US-12618677-B2 - Using magnetic-sensor data to correct for INS drift

US12618677B2US 12618677 B2US12618677 B2US 12618677B2US-12618677-B2

Abstract

In one embodiment, a method includes accessing a trajectory and comparing a first one of multiple portions of the trajectory with a second one of the portions. A greater level of confidence is associated with the first one of the portions than the second one of the portions. The method includes determining, based on the comparison, that the first and second ones of the portions approximately magnetically coincide with each other and are not approximately colocated with each other in the trajectory; comparing a first level of confidence associated with the first one of the portions with a second level of confidence associated with the second one of the portions; and, when the first level of confidence is greater than the second level of confidence, adjusting the second one of the portions to be approximately colocated with the first one of the portions in the trajectory.

Inventors

Alexandre TOUTOV
ANTON TOUTOV
Oleksiy Lutsyk
Maryna MUKHINA

Assignees

Astra Navigation, Inc.

Dates

Publication Date: 20260505
Application Date: 20241107

Claims (20)

1 . A method comprising: by an electronic device, accessing a trajectory; by the electronic device, comparing a first one of a plurality of portions of the trajectory with a second one of the portions, wherein a greater level of confidence is associated with the first one of the portions than the second one of the portions; by the electronic device, determining based on the comparison that the first and second ones of the portions approximately magnetically coincide with each other and are not approximately colocated with each other in the trajectory; by the electronic device, comparing a first level of confidence associated with the first one of the portions with a second level of confidence associated with the second one of the portions; and by the electronic device, when the first level of confidence is greater than the second level of confidence, adjusting the second one of the portions to be approximately colocated with the first one of the portions in the trajectory.
2 . The method of claim 1 , wherein, when the second one of the portions is adjusted to be approximately colocated with the first one of the portions in the trajectory, relative headings of points along the trajectory are kept approximately constant and one or more locations of one or more other portions of the trajectory are adjusted consistent with the relative headings and the adjustment of the second one of the portions.
3 . The method of claim 1 , wherein: each of the portions is an approximately straight segment of the trajectory between two turns in the trajectory; and each of the turns comprises a change in heading along the trajectory.
4 . The method of claim 1 , wherein comparing the first one of the portions with the second one of the portions comprises: accessing a first series of first magnetic values that represents a first magnetic recording corresponding to the first one of the portions, wherein the first magnetic recording comprises a first set of first magnetic measurements that are each represented by one of the first magnetic values; accessing a second series of second magnetic values that represents a second magnetic recording corresponding to the second one of the portions, wherein the second magnetic recording comprises a second set of second magnetic measurements that are each represented by one of the second magnetic values; by an electronic device, approximately aligning the first and second series with each other; by an electronic device, calculating a difference between the first and second series as aligned with each other; and by an electronic device, determining a similarity between the first and second series based on the difference.
5 . The method of claim 4 , wherein: the first and second sets are both time series or the first and second sets are both distance series; the difference comprises a sum of one or more Euclidean distances, dynamic time warping (DTW) distances, or Fréchet distances between corresponding first and second magnetic values; and determining the similarity between the first and second series based on the difference comprises: comparing the difference with a predetermined threshold; and determining that the first and second series are similar to each other when the difference is approximately less than the predetermined threshold.
6 . The method of claim 4 , wherein the first and second magnetic measurements comprise one or more of a total value of a magnetic field, an inclination angle of the magnetic field, a declination angle of the magnetic field, an x component of the magnetic field, a y component of the magnetic field, or a z component of the magnetic field or a magnetic-susceptibility or magnetic-conductivity value.
7 . The method of claim 4 , wherein: each of the first and second magnetic measurements comprises a plurality of magnetic components; and each of the first and second magnetic values is a square root of a sum of squares of the magnetic components of the first or second magnetic measurement represented by the first or second magnetic value.
8 . The method of claim 4 , wherein approximately aligning the first and second series with each other comprises spline interpolation of the first and second series to produce a same number of elements in the first and second series.
