IMU accelerometer modulus length out-of-tolerance suppression method based on constraint

Abstract

The invention relates to a constraint-based IMU (inertial measurement unit) accelerometer modulus length out-of-tolerance suppression method, which comprises the following steps of: 1, starting, electrifying an IMU, outputting original data of an accelerometer and a gyroscope, and setting an accelerometer modulus length constraint identification bit and maximum and small out-of-tolerance thresholds of the accelerometer modulus length out-of-tolerance; 2, directly outputting the original data of the accelerometer and the gyroscope if the modulus length constraint identification bit is invalid, and triggering the modulus length constraint identification bit and calculating the modulus length difference value of the accelerometer if the constraint identification bit is valid; 3, if the mode length difference value is larger than the maximum threshold value, a fault is reported, and when the mode length difference value is smaller than the minimum threshold value, zero offset of the accelerometer is achieved, and an original accelerometer observation result is output; 4, when the modulus length difference value is between the maximum threshold value and the small threshold value, dynamic and static judgment is carried out, when the accelerometer is in a static state, accelerometer modulus length constraint is carried out, and 5, normalization processing is carried out on output of the accelerometer, zero offset values needing to be corrected of the three axes are calculated respectively, and measured values of the accelerometer are output after being corrected. According to the invention, the accelerometer modulus length constraint function in any static state can be realized, and each vector only changes in size before and after correction, is not changed in direction, and is the same as the gravitational acceleration 1g in size.

Description

Technical Field

[0001] This invention relates to the detection and calibration of inertial measurement units (IMUs). Specifically, it relates to methods for modulus constraint and modulus deviation suppression of static accelerometers in IMUs. More specifically, it relates to a method for suppressing modulus deviation of static accelerometers in IMUs through adaptive constraint. That is, by using the self-test mechanism after the IMU is powered on, abnormalities in the modulus of the static accelerometer are monitored and analyzed in real time. Adaptive constraints are applied to different detected modulus deviations of different accelerometers to restore the modulus to normal, thereby effectively improving the detection and calibration efficiency of the IMU and enhancing its stability and reliability during long-term use. Background Technology

[0002] Inertial measurement units (IMUs) undergo rigorous testing and parameter calibration at the factory to ensure that the modulus deviation of the accelerometer is controlled within a certain range, thereby effectively improving the stability and reliability of the IMU in long-term use.

[0003] Modulus length deviation is the absolute value of the difference between the modulus length of the accelerometer in a static state and the acceleration due to gravity g. It must typically be strictly controlled within 1 mg (1 mg ≈ 0.001 × 9.80 m / s²). However, due to changes in the storage environment (such as temperature, humidity, magnetic fields, etc.) and other factors, the modulus length (in static state) of a very small number of IMUs may fluctuate irregularly after the customer receives the parameter-calibrated product.

[0004] If the modulus of the accelerometer cannot be controlled within a certain range, that is, if the modulus does not match the Earth's gravitational acceleration g, the correction results will not only fail to reflect the true gravitational modulus, but will also introduce additional errors. In severe cases, it may even distort the accelerometer output data and make it unusable, thereby affecting the output results of the IMU and the accuracy of the entire navigation system.

[0005] Such magnitude deviations can only be observed when the accelerometer is completely stationary. When the IMU is in motion, the magnitude deviations are usually negligible because they are on a smaller order of magnitude than the external acceleration, and therefore do not affect the normal use of the IMU.

[0006] To ensure the stability of the accelerometer module length during long-term use of the IMU, major IMU manufacturers generally employ module length constraint technology. (Existing module length constraint technology is mainly used when the IMU is static and the module length deviates significantly from the acceptable range. It adjusts the Z-axis acceleration value to reduce the overall module length to meet requirements. Alternatively, it performs a simple module length adjustment, which involves only adjusting the Z-axis acceleration value to reduce the overall module length to meet requirements.) On the other hand, the modulus constraint technique works effectively when the IMU is stationary and placed approximately horizontally. It can adjust the stationary data output by the accelerometer. In this case, the direction of gravitational acceleration is parallel to the accelerometer measurement axis, and errors can be eliminated by simple modulus adjustment.

