Method for Model Construction for a Travel-Time Database

• US 20150046387A1
• Filed: 08/15/2014
• Published: 02/12/2015
• Est. Priority Date: 02/17/2012
• Status: Active Grant

• Current Assignee: Bayerische Motoren Werke AG
• Original Assignee: Bayerische Motoren Werke AG
• Inventors: KATES, Ronald, BELZNER, Heidrun
• Granted Patent: US 9,122,983 B2
• Time in Patent Office: Days
• US Class Current: 706/52
• CPC Class Codes: G06F 16/23 Updating
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: G08G 1/0112 from the vehicle, e.g. floating car data [FCD]
• CPC Class Codes: G08G 1/0129 for creating historical data or processing based on historical data
• CPC Class Codes: G08G 1/0141 for traffic information dissemination

Abstract

According to a method for creating a model for a travel time database, a leg network comprising leg sections between a starting point and a destination point are analyzed. Multiple routes are ascertained between the starting and destination points. Each leg section of a route is associated with a relative travel time loss and is weighted. The ascertained travel time losses and the associated weightings are used as input data for a learning method by way of which an existing knowledge base is iteratively expanded.

Claims

1. A method for creating a model for a travel time database, comprising:

providing a leg network comprising leg sections between a starting point and a destination point;

ascertaining a plurality of possible routes between the starting point and the destination point, wherein each route comprises leg sections including at least one of the leg sections of the leg network;

ascertaining a respective relative travel time loss for each leg section of each route with a respective traffic characteristic on each leg section of each route;

associating the ascertained relative travel time losses of the leg sections of the routes with each leg section of the leg network;

ascertaining a respective weighting of the relative travel time losses (u_i,k) for each leg section of each route with a respective traffic characteristic on each leg section of each route;

associating the ascertained weightings of the leg sections of the routes with each leg section of the leg network; and

ascertaining a probability distribution of the relative travel time losses for each leg section of the leg network (H) with the traffic characteristic associated with the respective leg section and a calibration parameter associated with the respective leg section.

2. The method according to claim 1,

wherein the respective relative travel time loss is ascertained as a function of the travel time measured between the starting point and the destination point and a minimum travel time predefined on the respective leg section of each route between the starting point and the destination point.

3. The method according to claim 2,

wherein the minimum travel time between the starting point and the destination point is ascertained from a sum of respective free travel times associated with the leg sections of the route,

wherein the free travel time associated with the respective leg section is ascertained as a function of the length of the respective leg section of the route and the free travel speed associated with the respective leg section of the route.

4. The method according to claim 3,

wherein the free travel speed associated with the respective leg section of the route is obtained from a digital map.

5. The method according to claim 1, further comprising:

ascertaining the plurality of possible routes and the relative travel time losses on the respective leg sections of the routes using the following acts (a) to (d);

(a) ascertaining a temporally shortest route comprising leg sections by evaluating a predefined travel time database;

(b) ascertaining the relative travel time losses of the respective leg sections of the temporally shortest route;

(c) ascertaining further routes comprising respective leg sections, wherein the respective minimum travel time of the further routes is shorter than a travel time measured on the respective further route; and

(d) ascertaining the relative travel time losses of the respective leg sections of the further routes.

6. The method according to claim 5,

wherein a resistance of the respective leg sections of the routes is increased after the step of ascertaining the relative travel time loss of the respective leg sections of the routes.

7. The method according to claim 1, further comprising:

ascertaining an expected travel time value for each of the ascertained routes;

ascertaining the respective weighting of the relative travel time losses for each leg section of each route as a function of the ascertained expected travel time values.

10. The method according to claim 1,

wherein the relative travel time losses of a first and a last leg of each of the routes can remain without consideration when associating the respective relative travel time losses of the respective leg sections of the routes with the leg sections of the sub-network.

11. The method according to claim 1, further comprising:

ascertaining a respective estimated mean value of the relative travel time losses as a function of a respective number of relative travel time losses and weightings that were associated with each leg section of the leg network with the respective traffic characteristic on the leg section.

12. The method according to claim 11,

wherein outliers of relative travel time losses of the leg sections of the routes remain without consideration in the ascertainment of the estimated mean value of the relative travel time losses.

13. The method according to claim 1,

wherein the probability distribution of the relative travel time losses for each leg section of the leg network is modeled as a gamma distribution.

8. A method comprising:

ascertaining a probability distribution model of relative travel time losses for each leg section of a leg network with a respective traffic characteristic on each leg section of the leg network by evaluating a predefined travel time database;

ascertaining a plurality of possible routes and ascertaining the respective relative travel time losses on each leg section of the routes by repeating the following acts (a) to (c) multiple times;

(a) ascertaining a random travel time on each leg section of the leg network by applying the probability distribution model and associating each of a randomly ascertained travel times with each leg section of the leg network;

(b) ascertaining a temporally shortest route between a starting point and a destination point based on the travel times associated with the leg sections of the leg network; and

(c) ascertaining the relative travel time losses of the respective leg sections of the temporally shortest route.

9. The method according to claim 8, comprising:

ascertaining a respective weighting of the relative travel time losses of the leg sections of the routes as a function of a value that is proportional to the reciprocal of the number of multiple repetitions for ascertaining the plurality of possible routes and the relative travel time losses of the leg sections of the routes.