改变未来的危害和恶化条件）（例如，决策者对用户成本和干预措施效率的偏好）。因此，所提出的方法特别适合在不确定性较大的情况下进行管理。拟议的方法适用于规划宾夕法尼亚州利哈伊河上的桥梁在受气候变化影响的恶化和洪水冲刷下的干预措施。