Entropy-STEP Multiobjective Trade-Off Analysis Method for Optimal Transportation Investment Decisions

Abstract

The transportation decision-making process is complicated in that it often involves conflicting interests and expectations of different stakeholders. To ensure achieving sound decisions, various performance criteria in economic, social, and environmental dimensions are typically utilized to assess and prioritize investment alternatives. This paper introduces an entropy-STEP (e-STEP, where STEP is the uppercase of the word step) multiobjective trade-off analysis method for optimal transportation investment decisions. Its entropy component derives normalized relative importance for multiple noncommensurable performance criteria by refining the prior relative importance with additional context-dependent information on the relative importance. The STEP component conducts trade-off analysis and makes efficient investment decisions using a multiobjective optimization formulation. The model is converted to a linear programming model with refined relative importance for performance criteria to derive efficient solutions. The proposed method is applied in a computational study using data on a state-level toll highway capital investment program in the United States. The same data set is used to execute the traditional STEP method and surrogate worth trade-off (SWT) method used in the current practice. Cross comparisons of decision outcomes reveal that the e-STEP method outperforms the other two methods, suggesting that it could be adopted by transportation agencies to develop capital investment programs.