Frequency Optimization Models for Reducing Overcrowding Discomfort

Abstract

One of the most important aspects of improving public bus transport attractiveness is reducing overcrowding in buses. However, most of the mathematical models that focus on designing bus services minimize the total social cost without considering the overcrowding discomfort. Further, they are mostly non-linear in nature and use heuristic and meta-heuristic approaches. Thus, they are difficult to understand and use by practitioners. This work addresses these gaps through models that include overcrowding discomfort and are also easy to implement and contextualize by practitioners. The authors develop one non-linear and two linear models to determine the optimum frequency of buses and apply them over a network of 34 routes of Delhi. The results reveal that the existing number of buses in Delhi is not sufficient to cater to the existing travel demand of peak hours, even after their optimum allocation. The authors also present a step-by-step procedure to enable practitioners to determine the minimum additional number of buses required to reduce the target discomfort and waiting time.