ORIGIN DESTINATION ESTIMATION BASED ON VEHICLE TRAJECTORY DATA

• US 20160180705A1
• Filed: 12/18/2014
• Published: 06/23/2016
• Est. Priority Date: 12/18/2014
• Status: Active Grant

• Current Assignee: SAP SE
• Original Assignee: Jing Liu, Qiwei Zhang, Ye Jin
• Inventors: LIU, Jing, JIN, Ye, ZHANG, Qiwei, WANG, Zhongcun
• Granted Patent: US 9,564,048 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G01C 21/26 specially adapted for navigation in a road network
• CPC Class Codes: G08G 1/0116 from roadside infrastructure, e.g. beacons
• CPC Class Codes: G08G 1/0125 Traffic data processing
• CPC Class Codes: G08G 1/0129 for creating historical data or processing based on historical data
• CPC Class Codes: G08G 1/017 identifying vehicles G08G1/015, G08G1/054 take precedence
• CPC Class Codes: G08G 1/0175 by photographing vehicles, e.g. when violating traffic rules

Abstract

A framework for origin-destination (OD) analysis of vehicle trajectory data is described herein. In accordance with one aspect, a vehicle trajectory dataset is provided to an OD analyzer. The vehicle trajectory dataset includes vehicle trajectory data collected from a sensor network having a plurality of sensor stations for detecting vehicles. The sensor stations of the sensor network are distributed in a geographical area of interest, where the vehicle trajectory data include trajectories of vehicles captured by the sensor network. The vehicle trajectory dataset may be analyzed by the OD analyzer to determine an origin and a destination of trips for trajectories of the vehicles in the vehicle trajectory dataset. The analysis includes calculating a probability distribution of travel time between sensor pairs of the sensor network of a number of (dropped-out) intermediate stations, and determining a stop probability between a station pair in the trajectories of the vehicles, where a stop is a destination of a previous trip and an origin of a next trip in the trajectories.

Claims

1. A computer-implemented method for origin-destination (OD) analysis of vehicle trajectory data, comprising:

providing a vehicle trajectory data set to an OD analyzer, the vehicle trajectory data set includes vehicle trajectory data collected from a sensor network having a plurality of sensor stations for detecting vehicles, the sensor stations of the sensor network are distributed in a geographical area of interest, wherein the vehicle trajectory data include trajectories of vehicles captured by the sensor network; and

analyzing the vehicle trajectory data set by the OD analyzer to determine an origin and a destination of trips for trajectories of the vehicles in the vehicle trajectory data set, wherein the analysis comprises,

calculating a probability distribution of travel time between sensor station pairs of the sensor network over a number of (dropped-out) intermediate stations, and

determining a stop probability between a station pair in the trajectories of the vehicles, a stop is a destination of a previous trip and an origin of a next trip in the trajectories.

2. The computer-implemented method of claim 1 wherein a sensor station comprises a fixed traffic sensor.

3. The computer-implemented method of claim 2 wherein the fixed traffic sensor comprises an RFID sensor, a license plate recognition sensor or both.

4. The computer-implemented method of claim 1 wherein a data record of the trajectory data in the data set comprises structured data including:

vehicle id, the vehicle id is a unique identifier for identifying a vehicle;

sensor id, the sensor id is a unique identifier for a sensor station in the sensor network, wherein the sensor id is used to determine location information using a look up table, wherein location information includes longitude and latitude coordinates of the sensor; and

a time stamp, the time stamp indicates when the vehicle identified by the vehicle id is captured by the sensor station identified by the sensor id.

5. The computer-implemented method of claim 1 wherein a data record in the trajectory data set comprises structured data including:

vehicle id, the vehicle id is a unique identifier for identifying a vehicle;

location information, wherein the location information includes longitude and latitude coordinates of the sensor station which captures the vehicles identified by the vehicle id; and

a time stamp, the time stamp indicates when the vehicle is captured by the sensor station.

7. The computer-implemented method of claim 5 wherein the OD analyzer comprises a processor module, the processor module analyzes pre-processed data from the pre-processor module.

8. The computer-implemented method of claim 7 wherein the processor module performs:

calculating the probability distribution of travel time between sensor station pairs of the sensor network over a number of (dropped-out) intermediate stations; and

determining a stop probability between a station pair in the trajectories of the vehicles, a stop is a destination of a previous trip and an origin of a next trip in the trajectories.

