HK1166125A - Automatic data collection algorithm for 3d magnetic field calibration with reduced memory requirements - Google Patents

Description

Automatic data acquisition algorithm for 3D magnetic field calibration with reduced memory requirements

Background

Magnetic compasses are often integrated with other components that distort the earth's magnetic field, compromising the accuracy of the compass. These disturbances are typically corrected using a field compensation mechanism that determines compensation coefficients to correct the magnetic field readings. The quality of the compensation coefficients depends on the data set of the magnetic field samples taken, which typically vary enough in the orientation (orientation) of the samples across three-dimensional space. These methods may require the user to perform tedious and time consuming tasks, such as holding the device steady in several orientations while the data is being collected, moving the device in a predetermined manner, or manually selecting the data. Manually entering data is often burdensome because typical embedded applications have few user interface elements for user input. Furthermore, a large memory capacity may be required to store enough data samples in embedded applications where small size is often important.

Disclosure of Invention

One embodiment is directed to a magnetic compass that includes a magnetometer for taking magnetic field readings and a processing unit that calibrates the magnetic compass. The processing unit is configured to verify (valid) a predetermined number of magnetic field samples and to calculate calibration coefficients from the verified magnetic field samples. Each validated magnetic field sample is separated from each other validated magnetic field sample by at least a minimum separation angle.

Drawings

FIG. 1 is a block diagram of one embodiment of a magnetic compass integrated in a device.

Fig. 2A is a diagram illustrating hard iron interference (hard iron interference) of a magnetic field.

FIG. 2B is a diagram illustrating an example set of sample vectors of an interfered magnetic field.

FIG. 3 is a flow diagram of one embodiment of a method of verifying a magnetic field sample.

FIG. 4 is a flow chart of one embodiment of a method of calibrating a magnetic compass.

Like reference numbers and designations in the various drawings indicate like elements.

Detailed Description

Embodiments described herein calibrate a magnetic compass to compensate for hard-iron interference in the earth's magnetic field due to devices and components integrated with the magnetic compass. In some embodiments, samples of the magnetic field that are separated from all other samples by at least a minimum separation angle are used to calculate the compensation factor. In some embodiments, the minimum separation angle and the total number of samples are predetermined to ensure a three-dimensional span (span) of the magnetic field.

FIG. 1 is a block diagram of one embodiment of a magnetic compass 120 integrated in a device 100. The magnetic compass 120 includes at least one magnetometer 122 that measures the magnetic field to which the magnetometer is exposed and provides navigational (heading) information to the device 100. The magnetic compass 120 further comprises a processing unit 102 and a memory 104. In the embodiment shown in fig. 1, the magnetic compass 120 further comprises at least one accelerometer 124. The device 100 includes functional circuitry 142 and a display device 110. The device 100 is any system or device that uses guidance information, such as, for example, a navigation device, a vehicle, or any other device. The functional circuitry 142 is any circuitry that uses guidance information, for example, for navigation or aiming. The components of device 100 are communicatively coupled to each other using suitable interfaces and interconnections as desired.

For example, the display device 110 displays a magnetic compass reading or requests user input. Examples of display device 110 include a digital display, an LCD monitor, an LED display, or the like. The user interface 140 is integrated with the display device 110 and includes physical or logical buttons for user input.

In the embodiment illustrated in fig. 1, the calibration routine 134 and the calculation routine 133 are implemented in software 132 executed by the processing unit 102. The software 132 comprises program instructions stored on a suitable storage device or medium 130. Suitable storage devices or media 130 include, for example, forms of non-volatile memory including, by way of example, semiconductor memory devices (e.g., erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), and flash memory devices), magnetic disks (e.g., a local hard disk and a removable magnetic disk), and optical disks (e.g., a compact disk read-only memory (CD-ROM) disk). Further, the storage device or medium 130 need not be local to the device 100. In one embodiment, the storage device or medium 130 is integrated into the magnetic compass 120. Typically, a portion of the software 132 executed by the processing unit 102 and one or more data structures used by the software 132 during execution are stored in the memory 104. In one implementation of such an embodiment, the memory 104 comprises any suitable form of Random Access Memory (RAM), now known or later developed, such as Dynamic Random Access Memory (DRAM). In other embodiments, other types of memory are used.