[0007] However, with the changing application scenarios of IMUs, especially as IMUs are increasingly used in fields such as robotics and automobiles, tilted installation of IMUs is becoming more and more common. When the IMU is placed at an angle, existing module length constraint methods will not work effectively.

[0008] When the IMU is tilted, the direction of gravitational acceleration is no longer parallel to the accelerometer measurement axis, and existing modulus constraint schemes cannot accurately correct for this. Therefore, existing modulus constraint techniques only work when the IMU is stationary and approximately horizontal. When the IMU is installed at an angle, existing schemes incorrectly use non-gravitational components for correction, causing a shift in the accelerometer measurement direction. In this case, the correction result not only fails to reflect the true gravitational modulus but also introduces additional errors, potentially distorting the accelerometer output data and rendering it unusable, thus affecting the IMU's output and the accuracy of the entire navigation system.

[0009] Therefore, the industry needs a method that can monitor and analyze the accelerometer module length in real time when the IMU is installed at an angle to detect any abnormalities. If any abnormalities are found, an adaptive constraint method can be used to suppress the abnormality of the IMU accelerometer module length, thereby effectively improving the stability and reliability of the IMU in long-term use.

[0010] Typically, methods for suppressing accelerometer modulus deviation (i.e., abnormal output signal amplitude) mainly include static calibration, dynamic threshold adjustment, temperature compensation, bias compensation, multi-sensor fusion algorithms, and finite element simulation optimization. These methods suppress noise and errors by adjusting parameters or algorithmic strategies in real time. This invention employs static calibration, dynamic threshold algorithms, and bias compensation. Summary of the Invention

[0011] To overcome the aforementioned problems, the present invention aims to provide a constraint-based method for suppressing IMU accelerometer modulus length deviation. This invention utilizes a self-test mechanism after the IMU (static) power-on to monitor and analyze in real time whether there are any abnormalities in the accelerometer modulus length. If any abnormality is detected, adaptive constraints are applied to the detected modulus length abnormality, thereby effectively improving the stability and reliability of the device during long-term use.

[0012] The technical solution of the constraint-based (static) IMU accelerometer modulus length deviation suppression method of the present invention is as follows: A constraint-based method for suppressing IMU accelerometer modulus length deviation is characterized by comprising the following steps: Step 1. Start-up: Power on the IMU and output the raw data of the linear acceleration (acc_x, acc_y, acc_z) of the accelerometer along the XYZ axes and the angular velocity (gyro_x, gyro_y, gyro_z) of the gyroscope. Configure the parameter file to set the accelerometer module length constraint flag and the maximum and minimum out-of-tolerance thresholds for the accelerometer module length. Step 2. If the algorithm entry point detects that the accelerometer modulus constraint flag is invalid, the accelerometer directly sends the raw data from the accelerometer and gyroscope. If the accelerometer constraint flag is valid, the accelerometer module length constraint flag is triggered, and the accelerometer module length difference is calculated. Let the accelerometer measurements on the x, y, and z axes be (a_x), (a_y), and (a_z), respectively. Then the formula for calculating the accelerometer's modulus (a) is: , The actual modulus at rest is equal to g (the Earth's gravitational acceleration constant). The difference between these two values ​​is the modulus difference. Calculate the current acceleration modulus : If it deviates from 1g, normalization is performed: , , Step 3. If the accelerometer modulus difference is greater than the maximum out-of-tolerance threshold, report a fault directly; If the accelerometer modulus is less than the minimum out-of-tolerance threshold, the accelerometer zero bias is directly set to zero. Output the original linear accelerations (acc_x, acc_y, acc_z) of the XYZ axes from the accelerometer. Step 4. When the calculated accelerometer modulus difference is between the minimum and maximum out-of-tolerance thresholds, proceed to IMU dynamic / static judgment: When the IMU is in a dynamic state, it is not processed and is ignored; When the IMU is in a static state, the accelerometer module length is constrained: Step 5: Normalize the accelerometer output, calculate the zero bias values ​​that need to be corrected for each of the three axes, and finally output the accelerometer measurement values ​​after correction (zero bias values).

