Idealised models for public transport systems

Abstract

Results are presented from two idealised models for public transport. Models are evaluated for trips via a corridor and for a network system meeting a uniform trip demand.

The network transport model uses a grid-based synchronous system which could serve a whole city with a maximum of one transfer. This model has been proposed by others, but as far as is known no basic results have been published for its operational effectiveness. The analysis has provided a number of mathematical results for this system. The system proposed may have value in its own right, but is introduced to enable the effectiveness of meeting transport demand with transport networks of different density to be calibrated.

An interesting result, which may be new, is that the average trip length for uniform demand in a grid-based city is equal to one sixth of the city perimeter served, independent of grid density.

The results for the linear and network cases were found to be very similar. In both cases, average walk lengths are equal to average stop separation. Minimum trip times, including walk, wait, in-vehicle and transfer elements, were found to occur with a station separation of around 0.5 km. Maximum average speed for the total trip was found to be around 15 km/h.

The network results demonstrated that transport effectiveness increased with reducing vehicle size. Optimum vehicle capacity for 0.5 km spacing was projected to be seven passengers. This would require automatic control to be operationally effective. This makes a case for the consideration of small automatically controlled vehicles for public transport.