SGER: Soft Computing-Based Life-Cycle Cost Analysis of Transportation Infrastructure Investments

这项探索性研究小额赠款（SGER）侧重于使用软计算开发实用的经济分析工具，以支持交通基础设施资产管理。本研究的具体目标是：(1)建立将软计算技术纳入基础设施资产生命周期成本分析的总体框架；(2)制定用于生命周期成本分析的混合软计算算法原型；(3)通过简单的示例将该算法与传统的生命周期成本分析工具进行比较，以评估其实际潜力。围绕生命周期成本分析（LCCA）的主要问题之一是在工程经济分析中考虑的物理和经济方面的不确定性和主观性的处理。这已经在文献中使用概率方法进行了广泛的讨论。然而，由于某些不确定性具有模糊性，理论上，软计算应用比概率方法更合适。软计算为处理随机和模糊的不确定性提供了一种正式的方法，因此有望产生更健壮、可靠和稳定的结果。此外，以软计算为基础的LCCA系统可以促进将需要主观评价的非货币因素纳入分析。对生命周期成本分析工具进行修改以处理歧义，有望提高基础设施管理过程的效率。虽然这项工作将集中在一个特定的交通基础设施资产，路面，研究应该有更广泛的影响，由于方法和算法的推广到其他类型的基础设施资产和相关领域。通过使用这些工具获得的知识将有助于各机构了解基础设施管理决策的技术和经济影响，以及有效管理和更新现有基础设施系统的重要性。

This Small Grant for Exploratory Research (SGER) focuses on the use soft computing for developing practical economic analysis tools to support transportation infrastructure asset management. The specific objectives of the research are to: (1) develop an overall framework for the incorporation of soft computing techniques in the life-cycle cost analysis of infrastructure assets, (2) formulate a prototype hybrid soft computing algorithm for life-cycle cost analysis, and (3) compare the algorithm against traditional life-cycle cost analysis tools using simple examples to assess its practical potential. One of the main concerns surrounding life-cycle cost analysis (LCCA) is the treatment of the uncertainty and subjectivity in the physical and economic aspects considered in the engineering economic analysis. This has been addressed extensively in the literature using probabilistic approaches. However, since some of the uncertainty is of an ambiguous nature, soft computing applications are theoretically more appropriate than the probabilistic methods. Soft computing provides a formal approach for the treatment of both random and ambiguous uncertainty and thus is expected to yield more robust, reliable, and stable results. Furthermore, a soft computing-based LCCA system could facilitate the incorporation of non-monetary factors, which require subjective assessments, into the analysis.Life-cycle cost analysis tools modified to treat ambiguity are expected to increase the efficiency of the infrastructure management process. Although the work will concentrate on a specific transportation infrastructure asset, pavements, the research should have a broader impact due to the generalizability of the methods and algorithms to other types of infrastructure assets and related areas. The knowledge acquired through the use of these tools will help agencies understand the technical and economical implications of infrastructure management decisions and the importance of efficiently managing and renewing the existing infrastructure systems.