Estimating Passenger Exits via Aggregated Unallocated Counts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing passenger destination estimation systems in mass transit networks face inaccuracies when estimating the number of passengers exiting at small-volume stops due to low signal-to-noise ratios and reliance on incomplete or unreliable origin-destination information.
Innovation Solution
A passenger destination estimation system that aggregates unallocated-exit passengers across preceding stops, calculates weighted probabilities based on origin-destination probabilities, and incorporates ticket OD information and historical data to improve estimation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing systems estimate passengers exiting at each stop independently using validation information, then the estimation process is simple, but the accuracy is poor especially at stops with small numbers of exiting passengers
Solution Approach 1:
The patent merges the estimation of unallocated-exit passengers across multiple preceding stops into a single aggregated calculation. Instead of estimating exits at each stop independently, the system combines data from all preceding stops (S-1, S-2, S-3, etc.) to calculate a total unallocated-exit passenger number, which is then distributed to individual stops using weighted probabilities. This merging approach improves measurement precision by reducing the impact of low signal-to-noise ratios at individual stops with small passenger volumes.
Solution Approach 2:
The patent segments the passenger population into distinct categories: OD-allocated passengers (with known origin-destination information from validation systems) and unallocated-exit passengers (without specific destination information). This segmentation allows the system to apply different estimation approaches to different passenger groups, using actual OD information where available and aggregated probabilistic methods for the remainder, thereby improving overall estimation accuracy while maintaining system manageability.
2Loss of information
If validation information is used to obtain OD information, then some OD data is available, but the information is incomplete or unreliable for many passengers
Solution Approach 1:
The patent applies partial action by utilizing only the portion of passenger flow for which reliable OD information is available (OD-allocated passengers), while using a different approach for the remaining unallocated portion. Rather than attempting to force complete OD information for all passengers through unreliable validation data, the system accepts partial information availability and compensates through aggregated statistical methods, thereby maintaining reliability without being constrained by incomplete validation data.
3Measurement precision
If aggregation of unallocated-exit passengers across multiple preceding stops is performed, then accuracy at small-volume stops improves, but computational complexity increases
Solution Approach 1:
The patent performs preliminary aggregation of unallocated-exit passenger numbers across all preceding stops before distributing them to individual stops. By calculating the total unallocated-exit passenger number once at the aggregation stage and then using pre-computed weighted probabilities for distribution, the system avoids redundant calculations at each individual stop. This preliminary action approach maintains high measurement precision while improving computational efficiency compared to methods that would perform separate aggregations for each stop.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A passenger destination estimation system configured to estimate, from a total number of passengers within a mass transit network, a number of passengers exiting the mass transit network at a given stop of the mass transit network. The system comprises a passenger count unit configured to: obtain a number of passengers arriving at each of a plurality of stops preceding the given stop on one or more routes of the mass transit network; and calculate an unallocated-exit passenger number, which indicates a number of passengers whose exit stop in the mass transit network is unallocated, based on the number of passengers arriving at the plurality of preceding stops. The system further comprises a probability unit configured to: obtain, for each of the preceding stops, an origin-destination, OD, probability of a passenger arriving at that stop and exiting the network at the given stop; calculate a weighted probability of a passenger exiting the network at the given stop that is based on: the OD probability for each of the plurality of preceding stops, the unallocated-exit passenger number, and the number of passengers arriving at each of the plurality of preceding stops; and estimate the number of passengers exiting the network at the given stop based on the unallocated-exit passenger number and the weighted probability.