9. The computer-implemented method of claim 8 wherein calculating the probability distribution of travel time between sensor station pairs of the sensor network over a number of (dropped-out) intermediate stations comprises:

determining, by an intermediate step predictor (ISP) unit, a probability of an actual number of intermediate steps k between sensor station pairs; and

determining, by a time travel predictor (TTP) unit, for a given number of k steps with dropped-out intermediate sensor stations, a probability that a trip between sensor station pairs with travel time more extreme than Δ

t, where Δ

t is the travel time between the station pair of a vehicle.

10. The computer-implemented method of claim 9 wherein determining the probability of an actual number of intermediate steps k between sensor station pairs comprises:

determining a probability of a number of k intermediate steps using Eq 1

11. The computer-implemented method of claim 10 wherein determining the probability of a stop between station pairs comprises using Eq 4

12. The computer-implemented method of claim 11 wherein a stop between station pairs is determined when the probability of a stop is less than a probability stop threshold (P_ST).

13. The computer-implemented method of claim 8 wherein determining the probability of a stop between station pairs comprises determining by a stop determination (SD) unit.

6. The computer-implemented method of claim 6 wherein the OD analyzer comprises a pre-processor module, the pre-processor module sorts and aggregates vehicle trajectory data based on:

vehicle id; and

station pairs traveled by vehicles in the trajectory data.

14. An origin-destination (OD) analysis system comprising:

a data storage module for storing a vehicle trajectory data set, the vehicle trajectory data set includes vehicle trajectory data collected from a sensor network having a plurality of sensor stations for detecting vehicles, the sensor stations of the sensor network are distributed in a geographical area of interest, wherein the vehicle trajectory data include trajectories of vehicles captured by the sensor network;

a pre-processor module, wherein the pre-processor module processes the vehicle trajectory data set by sorting and aggregating the vehicle trajectory data; and

a processor module, the processor module analyzes the processed trajectory data from the pre-processor to

calculate a probability distribution of travel time between sensor station pairs of the sensor network over a number of (dropped-out) intermediate stations, and

determine a stop probability between a station pair in the trajectories of a vehicle, a stop is a destination of a previous trip and an origin of a next trip in the trajectories.

15. The OD analysis system of claim 14 wherein a sensor station comprises a fixed traffic sensor.

16. The OD analysis system of claim 14 wherein a data record of the vehicle trajectory data in the data set comprises:

vehicle id, the vehicle id is a unique identifier for identifying a vehicle;

sensor id, the sensor id is a unique identifier for a sensor station in the sensor network, wherein the sensor id is used to determine location information using a look up table, wherein location information includes longitude and latitude coordinates of the sensor station; and

a time stamp, the time stamp indicates when the vehicle identified by the vehicle id is captured by the sensor station identified by the sensor id.

17. The OD analysis system of claim 14 wherein the pre-processor module sorts and aggregates vehicle trajectory data based on:

vehicle id; and

station pairs traveled by vehicles in the trajectory data.

18. The OD analysis system of claim 17 wherein:

the processor comprises

an intermediate step predictor (ISP) unit for determining a probability of an actual number of intermediate steps k between sensor station pairs,

a time travel predictor (TTP) unit for determining, for a given number of k steps with dropped-out intermediate sensor stations, a probability that a trip between sensor station pairs with travel time more extreme than Δ

t, where Δ

t is the travel time between the station pair of a vehicle; and

the ISP and TTP units calculate the probability distribution of travel time between sensor station pairs of the sensor network over a number of (dropped-out) intermediate stations.

19. The OD analysis system of claim 18 wherein the processor comprises a stop determination (SD) unit for determining the probability of a stop between station pairs.

20. A non-transitory computer-readable medium having stored thereon program code, the program code executable by a computer for performing origin-destination (OD) analysis of vehicle trajectory data comprising:

pre-processing a vehicle trajectory data set by a pre-processor of an OD analyzer, the vehicle trajectory data set includes vehicle trajectory data collected from a sensor network having a plurality of sensor stations for detecting vehicles, the sensor stations of the sensor network are distributed in a geographical area of interest, wherein vehicle trajectory data include trajectories of vehicles captured by sensor network; and

analyzing the vehicle trajectory data set by a processor of the OD analyzer to determine an origin and a destination of trips for trajectories of the vehicles in the vehicle trajectory data set, wherein analyzing comprises,

calculating a probability distribution of travel time between sensor station pairs of the sensor network over a number of (dropped-out) intermediate stations,

determining a stop probability between a station pair in the trajectories of the vehicle, a stop is a destination of a previous trip and an origin of a next trip in the trajectories.