The calculation routine 133 calculates the heading from the magnetic compass readings. In other embodiments, the calculation routine 133 also calculates roll and pitch (pitch) of the device 100. Accurate compass readings involve compensating for the changing magnetic field of the operating environment of the integrated magnetic compass 120. For example, even though the magnetic compass 120 may be calibrated at the factory with maximum accuracy, once it is integrated with other components, interference introduced by the surrounding environment affects the magnetic compass 120. The calibration routine 134 compensates for magnetic interference with the operating environment of the magnetic compass 120.

The calibration routine 134 includes a selection routine 136 and a compensation routine 138. The selection routine 136 validates a plurality of magnetic field samples that are separated from each other by at least a minimum separation angle. Verifying that the magnetic field samples select samples for calibrating the magnetic compass and that the samples are stored in the memory 104; while invalid (invalid) magnetic field samples are discarded. The minimum separation angle is the minimum angle value for the angle between the validation magnetic field samples and each validated magnetic field sample. The compensation routine 138 calculates compensation coefficients from the validated magnetic field samples to compensate for hard-iron interference. The compensation routine 130 is calculated using any now known or later developed method, including, for example, a numerical search or a least squares method.

Hard iron interference originates from the permanent magnets and magnetized material on the compass platform. These disturbances remain constant and are fixed in all of the guiding orientations relative to the magnetic compass 120 for a given installation. In a magnetic field without interference, the earth's magnetic field has a constant magnitude. When rendered in 3D, the interference free field is a sphere centered at the origin. The presence of hard iron interference increases the constant magnitude field component that shifts the center of the ball. Soft iron interference arises from the interaction between the earth's magnetic field and soft magnetic materials in the vicinity of the compass. Soft iron interference causes the ball to deform into an ellipsoid and depends on the compass 120 orientation. The combination of soft and hard iron interference will produce an ellipsoid with the center displaced from the origin. The ellipsoid may also be rotated depending on the nature of the disturbance. The calibration routine 134 transforms the ellipsoidal interfering magnetic field into a sphere positioned at the origin.

Fig. 2A is a graph 200 illustrating hard-iron interference with a magnetic field. For simplicity, this diagram 200 is shown as a two-dimensional (2D) projection of a three-dimensional (3D) magnetic field, and the subject matter discussed herein may be extended to 3D. The illustrated unperturbed magnetic field 210 is centered about an origin and has a radius H_m. In one embodiment, unperturbed magnetic field 210 is the earth's magnetic field. In this example, the hard iron interference causes the unperturbed magnetic field 210 to move H in the positive x-direction_xMoving H in the positive y-direction_y. When the undisturbed magnetic field 210 is disturbed by hard iron interference, the center of the sphere moves, but the shape of the disturbed magnetic field 220 remains with radius H_mThe ball of (1). Magnetic compass measurements integrated with hard iron interference to (H)_x，H_y) A central disturbed magnetic field 220.

The angle between the origin and the highest and lowest points on the y-axis of the disturbed magnetic field 220, measured from the x-axis, is θ_max222 and theta_min224. These angles are given as follows:

for taking N samples, the minimum separation angle Δ θ may take a maximum value:

in one embodiment, the samples are at 0, Δ θ, 2 Δ θ, 3 Δ θ, etc. up to (N-1) Δ θ, and in other embodiments, the samples are at least Δ θ apart. As can be shown in equation (3), the values of N and Δ θ are selected to satisfy the calibration routine 136. The criteria for selecting N and Δ θ include considerations based on available memory, computational power, and data span of the 3D space.