[0013] According to the present invention, an algorithm and parameter file are set for limiting the measurement range or output data of an accelerometer. The algorithm includes an accelerometer module length constraint flag, and the parameter file includes setting constraint threshold parameters and various enable / disable parameters. In the parameter file, the maximum and minimum out-of-tolerance thresholds for the accelerometer module length are set.

[0014] The algorithm entry point will check whether the flag is true or false. The specific true or false flag is configured by the user and written into the configuration file.

[0015] IMU dynamic / static judgment: The common practice is to use the observation data of accelerometer and gyroscope to make a sliding window judgment, and to distinguish between dynamic and static by setting dynamic / static thresholds.

[0016] Zero-biasing the accelerometer to zero means setting the fixed error of the non-zero reading output by the sensor when there is no external acceleration, in order to improve measurement accuracy, eliminate systematic errors and enhance environmental adaptability.

[0017] According to the present invention, if the accelerometer modulus error is less than the minimum out-of-tolerance threshold, the accelerometer zero bias is directly set to zero, which refers to the fixed error of the non-zero reading output by the sensor when there is no external acceleration, so as to improve measurement accuracy, eliminate system errors and enhance environmental adaptability.

[0018] According to the constraint-based IMU accelerometer modulus length out-of-tolerance method described in this invention, the method is characterized in that... If the IMU module length exceeds the tolerance by 10mg and fails the calibration test, installation angle calibration is required. At this time, power on the IMU to trigger the module length constraint flag to start.

[0019] According to the constraint-based IMU accelerometer modulus length out-of-tolerance method described in this invention, the method is characterized in that... The minimum exceedance threshold is set to 6 mg, and the maximum exceedance threshold is set to 20 mg.

[0020] The threshold values ​​of 20mg and 6mg are common and within the range of experience, but can be freely configured according to the project indicators.

[0021] According to the constraint-based IMU accelerometer modulus length out-of-tolerance method described in this invention, the method is characterized in that... The installation angle calibration is performed under static conditions during dynamic and static testing, triggering the module length constraint normalization program, estimating the three-axis correction zero bias, and outputting the corrected data.

[0022] According to the constraint-based IMU accelerometer modulus length out-of-tolerance method described in this invention, the method is characterized in that... Step 5, the corrected data includes: Corrected accelerometer output vector , Acceleration-corrected XYZ axis observations , , , Corrected pitch angle value compared to the original value , , Corrected value of roll angle compared to original value , , raw accelerometer output vector The actual accelerometer vector It is the zero bias of the accelerometer. .

[0023] According to the constraint-based IMU accelerometer modulus length out-of-tolerance method described in this invention, the method is characterized in that... The algorithm and proof process of the constraint-based IMU accelerometer modulus length out-of-tolerance suppression method are as follows: accelerometer output vector after settlement correction : , in, ; It is the original accelerometer output vector. It is the actual accelerometer vector. It is an accelerometer with zero bias. , g is the acceleration due to gravity on Earth. norm is the accelerometer modulus. Substitute into the formulas for calculating pitch and roll: Pitch calculation: , in, , , These are the XYZ axis observations after acceleration correction. , These are the corrected and original values ​​of the pitch angle, respectively. Roll calculation: , , These are the corrected and original values ​​for the roll angle, respectively. = 1.000g, It is the corrected accelerometer module length.

[0024] The algorithm of this invention can not only effectively constrain the accelerometer module length, but more importantly, it can ensure that the roll and pitch installation angles remain strictly unchanged before and after the module length adjustment.

