Branched Transit Services: An Analysis

Abstract

A method for optimizing the route structure and schedule for zonal transit services has been extended from one‐dimensional linear cases to two‐dimensional branched networks. The values of zone lengths, subroute lengths, and headways which minimize the sum of operator and user costs are determined by calculus in simple cases and a quasi‐Newton computer algorithm in more complex cases. Algebraic relations are derived which provide useful guidelines for optimal system design and greatly simplify sensitivity analysis. It is shown that user wait costs, user access costs, and vehicle operating costs should be equalized to minimize total costs. A case study is used to demonstrate the applicability of the method. The results confirm theoretical relations and indicate that considerable flexibility exists in adapting transit routes to irregular road networks and demand patterns without substantial cost increases.