Fig. 2B is a graph 240 illustrating an interfered magnetic field 250. An example set of sample vectors 252-1 through 252-8 is validated against disturbed magnetic field 250. In this example, the selection routine takes samples and requires 8 samples with a minimum separation angle Δ θ of 20 degrees. In one embodiment, Δ θ and N are selected to ensure that the magnetic field samples are taken out of a single plane and thus span 3D.

FIG. 3 is a flow diagram of one embodiment of a method 300 of verifying a magnetic field sample. A selection routine is initiated (block 302). At startup, a first magnetic field sample, referred to as the test vector Vt, is taken. Because this is the first test vector taken, it is automatically verified and saved as v (0) and the sample count n is set to 1 (block 304). In this embodiment, a total number N of magnetic field samples with the smallest separation angle Δ θ are validated and saved for calculation of the calibration coefficients. In one embodiment, the first sample v (0) taken is always acceptable when n is 1.

The method 300 queries whether a total of N samples have been taken, in other words, whether N equals N (block 310). If the total number N of samples have been verified, the compensation routine begins calculating calibration coefficients (block 312). If N ≠ N, another magnetic compass reading, Vt, is taken (Block 320).

Verification of the test vectors is based on the minimum angular separation of all magnetic vectors. Thus, test vector Vt will be compared to previously validated magnetic field samples (e.g., v (0), v (1), etc.) to ensure what is referred to as between Vt and validated sample v (i)Is greater than or equal to delta theta. To ensure that the test vector Vt passes on all other samples v (i), the index used to count the previously passing samples, i, is set to 0 to start the verification process (block 322).

The method 300 determines whether i < N (block 324). If i is less than N, determining the angle between the test vector Vt and the verified sample v (i)Whether it is greater than or equal to Δ θ (block 330). For v (i), ifThe sample is rejected and another sample is taken (block 320). When the test vector fails, it is not stored in memory for use in calculating the calibration coefficients. For v (i), ifThe sample passes on v (i) but must still be validated on all other stored magnetic field samples v (i). Thus, for the comparison between Vt and the next verified sample, i is incremented by 1 (block 332), and the loop for the comparison between the test vector and v (i) begins again (block 324).

In one embodiment, the angles are separatedIs calculated by using the inner product between the unit vectors, V and u (i) point in the same direction as the magnetic samples Vt and V (i). For each previously validated magnetic vector sample v (I) (e.g., v (0), v (1), etc.), the corresponding set of unit vectors is stored as u (I), I ═ 1. The test vector will only be validated and saved if and only if:

acos (< V, u (i) >) ≧ Δ θ 1. (4)

Once the cover is closedN, the angle between the test vector and all previously validated samples,have been compared. In other words, once equation 4 is true, U (k +1) ═ V and the magnetic sample test vector Vt are stored as valid data points. In other words, when i is not less than n (block 342), Vt is stored as v (n) (block 340). N is then incremented by 1 (block 342) and a determination is made as to whether a maximum number N of samples has been stored by comparing whether N-N (block 310). Once the number of verified samples equals the desired number of samples N, data acquisition is complete and the calculation routine begins calculating calibration coefficients (block 312).

FIG. 4 is a flow chart of one embodiment of a method 400 of calibrating a magnetic compass 120. The calibration routine 134 is initiated (block 410), for example, upon a user command or power-up of the magnetic compass. In one embodiment, an indication is made that the selection routine 136 is running. In another embodiment, an indication is made to the user for the integrated magnetic compass 120. In yet another embodiment, the user selects how many total samples N are and what the minimum separation angle Δ θ is to be taken. In some exemplary embodiments, N is equal to 6 to 24, but may be any suitable number.

Once the calibration routine has started, the selection routine 412 begins. The selection routine includes rotating the magnetic compass 120 while taking the magnetic field samples (block 420). The selection routine 412 further includes verifying each magnetic field sample that is separated from all other samples by a minimum separation angle Δ θ (block 430). In one embodiment, the first magnetic field sample is automatically validated. Magnetic field samples are validated by recursively comparing the angle between the magnetic field sample and all previously validated magnetic field samples to ensure that the magnetic field sample is separated from all previously validated magnetic field samples by at least a minimum separation angle.