[0025] According to the constraint-based IMU accelerometer modulus length out-of-tolerance method described in this invention, the method is characterized in that... In step 1, by enabling the configuration parameter file, i.e. enabling the accelerometer modulus constraint function; setting the minimum modulus threshold and the maximum modulus threshold), the accelerometer modulus constraint function can be flexibly controlled by dynamic and static judgment, providing a more accurate initial value for subsequent user estimation of zero bias, and shortening the initialization time and convergence time.

[0026] According to the constraint-based IMU accelerometer modulus length out-of-tolerance method described in this invention, the method is characterized in that... In step 1, the configuration parameter file is set in the sensor configuration or data processing to ensure that the measured values ​​are within a safe or effective range.

[0027] The present invention provides a constraint-based method for suppressing IMU accelerometer modulus length deviation, characterized in that the present invention is applicable to situations where the accelerometer modulus length deviation is too large to proceed to the installation angle calibration, and is mainly used in the static coarse calibration of IMU installation angle.

[0028] According to this invention, determining the installation angle of an IMU generally consists of two steps: coarse calibration, also known as static calibration, which is typically performed during factory installation and provides an approximate solution; and precise calibration, also known as dynamic calibration, where the user needs to use external sensors to assist in dynamic long-distance motion testing to find a more accurate solution. The method of this invention primarily addresses the first step.

[0029] According to the present invention, the following features are provided: By providing an external configuration interface and utilizing dynamic and static judgments, the accelerometer's modulus constraint function can be flexibly controlled. Furthermore, this function can implement accelerometer modulus constraint under any static attitude. This provides more accurate initial values ​​for subsequent user estimation of bias, shortening initialization and convergence time.

[0030] According to the present invention, this invention is applicable to situations where the accelerometer modulus is found to be significantly out of tolerance, making installation angle calibration impossible. It is mainly used in static coarse calibration of IMU installation angles. According to the present invention, accelerometer modulus constraint functions can be implemented under any static attitude. Before and after calibration, only the magnitude of each vector changes, while its direction remains unchanged, and it is identical to the magnitude of gravitational acceleration 1g.

[0031] According to the present invention, an accelerometer modulus anomaly is monitored and analyzed in real time through a self-test mechanism after the IMU is powered on. Adaptive constraints are applied to the detected modulus anomalies, thereby effectively improving the stability and reliability of the device during long-term use. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of the actual IMU installation of the present invention.

[0033] Figure 2 This is a schematic diagram of the accelerometer module length constraint scheme of the present invention. Figure 3 This is a schematic diagram of an existing technology that only eliminates the Z-axis offset modulus.

[0034] Figure 4 This is a flowchart of an embodiment of the present invention.

[0035] Figures 5-6 The algorithm of this invention can be applied to any IMU tilt installation. The horizontal axis represents the installation angle under tilt conditions from 0 to 80 degrees, and the vertical axis represents the error between the roll angle calculated by the algorithm and the actual installation angle; the error is very small.

[0036] Figures 7-15 In both static and dynamic experiments, when the IMU is stationary, the accelerometer modulus within the modulus range of [6mg, 20mg] is constrained to within [-1.0mg, 1.0mg]. When the IMU is in motion, the original accelerometer measurement values ​​are output as a display graph.

[0037] In the diagram, X, Y, and Z represent the raw measurements from the accelerometer, norm is the raw modulus, and 1g is the modulus of Earth's gravitational acceleration. The arrows indicate the directions of the vectors.

[0038] After calibration using the algorithm of this invention, the normalization principle only changes the magnitude of the vector, not its direction. Since the direction attribute reflects the installation method and is a crucial attribute for users, it should not be modified. Therefore, this invention is primarily used for coarse calibration of the installation angle. Light-colored arrows represent the original acceleration measurement value and magnitude, while dark-colored arrows represent the accelerometer measurement value and magnitude after calibration. It can be seen that before and after calibration, only the magnitude of each vector changes, but its direction remains unchanged and is the same as the magnitude of gravitational acceleration 1g.