9 . The method of claim 4 , wherein approximately aligning the first and second series with each other comprises scaling the first and second series by dividing each of the first magnetic values by a first weighted average of the first magnetic values and dividing each of the second magnetic values by a second weighted average of the second magnetic values.
10 . The method of claim 1 , further comprising, when the second level of confidence is greater than the first level of confidence, adjusting the first one of the portions to be approximately colocated with the second one of the portions in the trajectory.
11 . The method of claim 1 , wherein the greater level of confidence is associated with the first one of the portions at least in part because the first one of the portions is closer than the second one of the portions to a start point of the trajectory.
12 . One or more computer-readable non-transitory storage media embodying software that is operable when executed to: access a trajectory; compare a first one of a plurality of portions of the trajectory with a second one of the portions, wherein a greater level of confidence is associated with the first one of the portions than the second one of the portions; determine based on the comparison that the first and second ones of the portions approximately magnetically coincide with each other and are not approximately colocated with each other in the trajectory; compare a first level of confidence associated with the first one of the portions with a second level of confidence associated with the second one of the portions; and when the first level of confidence is greater than the second level of confidence, adjust the second one of the portions to be approximately colocated with the first one of the portions in the trajectory.
13 . The media of claim 12 , wherein, when the second one of the portions is adjusted to be approximately colocated with the first one of the portions in the trajectory, relative headings of points along the trajectory are kept approximately constant and one or more locations of one or more other portions of the trajectory are adjusted consistent with the relative headings and the adjustment of the second one of the portions.
14 . The media of claim 12 , wherein: each of the portions is an approximately straight segment of the trajectory between two turns in the trajectory; and each of the turns comprises a change in heading along the trajectory.
15 . The media of claim 12 , wherein comparing the first one of the portions with the second one of the portions comprises: accessing a first series of first magnetic values that represents a first magnetic recording corresponding to the first one of the portions, wherein the first magnetic recording comprises a first set of first magnetic measurements that are each represented by one of the first magnetic values; accessing a second series of second magnetic values that represents a second magnetic recording corresponding to the second one of the portions, wherein the second magnetic recording comprises a second set of second magnetic measurements that are each represented by one of the second magnetic values; by an electronic device, approximately aligning the first and second series with each other; by an electronic device, calculating a difference between the first and second series as aligned with each other; and by an electronic device, determining a similarity between the first and second series based on the difference.
16 . The media of claim 15 , wherein: the first and second sets are both time series or the first and second sets are both distance series; the difference comprises a sum of one or more Euclidean distances, dynamic time warping (DTW) distances, or Fréchet distances between corresponding first and second magnetic values; and determining the similarity between the first and second series based on the difference comprises: comparing the difference with a predetermined threshold; and determining that the first and second series are similar to each other when the difference is approximately less than the predetermined threshold.
17 . The media of claim 15 , wherein: each of the first and second magnetic measurements comprises a plurality of magnetic components; and each of the first and second magnetic values is a square root of a sum of squares of the magnetic components of the first or second magnetic measurement represented by the first or second magnetic value.
18 . The media of claim 15 , wherein approximately aligning the first and second series with each other comprises spline interpolation of the first and second series to produce a same number of elements in the first and second series.
19 . The media of claim 15 , wherein approximately aligning the first and second series with each other comprises scaling the first and second series by dividing each of the first magnetic values by a first weighted average of the first magnetic values and dividing each of the second magnetic values by a second weighted average of the second magnetic values.
20 . The media of claim 12 , wherein the software is further operable when executed to, when the second level of confidence is greater than the first level of confidence, adjust the first one of the portions to be approximately colocated with the second one of the portions in the trajectory.

