Ranking Non-Dominated Solutions in Automated Highway Design Using the Analytic Network Process (ANP)

Abstract

Construction of new highway projects needs a lot of consideration to be taken into account during the design process. The complexity of the process and its repetitive (trial and error) nature put forward the need for the development of computer aided/automated design. During the past decades, many researchers have been interested in this problem. The most advanced research in automated highway alignment design is multi-objective 3D alignment optimization which produces a set of nondominated solutions (Pareto Front Optimality). This paper presents the use of Analytic Network Process (ANP) methodology to prioritize alignments from a set of non-dominated solutions. ANP provides comprehensive framework for the assessment of highway alignment design. From the result of multi-objective optimization process, alignments are provided with their objective values. In this study, the construction cost, user cost, environmental impact and social impact are the main objectives with sub-objectives inside. Based on the valuations from experts and previous literature reviews, relative priorities between objectives are set. The priorities in feedback loops are determined based on dominant features of each alternative. A case study was conducted to investigate the efficiency of the model. The alignments from a set of non-dominated solutions using multi-objective optimization were selected using the clustering technique. Finally, the ANP was used to determine the priorities of alternatives (alignments). ANP could further assist the decision makers to prioritize non-dominated solutions according to their preferences.