Calibrating multi-dimensional sensor for offset, sensitivity, and non-orthogonality

• US 8,645,093 B2
• Filed: 11/04/2009
• Issued: 02/04/2014
• Est. Priority Date: 11/04/2009
• Status: Expired due to Fees

• Current Assignee: Qualcomm, Inc.
• Original Assignee: Qualcomm, Inc.
• Inventors: Brunner, Christopher
• Primary Examiner(s): Kundu, Sujoy
• Assistant Examiner(s): BETSCH, REGIS J
• Application Number:
  US12/612,563
• Publication Number:
  US 20110106477A1
• Time in Patent Office: 1,553 Days
• Field of Search: 702/104, 702/92, 702/93, 702/196, 324/245, 324/246
• US Class Current: 702/104
• CPC Class Codes: G01R 33/0035 Calibration of single magnetic sensors, e.g. integrated calibration
• CPC Class Codes: G01R 33/0206 Three-component magnetometers
• CPC Class Codes: G06F 3/0346 with detection of the device orientation or free movement in a 3D space, e.g. 3D mice, 6-DOF [six degrees of freedom] pointers using gyroscopes, accelerometers or tilt-sensors
• CPC Class Codes: G06F 3/038 Control and interface arrangements therefor, e.g. drivers or device-embedded control circuitry

Abstract

A multi-dimensional sensor, a magnetometer or accelerometer, is calibrated based on the raw data provided by the sensor. Raw data is collected and may be used to generate ellipse or ellipsoid parameters, for a two-dimensional or three-dimensional sensor, respectively. An offset calibration factor is calculated based on the raw data, e.g., the determined ellipse or ellipsoid parameters. A sensitivity calibration factor is then calculated based on the offset calibration factor and the raw data. A non-orthogonality calibration factor can then be calculated based on the calculated offset and sensitivity calibration factors. Using the offset, sensitivity and non-orthogonality calibration factors, the raw data can be corrected to produce calibrated data.

Claims

1. A method comprising:

collecting raw data for a plurality of axes of a multi-dimensional sensor and providing the collected raw data to a processor;

iteratively determining at least one of ellipse parameters and ellipsoid parameters using the raw data and at least one of previously determined ellipse parameters and previously determined ellipsoid parameters, wherein iteratively determining at least one of ellipse parameters and ellipsoid parameters using the raw data uses a cyclic Jacobi sweep;

calculating with the processor an offset calibration factor for each of the plurality of axes of the multi-dimensional sensor based on the at least one of the ellipse parameters and the ellipsoid parameters;

calculating with the processor a sensitivity calibration factor for each of the plurality of axes of the multi-dimensional sensor based on the calculated offset calibration factor and the at least one of the ellipse parameters and the ellipsoid parameters;

calculating with the processor a non-orthogonality calibration factor for one or more pairs of axes of the multi-dimensional sensor based on the calculated offset calibration factor, the calculated sensitivity calibration factor and the at least one of the ellipse parameters and the ellipsoid parameters;

storing the calculated offset calibration factors, the sensitivity calibration factors and one or more non-orthogonality calibration factor in memory coupled to the processor; and

correcting raw data for the plurality of axes of the multi-dimensional sensor with the processor using the stored calculated offset calibration factors, the sensitivity calibration factors and the one or more non-orthogonality calibration factor.

2. The method of claim 1, wherein the multi-dimensional sensor is a magnetometer and the raw data is collected outside a presence of a disturbing magnetic near field.

3. The method of claim 1, wherein the multi-dimensional sensor is an accelerometer and the raw data is collected while the accelerometer is subjected to no acceleration beyond gravity.

4. The method of claim 1, wherein:

the offset calibration factor for each of the plurality of axes of the multi-dimensional sensor compensates for at least an offset along each axis of the multi-dimensional sensor;

the sensitivity calibration factor for each of the plurality of axes of the multi-dimensional sensor compensates for at least a sensitivity of each axis of the multi-dimensional sensor, the sensitivity of each axis being dissimilar; and

the non-orthogonality calibration factor for the one or more pairs of axes of the multi-dimensional sensor compensates for misalignment between the one or more pairs of axes of the multi-dimensional sensor.

