To avoid overcrowding of vehicle journey items, boarding or passenger transfers are prevented as soon as the capacity of the vehicle is reached. The passengers concerned only become aware of the overcrowding of the vehicle at this point and cannot include it in their route choice before. Passengers search for and choose alternative connections at the place and time they are denied boarding. This can be the next vehicle journey of the same route or a connection via other lines.
The aim is to show where and when connections are interrupted due to a lack of vehicle capacity. The procedure shows the relations that are particularly affected and calculates the travel time losses of the passengers affected.
How the procedure works
The procedure is based on the results of a timetable-based assignment. It can be carried out directly after an assignment or be based on an existing assignment result. Prerequisites are saved connections and their volumes. The procedure does not include any iterative steps and follows this sequence:
1. The vehicle journey items are sorted according to their start time.
2. Starting with the first vehicle journey item, this step is carried out for all subsequent items.
- The vehicle journey item is checked for overcrowding. If it is overcrowded, all connections that board or change to this vehicle journey item are reduced by the share of the overcrowding. All connections are equally affected proportionally. The reduction is carried out on past and future vehicle journey items. The previous course of the connection is saved.
-
Starting from the location and time of the vehicle journey item, i.e. the stop area and the departure time (+1s), the passengers concerned search for alternative connections. This also takes into account vehicle journey items that are still overcrowded. This reflects the fact that the volumes may still change in the future and the passenger has no knowledge of the volume capacity ratio. Separate cut off rules are defined for the search to reduce the set of connections.
- A path is chosen and the volume is propagated along the alternative routes through the network.
-
The alternative routes are linked to the previous sections and the attribute Number of missed connecting journeys counts the number of interruptions in a connection.
The alternative connection search only takes place up to an adjustable minimum share. This limits the number of connections to be processed. Volumes below these shares are restored to the original connection. This can lead to minimal overcrowding of the vehicle journey items.
3. The linked routes are saved in the connection set of the assignment. Volumes of passengers who have not found an alternative route at a transfer are no longer included in the set of connections. This leads to an underestimation of the volume of the path legs up to the interrupted transfer. The number of passengers without alternative connection is indicated.
Results and skims of the procedure
The output attributes of the procedure consist, in addition to the extended amount of connections, of indicators such as the number and percentage of passengers who were unable to board a vehicle. These are further divided into passengers who found an alternative connection or no alternative connection.
The number of passengers who cannot board:
Number of fail to board = Volume - Capacity of the vehicle
The share is calculated based on passengers who boarded and passengers who were denied boarding.
Number of fail to board = Number of fail to board/Passengers boarding + Number of fail to board
This value is 0% if the capacity has not been reached and is 100% if no passenger has been able to board the vehicle journey item.
The risk to which the passenger is exposed is also calculated. Risk is composed of probability and a certain threat. In this case, the share of people who were refused boarding and the extension of the travel time they would have to experience. These skims are first calculated at connection level for each transfer.
The boarding risk per person of a connection C is calculated as the sum of all transfers u and the boarding risk tC,e. The probability aC,u indicates the share of people who fail to make the transfer u in connection C. The resulting travel time extension t indicates the time lost compared to the initial arrival time at the destination.
The total risk of a connection is the sum of the product of the boarding risk per person and the existing volumes on the connection.
These skims are also available in the respective aggregations at connecting journeys, stop points, path legs, and as a skim matrix.
