Location detecting method

• US 4,999,783 A
• Filed: 12/22/1988
• Issued: 03/12/1991
• Est. Priority Date: 05/11/1987
• Status: Expired due to Term

• Current Assignee: Sumitomo Electric Industries Limited (Sumitomo Corporation)
• Original Assignee: Sumitomo Electric Industries Limited (Sumitomo Corporation)
• Inventors: Shimizu, Toshiyuki, Shimizu, Osamu, Tenmoku, Kenji, Mito, Kunihiko
• Primary Examiner(s): Lall, Parshotam S.
• Assistant Examiner(s): Zanelli, Michael
• Application Number:
  US07/295,206
• Time in Patent Office: 810 Days
• Field of Search: 364/424.01, 364/424.02, 364/449, 364/450, 364/459, 364/460, 364/447, 340/995, 340/988, 73/178 R
• US Class Current: 701/217
• CPC Class Codes: G01C 21/30 Map- or contour-matching

Abstract

A location detecting method for detecting the location of a vehicle within a predetermined area, comprising the steps of: outputting distance data at every predetermined interval, the distance data being representative of the distance that a vehicle has travelled; outputting heading angle data at every predetermined interval, the heading angle data being representative of the heading angle change of vehicle travel; storing road map data representative of roads; computing a present location data from the distance data and the heading angle data; computing a first limit error (E1) and a second limit error (E2) of the present location data which are determined on the basis of errors in the distance data, in the heading angle data and in the road map data; registering locations on the roads, which roads are located within the first limit error (E1) of the present location data, as estimate locations; computing correlation coefficients respectively corresponding to the registered estimate locations; selecting a correlation coefficient whose error is smallest with respect to the road from the computed correlation coefficients; and outputting the registered estimate location corresponding to the selected correlation coefficient as a present location.

Claims

1. A location detecting method for detecting the location of a vehicle within a predetermined area, comprising the steps of:

outputting distance data at a predetermined interval, the distance data being representative of the distance that said vehicle has traveled;

outputting heading angle data at a predetermined interval, the heading angle data being representative of the heading angle change of the vehicle travel;

storing road map data representative of roads;

computing a present location data from said distance data and said heading angle data;

computing a first limit error (E1) and a second limit error (E2) of said present location data which are determined on the basis of errors in said distance data, in said heading angle data and in said road map data;

registering locations of all roads located within said first limit error (E1), as estimate locations;

computing a correlation coefficient for each of said registered estimate locations to each of said roads located within said first limit error, by accumulating a function having as a parameter the errors between the estimate locations to the registered roads and a location on said road map at a particular point;

selecting one of said correlation coefficients indicating the error with respect to one of said roads is smallest of the errors indicated by said correlation coefficients;

outputting as a present location one of said estimate locations corresponding to the selected one correlation coefficient; and

if any of said roads location within said first limit error (E1) do not exist, outputting as a present location said present location data computed from said distance data and said heading angle data and also registering locations on all said roads located within said second limit error (E2) as second estimate locations, and if a second one of said correlation coefficients indicating that error with respect to each said road of the registered second estimate locations is less than a predetermined value and is smallest of the errors indicated by said correlation coefficients, is obtained, then outputting as a present location one of said second estimate locations corresponding to said second correlation coefficient.

2. A location detecting method as set forth in claim 1, in which said selected one correlation coefficient is further selected by selecting one of said correlation coefficients which is largest among said computed correlation coefficients and in which one of said estimate locations corresponding to said correlation coefficient which is largest amount said computed correlation coefficients is outputted as a present location.

3. A location detecting method as set forth in claim 2, in which when one of said computed correlation coefficients is less than a predetermined value, registration of the corresponding estimate location is erased.

4. A location detecting method as set forth in claim 2, in which when a difference between two of said computed correlation coefficients is more than a predetermined value, registration of the estimate location corresponding to the smaller of the two correlation coefficients is erased.

5. A location detecting method as set forth in claim 3, or 4, in which said registration of the corresponding estimate location is erased after the vehicle passes a predetermined location on the road.

6. A location detecting method as set forth in claim 2, in which said correlation coefficient which is largest is selected by providing a predetermined hysteresis characteristic, in the case that the estimate location corresponding to the largest correlation coefficient and the estimate location corresponding to the second largest correlation coefficient are inverted.

7. A location detecting method as set forth in claim 2, in which, in the case that there are a plurality of estimation locations corresponding to said correlation coefficient which is largest, among said computed correlation coefficients the estimate location which is nearest to a center of the road is outputted as a present location.

8. A location detecting method as set forth in claim 2, in which said distance data representative of the distance that said vehicle has travelled is outputted at a predetermined time and in which said heading angle data representative of the heading angle change of vehicle travel is outputted at the predetermined time.

9. A location detecting method as set forth in claim 2, in which said distance data representative of the distance that said vehicle has travelled is outputted at a predetermined distance and in which said heading angle data representative of the heading angle change of vehicle travel is outputted at the predetermined distance.

10. A location detecting method as set forth in claim 2, in which said second correlation coefficient is obtained by evaluating correlation coefficients corresponding to all said roads which are located within said second limit error.

11. A location detecting method as set forth in claim 2, in which said second correlation coefficient is obtained by classifying said roads located within said second limit error and evaluating correlation coefficients corresponding to the classified roads.

12. A location detecting method as set forth in claim 11, in which, in the case that said second limit error consists of a plurality of limit errors which are different in size, while the vehicle travels a predetermined distance, correlation coefficients corresponding to subject roads are evaluated in order of narrower limit error until an estimate location corresponding to a correlation coefficient whose error is less than a predetermined value and smallest. PG,48

13. A location detecting method as set forth in claim 12, in which said second correlation coefficient is obtained by evaluating correlation coefficients corresponding to arterial roads.

14. A location detecting method as set forth in claim 12, in which said second correlation coefficient is obtained by considering characteristic of vehicle travel and evaluating correlation coefficients corresponding to roads that the vehicle can travel with high probability.

15. A location detecting method as set forth in claim 1, in which said estimate locations are corrected by the amount of the error between the estimate location to each said road and a location on said road map at a last particular point.

16. A location detecting method as set forth in claim 15, in which in the case that a road of a correlation coefficient to be selected within said first limit error is the only one and that the correlation coefficient is maintained larger than a predetermined value while the vehicle travels a predetermined place of the road, an estimate location corresponding to the one road is corrected by the amount of the error between the estimate location to the corresponding road and a location on said road map at the last particular point.