When the magnetic field sample is verified, it is stored in memory 104 (block 440). In another embodiment, a message is displayed on the display device 110 indicating that the first sample has been verified and stored. The magnetic field samples are validated and stored until a total number N of samples is reached. In one embodiment, an indication that N samples have been collected is displayed on the display device 110. In another embodiment, if the device 100 is not rotated such that N samples are verified within a predetermined time, a timeout message is displayed on the display device 110 and the selection routine terminates.

Once all N samples have been validated and stored, the selection routine 412 is complete. The compensation routine 138 begins calculating calibration coefficients (block 450). The compensation routine 138 develops a transformation matrix that maps the disturbed magnetic field 220 to a sphere centered at the origin. The magnetic compass 120 is calibrated using the calibration coefficients (block 460). That is, the calibration coefficients are applied to future magnetic field samples to compensate for magnetic field disturbances. In another embodiment, the magnetic compass heading is calculated by the resulting calibration coefficients.

In one embodiment, the display device 110 displays a message indicating that the calibration coefficients are determined and asks the user if the calibration coefficients should be used. If the user has entered that the calibration coefficients are to be used, device 110 writes the calibration coefficients to a memory, such as memory 104.

In one embodiment of the method 400, the sample is taken autonomously without any user intervention. In another embodiment, the user may abort the calibration routine 134 at any time. In yet another embodiment, the device 100 does not display any messages regarding the progress of the calibration routine 134. In one embodiment, the method 400 is complete whenever the magnetic compass 120 is mounted on another device 100 or a source of magnetic interference is added or removed from the device 100.

The embodiments described herein are applicable to any application that uses magnetic compass data, including but not limited to navigation, pointing applications, and gun-mounted applications. Embodiments provided herein reduce the number of user inputs required and reduce the memory required for data storage because fewer data points are stored. One embodiment of magnetic compass calibration is used in embedded applications where memory is scarce and user input can be burdensome and complex, such as handheld applications. An algorithm was developed to automate the data acquisition to ensure a wide span of 3D magnetic fields.

There have been described various embodiments of the invention defined by the following claims. Nevertheless, it will be understood that various modifications to the described embodiments may be made without departing from the spirit and scope of the claimed invention. Accordingly, other embodiments are within the scope of the following claims.

Claims (3)

1. A program product for calibrating a magnetic compass (120), the program product comprising a processor-readable medium (130) having program instructions embodied thereon, wherein the program instructions, when executed by at least one programmable processor (102) coupled to the magnetic compass, are operable to cause the magnetic compass to:

taking a plurality of magnetic field samples (252-1 to 252-8);

validating the plurality of magnetic field samples (430), wherein validating comprises:

validating the magnetic field sample by recursively comparing angles between the magnetic field sample and all previously validated magnetic field samples to ensure that the magnetic field sample is separated from all previously validated magnetic field samples by at least a minimum separation angle; and

calculating calibration coefficients from the validated magnetic field samples;

and

calibrating the magnetic compass using the calibration coefficients.

2. The program product of claim 1, wherein the program instructions are further operable to cause the magnetic compass to:

verifying the first magnetic field sample; and

when both the angle between the next magnetic field sample and the first magnetic field sample and the angle between the next magnetic field sample and the second magnetic field sample are greater than or equal to the minimum separation angle, the next magnetic field sample is validated.

3. A magnetic compass (120) comprising:

a magnetometer (122) for taking magnetic field readings; and

a processing unit (102) for calibrating a magnetic compass, wherein the processing unit is configured to:

verifying a predetermined number of magnetic field samples (252-1 to 252-8), wherein each verified magnetic field sample is separated from each other verified magnetic field sample (430) by at least a minimum separation angle; and

calibration coefficients are calculated from the validated magnetic field samples (450).