5. The method of claim 1, wherein iteratively determining at least one of ellipse parameters and ellipsoid parameters using the raw data uses a cyclic Jacobi sweep, wherein the ellipse parameters are A, B, C, D, E, and L and are defined by:

AB_x²+BB_y²+2CB_xB_y+2DB_x+2EB_y+L=0

where B_x and B_y are the raw data measured along the plurality of axes of the multi-dimensional sensor, and wherein the ellipsoid parameters are A, B, C, D, E, F, G, H, K and L and are defined by;

AB_x²+BB_y²+CB_z²+2DB_xB_y+2EB_xB_z+2FB_yB_z+2GB_x+2HB_y+2KB_z+L=0

where B_x, B_y, and B_z are the raw data measured along the plurality of axes of the multi-dimensional sensor, and wherein the offset calibration factors, the sensitivity calibration factors and the one or more non-orthogonality calibration factor are calculated based on the at least one of ellipse parameters and ellipsoid parameters.

6. The method of claim 5, wherein ellipsoid parameters are determined, the sensitivity calibration factor for each of the plurality of axes of the multi-dimensional sensor is calculated as:

scale=√

{square root over (diag(pinv(Amat/c_dash)))}

where scale is the sensitivity calibration factor, diag denotes diagonal components of a matrix, pinv denotes pseudoinverse, and

7. The method of claim 6, wherein the non-orthogonality calibration factor for the one or more pairs of axes of the multi-dimensional sensor is calculated as:

Q=diag(1./scale)*pinv(Amat/C_dash)*diag(1./scale)

where Q is a matrix, scale is a vector with length 3, 1./scale indicates taking a reciprocal of each element, and diag( ) indicates turning the vector into a matrix of zeros except for the diagonal components, which correspond to the vector; and

ψ

=Q(1,2)/2;

θ

=Q(1,3)/2;

φ

=Q(2,3)/2;

where ψ

denotes non-orthogonality in radians in the x-y plane, θ

denotes non-orthogonality in radians in the x-z plane; and

φ

denotes non-orthogonality in radians in the y-z plane.

8. The method of claim 7, wherein the raw data for the plurality of axes of the multi-dimensional sensor are corrected as follows:

9. The method of claim 5, wherein ellipse parameters are determined and wherein the sensitivity calibration factor for each of the plurality of axes of the multi-dimensional sensor is calculated as:

scale=√

{square root over (diag(pinv(Amat/c_dash)))}

where scale is the sensitivity calibration factor, diag denotes diagonal components of a matrix, pinv denotes pseudoinverse, and

10. The method of claim 9, wherein the non-orthogonality calibration factor for the one or more pairs of axes of the multi-dimensional sensor is calculated as:

Q=diag(1./scale)*pinv(Amat/C_dash)*diag(1./scale)

where Q is a matrix, scale is a vector with length 3, 1./scale indicates taking a reciprocal of each element, and diag( ) indicates turning the vector into a matrix of zeros except for the diagonal components, which correspond to the vector; and

ψ

=Q(1,2)/2;

where ψ

denotes non-orthogonality in radians in the x-y plane.

11. The method of claim 10, wherein the raw data for the plurality of axes of the multi-dimensional sensor are corrected as follows:

12. A mobile station comprising:

a multi-dimensional sensor that provides raw data for each of a plurality of axes;

a processor connected to the multi-dimensional sensor, the processor receives from the multi-dimensional sensor the raw data for each of the plurality of axes;

memory connected to the processor; and

software held in the memory and run in the processor to calibrate the multi-dimensional sensor, the software including instructions to iteratively determine at least one of ellipse parameters and ellipsoid parameters using the raw data and at least one of previously determined ellipse parameters and previously determined ellipsoid parameters wherein the software includes instructions to iteratively determine at least one of ellipse parameters and ellipsoid parameters using the raw data using a cyclic Jacobi sweep, to calculate and store in the memory an offset calibration factor for each of the plurality of axes based on the at least one of the ellipse parameters and the ellipsoid parameters;

to calculate and store in the memory a sensitivity calibration factor for each of the plurality of axes based on the calculated offset calibration factor and the at least one of the ellipse parameters and the ellipsoid parameters; and

to calculate and store in the memory a non-orthogonality calibration factor for one or more pairs of axes based on the calculated offset calibration factor, the calculated sensitivity calibration factor and the at least one of the ellipse parameters and the ellipsoid parameters;

the software further including instructions to correct the raw data for each of the plurality of axes using the stored calculated offset calibration factors, the stored sensitivity calibration factors and the stored one or more non-orthogonality calibration factor.

13. The mobile station of claim 12, wherein the multi-dimensional sensor is at least one of a magnetometer and an accelerometer.

