Driving behavior prediction method and apparatus

• US 7,805,240 B2
• Filed: 11/20/2007
• Issued: 09/28/2010
• Est. Priority Date: 11/22/2006
• Status: Active Grant

• Current Assignee: DENSO Corporation
• Original Assignee: DENSO Corporation
• Inventors: Naitou, Takahiro, Itoh, Takafumi
• Primary Examiner(s): ARTHUR JEANGLAUDE, GERTRUDE
• Application Number:
  US11/984,576
• Publication Number:
  US 20080120025A1
• Time in Patent Office: 1,043 Days
• Field of Search: 701/29, 701/32, 701/33, 701/36, 701/207, 701/300, 340/425.5, 340/438, 340/439
• US Class Current: 701/207
• CPC Class Codes: G01C 21/12 executed aboard the object being navigated; Dead reckoning

Abstract

A driving behavior prediction apparatus includes, for accurately predicting a driving behavior, a position calculation unit for a subject vehicle position calculation, a route setting unit for setting a navigation route, a distance calculation unit for calculating a distance to a nearest object point, a parameter storage unit for storing a template weighting factor that reflects a driving operation tendency of a driver, a driving behavior prediction unit for predicting a driver'"'"'s behavior based on vehicle information and the template weighting factor, a driving behavior recognition unit for recognizing driver'"'"'s behavior at the object point, and a driving behavior learning unit for updating the template weighting factor so as to study the driving operation tendency of the driver in a case that a prediction result by the driving behavior prediction unit agrees with a recognition result by the driving behavior recognition unit.

Claims

1. A driving behavior prediction method comprising:

a first step of acquiring a current position of a subject vehicle;

a second step of acquiring a driving route to be followed by the subject vehicle;

a third step of calculating an object point distance which is a distance to an object point indicated by the route information based on position information acquired by the first step and route information acquired by the second step;

a fourth step of acquiring driving data including information on a speed of the subject vehicle and information on a vehicle operation performed by a driver to affect a running condition of the subject vehicle;

a fifth step of predicting the driving behavior of the driver at the object point based on a history of the driving data acquired by the fourth step, a plurality of templates, and a weighting factor for each of the plurality of templates when a distance determined by the third step is within a predetermined permissible distance;

a sixth step of recognizing the driving behavior actually practiced by the driver at the object point; and

a seventh step of updating the weighting factor for each of the plurality of templates, wherein, when the predicted driving behavior agrees with the recognized driving behavior in a comparison between a prediction result by the fifth step and a recognition result by the sixth step, the seventh step updates the weighting factor for each of relevant templates that is used for generating the prediction result in a manner that increases a reliability of a prediction of the driving behavior by using the plurality of templates.

2. A driving behavior prediction apparatus comprising:

a position information acquisition unit that acquires a current position of a subject vehicle;

a route information acquisition unit that acquires a driving route to be followed by the subject vehicle;

a distance calculation unit that calculates an object point distance which is a distance to an object point indicated by the route information based on position information acquired by the position information acquisition unit and route information acquired by the route information acquisition unit;

a driving data acquisition unit that acquires driving data including information on a speed of the subject vehicle and information on a vehicle operation performed by a driver to affect a running condition of the subject vehicle;

a template storage unit that stores a plurality of templates representing typical patterns of a driving behavior of the driver;

a weight storage unit that stores a weighting factor for each of the plurality of templates;

a driving behavior prediction unit that predicts the driving behavior of the driver at the object point based on a history of the driving data acquired by the driving data acquisition unit, the plurality of templates stored in the template storage unit, and the weighting factor for each of the plurality of templates stored in the weight storage unit when a distance determined by the distance calculation unit is within a predetermined permissible distance;

a driving behavior recognition unit recognizes the driving behavior actually practiced by the driver at the object point; and

a first weight update unit that updates the weighting factor for each of the plurality of templates, wherein, when the predicted driving behavior agrees with the recognized driving behavior in a comparison between a prediction result by the driving behavior prediction unit and a recognition result by the driving behavior recognition unit, the first weight update unit updates the weighting factor of the for each of relevant templates that is used for generating the prediction result in a manner that increases a reliability of a prediction of the driving behavior by using the plurality of templates.

