Institute of Transportation, MOTC-Digital Archive-Transportation Dissertation-Commuter Bus Routing Problem

Digital Archive

:::

Transportation Dissertation

Title	Commuter Bus Routing Problem
Year	2006
Summary	Shu-Shih Chang, 2006.06 Institute of Transportation Technology and Management National Chiao Tung University 　　Commuter Bus Routing Problem (CBRP) is defined as a multi-workplace and multi-vehicle routing problem commonly faced by many companies for their employee daily commuting. CBRP can be considered as a variant of Vehicle Routing Problem (VRP), but traditional solution methods of VRP are not applicable to solve CBRP. School Bus Routing Problem (SBRP) is similar to CBRP, but most of the research on SBRP only considered single-school and single-vehicle. In this study, we presented an integer programming (IP) formulation for CBRP, and developed heuristic methods for solving large scare CBRP. 　　We first formulated an IP model with the objective of minimizing total operating cost, and then generated a bank of 24 small example problems for testing and validation of our IP model. We also developed heuristic methods to solve the CBRP. The heuristic methods consist of two phases. In route construction phase, we first selected seed nodes and applied nearest neighbor procedures. In route improvement phase, we considered surplus vehicle capacity and surplus route time to improve the incumbent solution. Comparing the results of IP model and the heuristic methods, we found that our heuristic methods can solve all the 24 small example problems correctly in about 0.3 seconds. 　　In this study, we also tested a real-world problem of a company in Hsinchu Science Park. There were 14 routes providing commuting services to 244 employees in the case problem, and the annual operating cost was about 5 million. The case problem would have 38,296,776 variables and 2,247,204 constraints in the IP model, and is difficult to obtain the exact solution. We applied our heuristic methods to successfully solve the case problem in 1.5 seconds, and the results could save about 20% total operating cost. 　　In this study, we only considered the objective of minimizing total operating cost. For service consideration, further research may consider the objective of minimizing total travel time or multiple objectives.

Title

Commuter Bus Routing Problem

Year

2006

Summary

Shu-Shih Chang, 2006.06

Institute of Transportation Technology and Management
National Chiao Tung University

　　Commuter Bus Routing Problem (CBRP) is defined as a multi-workplace and multi-vehicle routing problem commonly faced by many companies for their employee daily commuting. CBRP can be considered as a variant of Vehicle Routing Problem (VRP), but traditional solution methods of VRP are not applicable to solve CBRP. School Bus Routing Problem (SBRP) is similar to CBRP, but most of the research on SBRP only considered single-school and single-vehicle. In this study, we presented an integer programming (IP) formulation for CBRP, and developed heuristic methods for solving large scare CBRP.
　　We first formulated an IP model with the objective of minimizing total operating cost, and then generated a bank of 24 small example problems for testing and validation of our IP model. We also developed heuristic methods to solve the CBRP. The heuristic methods consist of two phases. In route construction phase, we first selected seed nodes and applied nearest neighbor procedures. In route improvement phase, we considered surplus vehicle capacity and surplus route time to improve the incumbent solution. Comparing the results of IP model and the heuristic methods, we found that our heuristic methods can solve all the 24 small example problems correctly in about 0.3 seconds.
　　In this study, we also tested a real-world problem of a company in Hsinchu Science Park. There were 14 routes providing commuting services to 244 employees in the case problem, and the annual operating cost was about 5 million. The case problem would have 38,296,776 variables and 2,247,204 constraints in the IP model, and is difficult to obtain the exact solution. We applied our heuristic methods to successfully solve the case problem in 1.5 seconds, and the results could save about 20% total operating cost.
　　In this study, we only considered the objective of minimizing total operating cost. For service consideration, further research may consider the objective of minimizing total travel time or multiple objectives.

Count Views:333

Pre Page