14. The mobile station of claim 12, wherein:

the offset calibration factor for each of the plurality of axes of the multi-dimensional sensor compensates for at least an offset along each axis of the multi-dimensional sensor;

the sensitivity calibration factor for each of the plurality of axes of the multi-dimensional sensor compensates for at least a sensitivity of each axis of the multi-dimensional sensor, the sensitivity of each axis being dissimilar; and

the non-orthogonality calibration factor for the one or more pairs of axes of the multi-dimensional sensor compensates for misalignment between the one or more pairs of axes of the multi-dimensional sensor.

15. The mobile station of claim 12, wherein the instructions to determine at least one of ellipse parameters and ellipsoid parameters using the raw data uses a cyclic Jacobi sweep, wherein the ellipse parameters are A, B, C, D, E, and L and are defined by:

AB_x²+BB_y²+2CB_xB_y+2DB_x+2EB_y+L=0

where B_x and B_y are the raw data measured along the plurality of axes of the multi-dimensional sensor, and wherein the ellipsoid parameters are A, B, C, D, E, F, G, H, K and L and are defined by;

AB_x²+BB_y²+CB_z²+2DB_xB_y+2EB_xB_z+2FB_yB_z+2GB_x+2HB_y+2KB_z+L=0

where B_x, B_y, and B_z are the raw data measured along the plurality of axes of the multi-dimensional sensor, and wherein the offset calibration factors, the sensitivity calibration factors and the one or more non-orthogonality calibration factor are calculated based on the at least one of ellipse parameters and ellipsoid parameters.

16. The mobile station of claim 15, wherein the software includes instructions to calculate the sensitivity calibration factor for each of the plurality of axes of the multi-dimensional sensor as:

scale=√

{square root over (diag(pinv(Amat/c_dash)))}

where scale is the sensitivity calibration factor, diag denotes diagonal components of a matrix, pinv denotes pseudoinverse, and

17. The mobile station of claim 16, wherein software includes instructions to calculate the non-orthogonality calibration factor for the one or more pairs of axes of the multi-dimensional sensor as:

Q=diag(1./scale)*pinv(Amat/C_dash)*diag(1./scale)

where Q is a matrix, scale is a vector with length 3, 1./scale indicates taking a reciprocal of each element, and diag( ) indicates turning the vector into a matrix of zeros except for the diagonal components, which correspond to the vector; and

ψ

=Q(1,2)/2;

θ

=Q(1,3)/2;

φ

=Q(2,3)/2;

where ψ

denotes non-orthogonality in radians in the x-y plane, θ

denotes non-orthogonality in radians in the x-z plane; and

φ

denotes non-orthogonality in radians in the y-z plane.

18. The mobile station of claim 17, wherein the software includes instructions to correct the raw data for each of the plurality of axes using the stored calculated offset calibration factors, the stored sensitivity calibration factors and the stored one or more non-orthogonality calibration factor as follows:

19. The mobile station of claim 15, wherein the software includes instructions to calculate the sensitivity calibration factor for each of the plurality of axes of the multi-dimensional sensor as:

scale=√

{square root over (diag(pinv(Amat/c_dash)))}

where scale is the sensitivity calibration factor, diag denotes diagonal components of a matrix, pinv denotes pseudoinverse, and

20. The mobile station of claim 19, wherein software includes instructions to calculate the non-orthogonality calibration factor for the one or more pairs of axes of the multi-dimensional sensor as:

Q=diag(1./scale)*pinv(Amat/C_dash)*diag(1./scale)

where Q is a matrix, scale is a vector with length 3, 1./scale indicates taking a reciprocal of each element, and diag( ) indicates turning the vector into a matrix of zeros except for the diagonal components, which correspond to the vector; and

ψ

=Q(1,2)/2;

where ψ

denotes non-orthogonality in radians in the x-y plane.

21. The mobile station of claim 20, wherein the software includes instructions to correct the raw data for each of the plurality of axes using the stored calculated offset calibration factors, the stored sensitivity calibration factors and the stored one or more non-orthogonality calibration factor as follows:

22. A mobile station comprising:

means for collecting raw data for a plurality of axes of a multi-dimensional sensor;

means for iteratively determining at least one of ellipse parameters and ellipsoid parameters using the raw data and at least one of previously determined ellipse parameters and previously determined ellipsoid parameters, wherein iteratively determining at least one of ellipse parameters and ellipsoid parameters using the raw data uses a cyclic Jacobi sweep;

means for calculating an offset calibration factor for each of the plurality of axes of the multi-dimensional sensor based on the at least one of the ellipse parameters and the ellipsoid parameters;

means for calculating a sensitivity calibration factor for each of the plurality of axes of the multi-dimensional sensor based on the calculated offset calibration factor and the at least one of the ellipse parameters and the ellipsoid parameters;

means for calculating a non-orthogonality calibration factor for one or more pairs of axes of the multi-dimensional sensor based on the calculated offset calibration factor, the calculated sensitivity calibration factor and the at least one of the ellipse parameters and the ellipsoid parameters; and

means for correcting raw data for the plurality of axes of the multi-dimensional sensor using the calculated offset calibration factors, the sensitivity calibration factors and the one or more non-orthogonality calibration factor.