Description

PRIORITY This application is a continuation, under 35 U.S.C. § 120, of U.S. patent application Ser. No. 18/795,387, filed 6 Aug. 2024, which claims the benefit, under 35 U.S.C. § 119(e), of U.S. Provisional Patent Application No. 63/531,316, filed 8 Aug. 2023, which is incorporated herein by reference. TECHNICAL FIELD This disclosure generally relates to the use of magnetic sensors. BACKGROUND The Earth's geomagnetic field (GMF) is a continuous potential field that is locally unique due to its physical nature and the influence of surrounding objects (e.g. buildings, ferromagnetic materials, pillars, etc.). The GMF is stable, and changes over time are predictable and well documented. Magnetic anomalies can lead to local anomalies in the GMF. As a result, the GMF is a source of specific data that can be recorded, processed, and analyzed. A fusion filter is an algorithm that may be used in sensor fusion to combine data from multiple sensors, resulting in a more accurate estimate of a system being measured. The goal of a fusion filter is to reduce errors and uncertainty and thereby improve the accuracy and reliability of the system. A fusion filter may combine data from multiple sensors measuring the same variable(s), such as position or orientation, and then apply statistical techniques to that data to estimate the most likely value of the variable(s). An advantage of fusion filters is their ability to deal with noisy or unreliable sensor data. By combining data from multiple sources, a fusion filter may reduce the impact of any individual sensor's errors or biases, resulting in a more accurate estimate of the system being measured. A magnetic field in an indoor environment tends to be significantly non-uniform and anomalous. As a result, when a fusion filter is used in such an environment, the fusion filter will tend to function incorrectly and orientation (or one or more other variables) will not be able to be determined accurately. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 illustrates example measurements of an example magnetic field at an example point in space. FIG. 2 illustrates an example system for authenticating a user or device using a magnetic recording. FIG. 3 illustrates example plots of example magnetic recordings of an example trajectory. FIG. 4 illustrates an example two-dimensional trajectory. FIG. 5 illustrates example plots of example magnetic recordings of the trajectory of FIG. 4. FIGS. 6A-6B and 7A-7B illustrate example conversion of example time series of example magnetic recordings to example distance series. FIG. 8 illustrates example scaling of the plots of FIG. 5. FIG. 9 illustrates another example two-dimensional trajectory. FIG. 10 illustrates example plots of example magnetic recordings of the trajectory of FIG. 9. FIG. 11 illustrates another example two-dimensional trajectory. FIG. 12 illustrates an example plot of an example magnetic recording of the trajectory of FIG. 11. FIG. 13 illustrates an example two-dimensional reference trajectory and example two-dimensional candidate trajectories offset from the reference trajectory. FIG. 14 illustrates example plots of example magnetic recordings of the trajectories of FIG. 13. FIGS. 15A and 15B illustrate other example two-dimensional trajectories. FIGS. 16A-16F illustrate example comparisons of example plots of four example magnetic recordings of the trajectory of FIG. 15A with each other. FIGS. 17A-17C illustrate example comparisons of example plots of three example magnetic recordings of the trajectory of FIG. 15B with each other. FIGS. 18A-18F illustrate example comparisons of example plots of example magnetic recordings of the trajectory of FIG. 15A with example magnetic recordings of the trajectory of FIG. 15B. FIG. 19 illustrates an example method for authentication using a magnetic recording. FIG. 20 illustrates an example device including a magnetic-sensor system and inertial navigation system (INS). FIGS. 21A-21T illustrate example correction of INS drift. FIG. 22 illustrates an example magnetometer. FIGS. 23A-23B illustrate an example module with an example magnetometer. FIG. 24 illustrates an example sensor set including four example magnetometers. FIG. 25 illustrates an example sensor set including eight example magnetometers. FIG. 26 illustrates an example sensor set including magnetometers for measuring a distance traversed or speed. FIG. 27 illustrates example magnetic measurements by an example pair of magnetometers. FIG. 28 illustrates an example system for controlling motion with magnetometers. FIG. 29 illustrates an example method for controlling motion with magnetometers. FIG. 30 illustrates an example method for measuring a distance traversed or a speed. FIG. 31 illustrates an example method for determining confidence in magnetic-sensor data. FIG. 32 illustrates an example method for using magnetic-sensor data to correct for INS drift. FIG. 33 illustrates example axes of an example smartphone. FIG. 34 illustrates