Method and apparatus of improving accuracy of accelerometer

• US 8,117,888 B2
• Filed: 08/28/2009
• Issued: 02/21/2012
• Est. Priority Date: 02/11/2009
• Status: Active Grant

• Current Assignee: Well Being Digital Limited
• Original Assignee: Perception Digital Limited (Atom Xquare Corp. Ltd.)
• Inventors: Chan, Kai Kin, Ma, Chor Tin
• Primary Examiner(s): WILLIAMS, HEZRON
• Assistant Examiner(s): Frank, Rodney T
• Application Number:
  US12/549,720
• Publication Number:
  US 20100199744A1
• Time in Patent Office: 907 Days
• Field of Search: 73/13.8, 73/10.1
• US Class Current: 73/1.38
• CPC Class Codes: G01P 15/18 in two or more dimensions
• CPC Class Codes: G01P 21/00 Testing or calibrating of apparatus or devices covered by the preceding groups

Abstract

A method and apparatus for calibrating or adjusting an accelerometer, wherein the accelerometer is held stationary to obtain the signal outputs from the accelerometer, representing component vectors making up the composite vector of 1 g. Thus, gain or sensitivity, and the zero-g signal offset along for each axis of the accelerometer is determined and adjustable.

Claims

1. A method of calibrating the accuracy of a 3-axis accelerometer comprising the steps of:

a) holding the 3-axis accelerometer in an orientation and relatively still to the earth;

b) obtaining 3-axis accelerometer outputs, V_x, V_y, V_z, for each respective axis x, y, z;

c) equating outputs V_x, V_y, V_z to gravitational force of 1 g according to the equation,

1=G_x²(V_x−

V_xo)²+G_y²(V_y−

V_yo)²+G_z²(V_z−

V_zo)²;

where

G_x=gain of the 3-axis accelerometer along the x axis;

G_y=gain of the 3-axis accelerometer along the y axis;

G_z=gain of the 3-axis accelerometer along the z axis;

V_x0=zero-g output of the x axis;

V_y0=zero-g output of the y axis;

V_z0=zero-g output of the z axis;

d) solving the values of G_x, G_y, G_z, V_x0, V_y0 and V_z0;

whereby

subsequent 3-axis accelerometer output is treated by the values of G_x, G_y, G_z, V_x0, V_y0 and V_z0 to obtain the component force vectors, a_x, a_y, a_z, along the axes of x, y and z, according to the relationship;

a_x=G_x(V_x−

V_x0);

a_y=G_y(V_y−

V_y0); and

a_z=G_z(V_z−

V_z0).

2. A method of calibrating the accuracy of a 3-axis accelerometer as claimed in claim 1, further comprising the steps of:

e) repeating steps a) to c) to obtain and store outputs V_xm, V_ym, V_zm (m=1 to ≧

6) from the 3-axis accelerometer in the m^th orientation to provide a full rank matrix, in order to obtaining the values of G_x, G_y, G_z, V_x0, V_y0 and V_z0 from the full rank matrix.

3. A method of calibrating the accuracy of a 3-axis accelerometer as claimed in claim 1,

wherein solving out the values of G_x, G_y, G_z, V_x0, V_y0 and V_z0 comprises using the least-squares method.

4. A method of calibrating the accuracy of a 3-axis accelerometer as claimed in claim 2, wherein

if m≧

6 at step e) and if the number of stored sets of V_xm, V_ym, V_zm are enough to provide a full rank matrix, the method of claim 2 is used; and

if m<

6 at step e), the method of claim 3 is used.

5. A method of calibrating the accuracy of a 3-axis accelerometer as claimed in claim 4, further comprising the steps of:

detecting that the accelerometer is rotated into a new orientation before collecting the outputs, V_xm, V_ym, V_zm in step (e).

6. A method of calibrating the accuracy of a 3-axis accelerometer as claimed in claim 1, wherein

determining that the 3-axis accelerometer is held still in the orientation comprises the steps of:

calculating the composite acceleration vector represented by V_x, V_y, V_z;

wherein

the 3-axis accelerometer is deemed to be still if the composite acceleration vector is within a pre-determined threshold value.

7. A method of calibrating the accuracy of a 3-axis accelerometer as claimed in claim 1, wherein determining that the 3-axis accelerometer is held still in the orientation comprises the steps of:

collecting a plurality of accelerometer outputs for each axis within a pre-determined period of time,

calculating the mean and standard deviation between the plurality of accelerometer outputs; and

if the standard deviation between the plurality of accelerometer outputs is within a pre-determined threshold, the accelerometer output is deemed to have been collected with the accelerometer being sufficiently still, whereby

the mean value of the outputs for each axis is accepted as accelerometer output, V_x, V_y, V_z in step (b).

8. A method of calibrating the accuracy of a 3-axis accelerometer as claimed in claim 1, wherein the method is triggered by a timer.

9. An accuracy improvement unit for a 3-axis accelerometer comprising

an x input for receiving a V_x output from the 3-axis accelerometer representing the force acting on an x axis;

a y input for receiving a V_y output from the 3-axis accelerometer representing the force acting on a y axis;

a z input for receiving a V_z output from the 3-axis accelerometer representing the force acting on a z axis;

the accuracy improvement unit being capable of treating outputs V_x, V_y, V_z as indicative of component vectors of an acceleration force of 1 g according to the relationship;

1=G_x²(V_x−

V_xo)²+G_y²(V_y−

V_yo)²+G_z²(V_z−

V_zo)²

G_x=sensitivity of the accelerometer along the x axis;

G_y=sensitivity of the accelerometer along the y axis;

G_z=sensitivity of the accelerometer along the z axis;

V_x0=zero-g output the x axis;

V_y0=zero-g output the y axis; and

V_z0=zero-g output the z axis;

the processing unit is capable of determining G_x, G_y, G_z, V_x0, V_y0 and V_z0 from said relationship;

wherein

the accuracy improvement unit being further capable of applying G_x, G_y, G_z, V_x0, V_y0 and V_z0 to subsequent outputs V_x, V_y, V_z from the accelerometer to provide component force vectors a_x, a_y, a_z according to the relationship;

a_x=G_x(V_x−

V_x0)

a_y=G_y(V_y−

V_y0)

a_z=G_z(V_z−

V_z0).

10. An accuracy improvement unit for a 3-axis accelerometer as claimed in claim 9, wherein

the accuracy improvement unit is capable of storing m sets of outputs V_zm, V_ym, V_zm;

the accuracy improvement unit is capable of providing a full rank matrix from m (m=1 to ≧

6) sets of outputs V_zm, V_ym, V_zm to determine G_x, G_y, G_z, V_x0, V_y0 and V_z0;

wherein

the m sets of outputs are obtained when the 3-axis accelerometer is held relatively still to the earth, in respective m orientations.

11. An accuracy improvement unit for a 3-axis accelerometer as claimed in claim 9, wherein

solving out the values of G_x, G_y, G_z, V_x0, V_y0 and V_z0 comprises using the least squares method.

12. An accuracy improvement unit for a 3-axis accelerometer as claimed in claim 9, wherein

the values of G_x, G_y, G_z, V_x0, V_y0 and V_z0 are solved by the method of claim 11 if the stored m sets of V_zm, V_ym, V_zm are not sufficient to provide a full rank matrix;

and

the values of G_x, G_y, G_z, V_x0, V_y0 and V_z0 are solved by the method of claim 10 if the stored m sets of V_zm, V_ym, V_zm are sufficient to provide a full rank matrix.

13. An accuracy improvement unit for a 3-axis accelerometer as claimed in anyone of claim 9, further comprising

an output unit for indicating the count of m.

14. An accuracy improvement unit for a 3-axis accelerometer as claimed in anyone of claim 9, further comprising

an input device for triggering the start of a calibration.

15. An accuracy improvement unit for a 3-axis accelerometer as claimed in anyone of claim 9, further comprising

a time-based trigger for starting calibration at a pre-determined time.

16. An accuracy improvement unit for a 3-axis accelerometer as claimed in anyone of claim 9, wherein

the accuracy improvement unit determines that the 3-axis accelerometer is rotated into a new orientation before collecting the outputs V_x, V_y, V_z.

17. An accuracy improvement unit for a 3-axis accelerometer as claimed in anyone of claim 9, wherein:

the accuracy improvement unit determines that the 3-axis accelerometer is held still before collecting outputs V_x, V_y, V_z.

18. An accuracy improvement unit for a 3-axis accelerometer as claimed in claim 17, wherein

the accuracy improvement unit determines that the 3-axis accelerometer is held still if V_x, V_y, V_z indicate a composite vector within an acceptable value.

19. An accuracy improvement unit for a 3-axis accelerometer as claimed in claim 17, wherein

the accuracy improvement unit determines that the 3-axis accelerometer is held still in an orientation if a plurality of successive accelerometer outputs collected within a pre-determined period of time for each axis has a standard deviation within a pre-determined threshold;

wherein

the mean value of the plurality of successive accelerometer outputs for each axis is accepted as accelerometer outputs, V_x, V_y, V_z.

20. An accuracy improvement unit for a 3-axis accelerometer as claimed in anyone of claim 9, wherein

the accuracy improvement unit is a separate device to be connected to the 3-axis accelerometer for use.

21. An accuracy improvement unit for a 3-axis accelerometer as claimed in claim 9, wherein

the accuracy improvement unit is integrated with or embedded in the 3-axis accelerometer.

22. An accuracy improvement unit for a 3-axis accelerometer as claimed in claim 9, wherein

the accuracy improvement unit is integrated with or embedded in a processor which connects to the 3-axis accelerometer.