23. The mobile station of claim 22, wherein the multi-dimensional sensor is at least one of a magnetometer and an accelerometer.

24. The mobile station of claim 22, wherein:

the offset calibration factor for each of the plurality of axes of the multi-dimensional sensor compensates for at least an offset along each axis of the multi-dimensional sensor;

the sensitivity calibration factor for each of the plurality of axes of the multi-dimensional sensor compensates for at least a sensitivity of each axis of the multi-dimensional sensor, the sensitivity of each axis being dissimilar; and

the non-orthogonality calibration factor for the one or more pairs of axes of the multi-dimensional sensor compensates for misalignment between the one or more pairs of axes of the multi-dimensional sensor.

25. The mobile station of claim 22, wherein the means for iteratively determining at least one of ellipse parameters and ellipsoid parameters using the raw data uses a cyclic Jacobi sweep, wherein the ellipse parameters are A, B, C, D, E, and L and are defined by:

AB_x²+BB_y²+2CB_xB_y+2DB_x+2EB_y+L=0

where B_x and B_y are the raw data measured along the plurality of axes of the multi-dimensional sensor, and wherein the ellipsoid parameters are A, B, C, D, E, F, G, H, K and L and are defined by;

AB_x²+BB_y²+CB_z²+2DB_xB_y+2EB_xB_x+2FB_yB_z+2GB_x+2HB_y+2KB_z+L=0

where B_x, B_y, and B_z are the raw data measured along the plurality of axes of the multi-dimensional sensor, and wherein the offset calibration factors, the sensitivity calibration factors and the one or more non-orthogonality calibration factor are calculated based on the at least one of ellipse parameters and ellipsoid parameters.

26. The mobile station of claim 25, wherein the means for calculating the sensitivity calibration factor calculates the sensitivity calibration factor as:

scale=√

{square root over (diag(pinv(Amat/c_dash)))}

where scale is the sensitivity calibration factor, diag denotes diagonal components of a matrix, pinv denotes pseudoinverse, and

27. The mobile station of claim 26, wherein the means for calculating the non-orthogonality calibration factor calculates the non-orthogonality calibration factor as:

Q=diag(1./scale)*pinv(Amat/C_dash)*diag(1./scale)

where Q is a matrix, scale is a vector with length 3, 1./scale indicates taking a reciprocal of each element, and diag( ) indicates turning the vector into a matrix of zeros except for the diagonal components, which correspond to the vector; and

ψ

=Q(1,2)/2;

θ

=Q(1,3)/2;

φ

=Q(2,3)/2;

where ψ

denotes non-orthogonality in radians in the x-y plane, θ

denotes non-orthogonality in radians in the x-z plane; and

φ

denotes non-orthogonality in radians in the y-z plane.

28. The mobile station of claim 26, further comprising a means for correcting the raw data that corrects the raw data as:

29. The mobile station of claim 25, wherein the means for calculating the sensitivity calibration factor calculates the sensitivity calibration factor as:

scale=√

{square root over (diag(pinv(Amat/c_dash)))}

where scale is the sensitivity calibration factor, diag denotes diagonal components of a matrix, pinv denotes pseudoinverse, and

30. The mobile station of claim 29, wherein the means for calculating the non-orthogonality calibration factor calculates the non-orthogonality calibration factor as:

Q=diag(1./scale)*pinv(Amat/C_dash)*diag(1./scale)

where Q is a matrix, scale is a vector with length 3, 1./scale indicates taking a reciprocal of each element, and diag( ) indicates turning the vector into a matrix of zeros except for the diagonal components, which correspond to the vector; and

ψ

=Q(1,2)/2;

where ψ

denotes non-orthogonality in radians in the x-y plane.

