||"Improving bus timetables"
||Guillaumet, Joan Ballano
||Rich, Jeppe (Traffic Modelling Group, Centre for Traffic and Transport, Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark)
Nielsen, Otto Anker (Traffic Modelling Group, Centre for Traffic and Transport, Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark)
||Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark
||It is widely known the importance of the optimization of the cost of the public transport. This includes the time saving of the users as well as the real cost of the infrastructures. Literature shows that genetic algorithms are useful tool to reach the optimum in similar situations [ref. 01]. GA’s method creates a set of solutions, and in the same way as the real genetics, evolves solutions trough
several generations until the optimum. This process makes the solution space explored much more extensive, thus becoming a good method for this problem. Applying GA’s to an existing single bus line, results showed that the behaviour of the GA converges to a minimum cost and the optimum timetable is similar to the existing one. These satisfactory results motivated the application of GA to a
bus network. Although the optimal timetable obtained in this simulation does not bring relevant information since it is not based on a real network, the trend of the GA evolution is to converge in a minimum cost, in the same way as the optimization of a single bus line. In conclusion GA’s can be considered as a useful tool to optimize bus network scheduling problems. Nevertheless, further work in this subject using real networks and reliable data is encouraged so that more conclusive and realist results can be obtained.
Creation date: 2007-07-06
Update date: 2007-07-06