Using clustering techniques to improve magnetometer bias estimation

• US 9,423,252 B2
• Filed: 09/11/2012
• Issued: 08/23/2016
• Est. Priority Date: 09/11/2012
• Status: Active Grant

• Current Assignee: Apple Inc.
• Original Assignee: Apple Inc.
• Inventors: Chow, Sunny Kai Pang, Tu, Xiaoyuan
• Primary Examiner(s): LE, TOAN M
• Application Number:
  US13/610,678
• Publication Number:
  US 20140074429A1
• Time in Patent Office: 1,442 Days
• Field of Search: 702/154, 702/179
• US Class Current: 1/1
• CPC Class Codes: G01C 17/02 Magnetic compasses
• CPC Class Codes: G01C 17/38 Testing, calibrating, or compensating of compasses
• CPC Class Codes: G01R 33/0035 Calibration of single magnetic sensors, e.g. integrated calibration
• CPC Class Codes: G01R 33/028 Electrodynamic magnetometers
• CPC Class Codes: G06F 17/16 Matrix or vector computation , e.g. matrix-matrix or matrix-vector multiplication, matrix factorization matrix transposition G06F7/78
• CPC Class Codes: G06F 17/18 for evaluating statistical data , e.g. average values, frequency distributions, probability functions, regression analysis forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04

Abstract

In some implementations, a computer-implemented method includes receiving a reading from a magnetometer of a mobile device. A cluster from a plurality of clusters of bias offsets generated from previously-calibrated readings is selected. The selected cluster has a representative bias offset, a mean of magnitudes in the selected cluster, and a magnitude threshold. An external magnetic field is estimated based on the reading and the representative bias offset for the selected cluster. Whether a magnitude of the estimated external field is within a magnitude range defined by the mean magnitude and the mean magnitude plus the magnitude threshold is determined.

Claims

1. A computer-implemented method, comprising:

receiving, by a mobile device, a reading from a magnetometer of the mobile device;

selecting, by a processor of the mobile device, a cluster from a plurality of clusters of bias offsets, wherein;

each of the clusters includes a plurality of bias offsets including a representative bias offset,

the plurality of clusters are generated by clustering on bias vectors calculated from previous magnetometer readings, and

selecting the cluster from the clusters includes matching the reading from the magnetometer to a respective set of one or more mean values of each of the clusters and identifying the cluster from the clusters according to a respective set of one or more threshold values;

estimating, by the processor, a current external magnetic field based on the reading and the representative bias offset for the selected cluster; and

determining, by the processor, a heading for the mobile device using the estimated current external field.

2. The computer-implemented method of claim 1, further comprising:

determining that a magnitude of the estimated current external field is within a magnitude range defined by a mean magnitude for the selected cluster and the mean magnitude plus a magnitude threshold;

determining a gravitational vector;

determining an inclination angle between the gravitational vector and the estimated current external magnetic field; and

determining that the inclination angle is within an angle range defined by a mean inclination angle for the selected cluster and the mean angle plus an angle threshold.

3. The computer-implemented method of claim 2, further comprising:

determining the estimated current external magnetic field is not within the magnitude range or the inclination angle is not in the angle range;

iteratively estimating the current external field using representative bias offsets for different clusters in the plurality of clusters; and

for each iteration, determining whether the estimated current external field is in the magnitude range or the inclination angle is in the angle range for a cluster selected during that iteration.

4. The computer-implemented method of claim 1, further comprising:

receiving calibrated magnetometer data including a bias offset;

comparing the bias offset to the plurality of clusters;

determining the bias offset does not match any cluster in the plurality of clusters; and

identifying the bias offset as novel.

5. The computer-implemented method of claim 4, in response to determining a number of novel bias offsets exceeds a specified number, automatically applying a clustering technique to historical magnetometer data to generate new clusters of bias offsets.

6. The computer-implemented method of claim 1, wherein the plurality of clusters are formed using a quality threshold clustering.

7. The computer-implemented method of claim 1, wherein the current external magnetic field is estimated by subtracting the representative bias offset for the selected cluster from the readings.

8. The computer-implemented method of claim 1, wherein the magnitude threshold is based on a standard deviation of magnitudes in the selected cluster.

9. A computer program product encoded on a tangible, non-transitory storage medium, the product comprising computer readable instructions for causing one or more processors to perform operations comprising:

receiving, by a mobile device, a reading from a magnetometer of the mobile device;

selecting, by a processor of the mobile device, a cluster from a plurality of clusters of bias offsets, wherein;

each of the clusters includes a plurality of bias offsets including a representative bias offset,

the plurality of clusters are generated by clustering on bias vectors calculated from previous magnetometer readings, and

selecting the cluster from the clusters includes matching the reading from the magnetometer to a respective set of one or more mean values of each of the clusters and identifying the cluster from the clusters according to a respective set of one or more threshold values;

estimating, by the processor, a current external magnetic field based on the reading and a representative bias offset for the selected cluster; and

determining, by the processor, a heading for the mobile device using the estimated current external field.