31. The mobile station of claim 30, further comprising a means for correcting the raw data that corrects the raw data as:

32. A non-transitory computer-readable medium including program code stored thereon, comprising:

program code to iteratively determine at least one of ellipse parameters and ellipsoid parameters using collected raw data and at least one of previously determined ellipse parameters and previously determined ellipsoid parameters, wherein the program code to iteratively determine at least one of ellipse parameters and ellipsoid parameters using the collected raw data uses a cyclic Jacobi sweep;

program code to calculate an offset calibration factor for each of a plurality of axes of a multi-dimensional sensor based on the at least one of the ellipse parameters and the ellipsoid parameters for the plurality of axes of the multi-dimensional sensor;

program code to calculate a sensitivity calibration factor for each of the plurality of axes of the multi-dimensional sensor based on the calculated offset calibration factor and the at least one of the ellipse parameters and the ellipsoid parameters for the plurality of axes of the multi-dimensional sensor;

program code to calculate a non-orthogonality calibration factor for one or more pairs of axes of the multi-dimensional sensor based on the calculated offset calibration factor, the calculated sensitivity calibration factor and the at least one of the ellipse parameters and the ellipsoid parameters for the plurality of axes of the multi-dimensional sensor; and

program code to correct raw data for the plurality of axes of the multi-dimensional sensor using the calculated offset calibration factors, the sensitivity calibration factors and the one or more non-orthogonality calibration factor.

33. The non-transitory computer-readable medium of claim 32, wherein the program code to iteratively determine at least one of ellipse parameters and ellipsoid parameters using the raw data uses a cyclic Jacobi sweep, wherein the ellipse parameters are A, B, C, D, E, and L and are defined by:

AB_x²+BB_y²+2CB_xB_y+2DB_x+2EB_y+L=0

where B_x and B_y are the raw data measured along the plurality of axes of the multi-dimensional sensor, and wherein the ellipsoid parameters are A, B, C, D, E, F, G, H, K and L and are defined by;

AB_x²+BB_y²+CB_z²+2DB_xB_y+2EB_xB_x+2FB_yB_z+2GB_x+2HB_y+2KB_z+L=0

where B_x, B_y, and B_z are the raw data measured along the plurality of axes of the multi-dimensional sensor, and wherein the offset calibration factors, the sensitivity calibration factors and the one or more non-orthogonality calibration factor are calculated based on the at least one of ellipse parameters and ellipsoid parameters.

34. The non-transitory computer-readable medium of claim 33, wherein the program code to calculate the sensitivity calibration factor calculates the sensitivity calibration factor as:

scale=√

{square root over (diag(pinv(Amat/c_dash)))}

where scale is the sensitivity calibration factor, diag denotes diagonal components of a matrix, pinv denotes pseudoinverse, and

35. The non-transitory computer-readable medium of claim 34, wherein the program code to calculate the non-orthogonality calibration factor calculates the non-orthogonality calibration factor as:

Q=diag(1./scale)*pinv(Amat/C_dash)*diag(1./scale)

where Q is a matrix, scale is a vector with length 3, 1./scale indicates taking a reciprocal of each element, and diag( ) indicates turning the vector into a matrix of zeros except for the diagonal components, which correspond to the vector; and

ψ

=Q(1,2)/2;

θ

=Q(1,3)/2;

φ

=Q(2,3)/2;

where ψ

denotes non-orthogonality in radians in the x-y plane, θ

denotes non-orthogonality in radians in the x-z plane; and

φ

denotes non-orthogonality in radians in the y-z plane.

36. The non-transitory computer-readable medium of claim 35, wherein the program code to correct the raw data corrects the raw data as:

37. The non-transitory computer-readable medium of claim 33, wherein the program code to calculate the sensitivity calibration factor calculates the sensitivity calibration factor as:

scale=√

{square root over (diag(pinv(Amat/c_dash)))}

where scale is the sensitivity calibration factor, diag denotes diagonal components of a matrix, pinv denotes pseudoinverse, and

38. The non-transitory computer-readable medium of claim 37, wherein the program code to calculate the non-orthogonality calibration factor calculates the non-orthogonality calibration factor as:

Q=diag(1./scale)*pinv(Amat/C_dash)*diag(1./scale)

where Q is a matrix, scale is a vector with length 3, 1./scale indicates taking a reciprocal of each element, and diag( ) indicates turning the vector into a matrix of zeros except for the diagonal components, which correspond to the vector; and

ψ

=Q(1,2)/2;

where ψ

denotes non-orthogonality in radians in the x-y plane.

39. The non-transitory computer-readable medium of claim 38, wherein the program code to correct the raw data corrects the raw data as: