Method for calibrating inertial sensor installed in arbitrary position on board vehicle and sensor system of dynamics of vehicle able to be installed on board in arbitrary position

• US 8,825,274 B2
• Filed: 12/23/2011
• Issued: 09/02/2014
• Est. Priority Date: 12/24/2010
• Status: Active Grant

• Current Assignee: Magneti Marelli S.p.A.
• Original Assignee: Magneti Marelli S.p.A.
• Inventors: De Tommasi, Claudio
• Primary Examiner(s): Jabr, Fadey
• Assistant Examiner(s): Soofi, Yazan A
• Application Number:
  US13/336,034
• Publication Number:
  US 20120330497A1
• Time in Patent Office: 984 Days
• Field of Search: 701/33.1
• US Class Current: 701/31.1
• CPC Class Codes: G01C 25/005 initial alignment, calibration or starting-up of inertial devices

Abstract

A method calibrates an inertial-sensor device installed in an arbitrary position on board a vehicle and adapted to detect at least one entity indicative of vehicle dynamics along at least one direction of a local reference-coordinate system. The method includes steps of: acquiring by the inertial-sensor device at predetermined measuring times real-measured data indicative of the vehicle dynamics in the local reference-coordinate system; acquiring by a calibrating detector different from the inertial-sensor device, at sampling times coinciding with the measuring times, reference-measured data indicative of the vehicle dynamics in a vehicle-coordinate system; and generating a coordinate-transformation matrix adapted to correlate the entity measured in the local reference-coordinate system with a corresponding entity in the vehicle-coordinate system. Also, a sensor system of dynamics of the vehicle is able to be installed on board in an arbitrary position.

Claims

1. A method for calibrating an inertial-sensor device installed in an arbitrary position on board a vehicle and adapted to detect at least one entity indicative of dynamics of the vehicle along at least one direction of a local reference-coordinate system, said method comprising steps of:

acquiring by the inertial-sensor device at predetermined measuring times real-measured data indicative of the vehicle dynamics in the local reference-coordinate system;

acquiring by a calibrating detector different from the inertial-sensor device, at sampling times coinciding with the measuring times, reference-measured data indicative of the vehicle dynamics in a vehicle-coordinate system; and

generating by a processing unit a coordinate-transformation matrix adapted to correlate the entity measured in the local reference-coordinate system with a corresponding entity in the vehicle-coordinate system, wherein all elements of the coordinate-transformation matrix are computed by comparing the real-measured data acquired by the initial-sensor device with corresponding the reference-measured data obtained in a plurality of different driving conditions, the real-measured data include a quantity of items of data substantially equal to a quantity of the elements of the coordinate-transformation matrix, and a value of each of the elements of the coordinate-transformation matrix is adjusted by forcing a constraint of orthogonality to obtain an orthogonal-transformation matrix that is closest to the coordinate-transformation matrix obtained based on the measured data, wherein Ro=R(R^TR)<sup>−

1/2</sup> is applied to determine the orthogonal-transformation matrix that is closest to the coordinate-transformation matrix obtained based on the measured data.

2. A calibration method as set forth in claim 1, wherein the calibrating detector includes a satellite-positioning system.

3. A calibration method as set forth in claim 1, wherein the calibrating detector includes odometric sensors integrated on board the vehicle.

4. A calibration method as set forth in claim 1, wherein said method is carried out if a travelling speed of the vehicle is greater than a minimum threshold speed of the vehicle.

5. A calibration method as set forth in claim 1, wherein said method comprises further periodically updating the value of each of the elements of the coordinate-transformation matrix and computing the orthogonal-transformation matrix after a predetermined number of the updates.

6. A calibration method as set forth in claim 1, wherein computing the value of the elements of the coordinate-transformation matrix at each iteration of the calibration includes applying a filtering operation, taking into account a number of previous measuring events, according to:

Output=(Output*nstep÷

New_Value)/(nstep+1).

7. A calibration method as set forth in claim 1, wherein said method is applied for calibrating a tri-axial accelerometer sensor.

8. A calibration method as set forth in claim 7, wherein the measured data indicative of the vehicle dynamics are substantially linear and centripetal-acceleration data detected in at least one of following dynamic conditions:

the vehicle either of being substantially at rest and moving at a substantially constant and uniform speed;

the vehicle either of substantially linearly accelerating and decelerating; and

the vehicle changing its travel direction at a substantially constant speed.

9. A calibration method as set forth in claim 8, wherein the dynamic condition of the vehicle being substantially at rest is determined when a detected travelling speed of the vehicle is lower than a first speed threshold of the vehicle and an absolute value of a detected substantially linear acceleration of the vehicle is lower than a first linear-acceleration threshold of the vehicle.

10. A calibration method as set forth in claim 8, wherein the dynamic condition of the vehicle substantially linearly accelerating is determined when a detected travelling speed of the vehicle is greater than a second speed threshold of the vehicle, an absolute value of a detected substantially linear acceleration of the vehicle is greater than a second linear-acceleration threshold of the vehicle, and an absolute ratio between a detected centripetal acceleration and the detected substantially linear acceleration of the vehicle is lower than a comparison-threshold value.

11. A calibration method as set forth in claim 8, wherein the dynamic condition of centripetal acceleration is determined when a detected travelling speed of the vehicle is greater than a third speed threshold of the vehicle, an absolute value of a detected centripetal acceleration of the vehicle is greater than a first centripetal-acceleration threshold of the vehicle, and an absolute ratio between a detected linear acceleration and the detected centripetal acceleration of the vehicle is lower than a comparison-threshold value.

12. A calibration method as set forth in claim 1, wherein calibration-acceleration data detected at a plurality of subsequent sampling times and acceleration data measured at a plurality of subsequent measuring times are processed by applying a “

Butterworth”

filter of second order implementing following relationship;

OutData(i)=Input(i)*a(1)+Input(i−

1)*a(2)+Input(i−

2)*a(3)−

OutData(i−

1)*b(2)+−

OutData(i−

2)*b(3), where “

Input” and

“

Outdata”

are dynamic input and output entities, respectively.

13. A calibration method as set forth in claim 1, wherein said method comprises further steps of determining contribution of gravity acceleration in a condition of the vehicle being at rest from a predetermined minimum number of real-measured-data samples and computing of the elements of the coordinate-transformation matrix as a function of the measured data obtained in driving conditions of linear and centripetal acceleration of the vehicle being carried out by taking into consideration the determined contribution of the gravity acceleration.

14. A sensor system of dynamics of a vehicle capable of being installed at an arbitrary position on board the vehicle, said sensor system comprising:

an inertial-sensor device adapted to detect at least one entity indicative of the vehicle dynamics along at least one direction of a local reference-coordinate system including a processor arranged for acquiring data detected by said inertial-sensor device in said local reference-coordinate system and converting said data in a vehicle reference system through a coordinate-transformation matrix computed by a calibration method, wherein Ro=R(R^TR)<sup>−

1/2</sup> is applied to determine the orthogonal-transformation matrix that is closest to the coordinate-transformation matrix obtained based on the measured data.