10. The computer program product of claim 9, the instructions further comprising:

determining that a magnitude of the estimated current external field is within a magnitude range defined by a mean magnitude for the selected cluster and the mean magnitude plus a magnitude threshold;

determining a gravitational vector;

determining an inclination angle between the gravitational vector and the estimated current external magnetic field; and

determining that the inclination angle is within an angle range defined by a mean inclination angle for the selected cluster and the mean angle plus an angle threshold.

11. The computer program product of claim 9, the instructions further comprising:

determining the estimated current external magnetic field is not within the magnitude range or the inclination angle is not in the angle range;

iteratively estimating the current external field using representative bias offsets for different clusters in the plurality of clusters; and

for each iteration, determining whether the estimated current external field is in the magnitude range or the inclination angle is in the angle range for a cluster selected during that iteration.

12. The computer program product of claim 9, the instructions further comprising:

receiving calibrated magnetometer data including a bias offset;

comparing the bias offset to the plurality of clusters;

determining the bias offset does not match any cluster in the plurality of clusters; and

identifying the bias offset as novel.

13. The computer program product of claim 12, in response to determining a number of novel bias offsets exceeds a specified number, the instructions further comprising automatically applying a clustering technique to historical magnetometer data to generate new clusters of bias offsets.

14. The computer program product of claim 9, wherein the plurality of clusters are formed using a quality threshold clustering.

15. The computer program product of claim 9, wherein the current external magnetic field is estimated by subtracting the representative bias offset for the selected cluster from the readings.

16. The computer program product of claim 9, wherein the magnitude threshold is based on a standard deviation of magnitudes in the selected cluster.

17. A mobile device, comprising:

a magnetometer;

memory configured to a plurality of clusters of bias offsets generated from previously-calibrated readings; and

one or more processors configured to perform operations comprising;

receiving a reading from the magnetometer;

selecting a cluster from a plurality of clusters of bias offsets, wherein;

each of the clusters includes a plurality of bias offsets including a representative bias offset,

the plurality of clusters are generated by clustering on bias vectors calculated from previous magnetometer readings, and

selecting the cluster from the clusters includes matching the reading from the magnetometer to a respective set of one or more mean values of each of the clusters and identifying the cluster from the clusters according to a respective set of one or more threshold values;

estimating a current external magnetic field based on the reading and the representative bias offset for the selected cluster; and

determining a heading for the mobile device using the estimated current external field.

18. The mobile device of claim 17, the processors further configured to:

determine that a magnitude of the estimated current external field is within a magnitude range defined by a mean magnitude for the selected cluster and the mean magnitude plus a magnitude threshold;

determine a gravitational vector;

determine an inclination angle between the gravitational vector and the estimated current external magnetic field; and

determine that the inclination angle is within an angle range defined by a mean inclination angle for the selected cluster and the mean angle plus an angle threshold.

19. The mobile device of claim 17, the processors further configured to:

determine the estimated current external magnetic field is not within the magnitude range or the inclination angle is not in the angle range;

iteratively estimate the current external field using representative bias offsets for different clusters in the plurality of clusters; and

for each iteration, determine whether the estimated current external field is in the magnitude range or the inclination angle is in the angle range for a cluster selected during that iteration.

20. The mobile device of claim 17, the processors further configured to:

receive calibrated magnetometer data including a bias offset;

compare the bias offset to the plurality of clusters;

determine the bias offset does not match any cluster in the plurality of clusters; and

identify the bias offset as novel.

21. The mobile device of claim 20, in response to determine a number of novel basis offsets exceeds a specified number, the processors further configured to automatically apply a clustering technique to historical magnetometer data to generate new clusters of bias offsets.

22. The mobile device of claim 17, wherein the plurality of clusters are formed using a quality threshold clustering.

23. The mobile device of claim 17, wherein the current external magnetic field is estimated by subtracting the representative bias offset for the selected cluster from the readings.

24. The mobile device of claim 17, wherein the magnitude threshold is based on a standard deviation of magnitudes in the selected cluster.

25. The mobile device of claim 17, wherein each set of one or more mean values includes at least one of a mean magnitude value or a mean inclination angle value.

26. The mobile device of claim 25, wherein matching the reading from the magnetometer to a respective set of one or more mean values of each of the clusters and identifying the cluster from the clusters according to a respective set of one or more threshold values comprise, for each cluster:

determining a magnitude and an inclination angle for an estimated geomagnetic field; and

for each cluster;

determine whether the magnitude of the estimated geomagnetic field is within a range between the mean magnitude value of the cluster and the mean magnitude value plus a threshold value for the cluster;

determine whether the inclination angle of the estimated geomagnetic field is within a range between the mean inclination angle of the cluster and the mean inclination angle plus a threshold value for the cluster; and

identifying the cluster upon determining that both the magnitude and the inclination angle are in range.