3. The driving behavior prediction apparatus of claim 2,

wherein the first weight update unit updates the weighting factor for each of the relevant templates in a manner that increases the reliability in proportion to the object point distance at the time of predicting the driving behavior.

4. The driving behavior prediction apparatus of claim 3,

wherein the first weight update unit updates the weighting factor for each of the relevant templates by adding a reward for increasing the reliability of each of the relevant templates, and

the reward is configured to increase in geometric progression in proportion to the object point distance.

5. The driving behavior prediction apparatus of claim 2,

wherein the first weight update unit updates the weighting factor for each of the relevant templates every time the subject vehicle passes the object point.

6. The driving behavior prediction apparatus of claim 2 further comprising:

a second weight update unit that updates the weighting factor of a for each of the relevant templates, wherein, when the predicted driving behavior disagrees with the recognized driving behavior in a comparison between a prediction result by the driving behavior prediction unit and a recognition result by the driving behavior recognition unit, the second weight update unit updates the weighting factor for each of the relevant templates that is used for generating the prediction result in a manner that decreases a reliability of a prediction of the driving behavior by using the relevant templates.

7. The driving behavior prediction apparatus of claim 6,

wherein the driving behavior prediction unit determines a similarity between the plurality of templates and the history of the driving data,

the similarity is determined based on a product of a discrepancy value and the weighting factor for each of the relevant templates,

the discrepancy value is defined as a value that indicates discrepancy between the history of the driving data acquired by the driving data acquisition unit and the plurality of templates,

the driving behavior prediction unit predicts that the driving behavior of the driver at the object point becomes substantially same as the driving behavior indicated by the plurality of templates when a template having the similarity that is greater than a predetermined criterion exists, and

the second weight update unit updates the weighting factor for each of the relevant templates based on a difference between the similarity and the predetermined criterion.

8. The driving behavior prediction apparatus of claim 6,

wherein the second weight update unit updates the weighting factor every time the subject vehicle passes the object point.

9. The driving behavior prediction apparatus of claim 2 further comprising:

a driver recognition unit that recognizes a driver of the subject vehicle,

wherein the weight storage unit stores the weighting factor that corresponds to respective drivers recognized by the driver recognition unit, and

the driving behavior prediction unit predicts the driving behavior based on the weighting factor that corresponds to the driver recognized by the driver recognition unit.

10. The driving behavior prediction apparatus of claim 2,

wherein each of the plurality of templates is a time-series pattern that is generated by classifying driving patterns identified by collected data that includes at least one of pre-collected driving data and pseudo-driving data that simulates the driving data.

11. The driving behavior prediction apparatus of claim 10,

wherein each of the plurality of templates is generated by extracting the time-series pattern from the driving data in a manner that shifts the time-series pattern by a predetermined shift period.

12. The driving behavior prediction apparatus of claim 2,

wherein the object point at least includes a branch point in the driving route, and

the driving behavior prediction unit predicts as the driving behavior whether the subject vehicle turns at the branch point.

13. The driving behavior prediction apparatus of claim 2,

wherein the information on the vehicle operation includes operation information on at least one of an accelerator pedal and a brake pedal.

14. The driving behavior prediction apparatus of claim 2 further comprising:

a voice output unit that outputs a voice message of discrepancy that indicates discrepancy of a prediction result and a driver'"'"'s expected behavior when the prediction result by the driving behavior prediction unit disagree with the driver'"'"'s expected behavior at the object point.

15. The driving behavior prediction apparatus of claim 2 further comprising:

an image output unit that displays a message of discrepancy that indicates discrepancy of a prediction result and a driver'"'"'s expected behavior when the prediction result by the driving behavior prediction unit disagree with the driver'"'"'s expected behavior at the object point.