[0039] Figure 3 In this process, because the existing technology does not consider the horizontal axis but only the gravity axis and only adjusts the size of the Z-axis, the direction of the measurement vector changes after the measurement calibration of the entire IMU, resulting in distortion of the subsequent static calibration of the mounting angle. Detailed Implementation

[0040] Example

[0041] In any stationary state, an IMU with out-of-tolerance module length is placed on a three-axis turntable, and the turntable's three-axis attitude angles are known. The IMU corrected using this method has an estimated mounting angle that is almost identical to the actual turntable angle. However, the existing method only roughly matches the true value when all three axes have approximately horizontal small angles; it fails once any axis is tilted.

[0042] The following describes a constraint-based IMU accelerometer modulus out-of-tolerance suppression method for a batch of 10,000 IMU products. One IMU had a modulus out-of-tolerance of 10 mg, which directly affected the coarse calibration of the mounting angle, causing it to fail the calibration test. However, this product already met the usage requirements, and its subsequent processes are primarily for its application scenario, which has a high tolerance for modulus out-of-tolerance. Nevertheless, initial mounting angle calibration is still required. The IMU is powered on, and the modulus constraint flag is manually triggered, setting a minimum out-of-tolerance threshold of 6 mg and a maximum out-of-tolerance threshold of 20 mg. The mounting angle calibration is performed under static conditions, where dynamic and static detection is static. The modulus constraint normalization program is triggered to estimate the triaxial correction zero bias, and then the corrected data is output.

[0043] 1. Static attitude angle error analysis Calibration purpose: To verify whether the modulus zeroing algorithm affects the accuracy of Roll and Pitch calculated by the IMU, a comparative analysis was conducted on the Roll and Pitch calculated from the raw and corrected data of the accelerometer at different X-axis tilt angles (from 0.5° to 80°).

[0044] Calibration steps: The IMU is rotated around the Y-axis to the following angles in sequence, and held at each position for about 1 minute. After obtaining the data, the zero-biased modulus length of the added table is forcibly increased to more than 10mg, and a single-group engineering simulation is performed for verification.

[0045] Calibration results: Modulus length correction: Before correction, the modulus length of the original data showed a significant deviation, with the maximum error reaching approximately 10.7 mg. After correction, the modulus length of the corrected data became very stable, generally close to 1.0005 g, and the error decreased and stabilized at around 0.5 mg.

[0046] Attitude angle differences before and after correction: The table below details the acceleration values ​​along the X, Y, and Z axes, as well as the changes in Roll and Pitch angles calculated from these values ​​before and after correction. Through this modulus correction and simulation, the two core deviation parameters of the accelerometer were calculated: Roll deviation = -0.0765° Pitch deviation = 0.11645° Simply put, the magnitude of the attitude angle is not affected before or after the correction.

[0047]

[0048] 2. Experiment Details A detailed comparison of pitch and roll before and after each engineering project was completed.

[0049] 3. Dynamic and static experiments Experimental objective: To verify the effectiveness and feasibility of the algorithm by using dynamic and static judgments to constrain the accelerometer to zero bias under static conditions.

[0050] Experimental subjects: 9 IMU3200 devices with a static modulus greater than 3 mg, at a frequency of 100 Hz.

[0051] Experimental method: After each IMU is powered on, it is first held statically for a period of time, then the IMU is shaken to make it move, and then the movement is maintained. This process is repeated, and data is collected for 60 seconds.

[0052] Algorithm flow: see Figure 4 .

[0053] 4. Results Analysis Conclusion: When the IMU is stationary, the accelerometer modulus within the modulus range of [6mg, 20mg] is constrained to within the range of [-1.0mg, 1.0mg]. When the IMU is in motion, the original accelerometer measurement value is output.

[0054] Legend (see appendix) Figure 7-15 ): status_flag: (red) Dashed line 1, IMU fault code.

[0055] origin_residual_mag(mg): (orange) solid line 2, the original accelerometer magnitude difference, in mg.

[0056] constrain_residual_mag(mg): (blue) solid line 3, the difference in accelerometer magnitude after constraint, in mg.

[0057] 5. API Interface This interface function is primarily used to constrain the accelerometer module length when it is found to be out of tolerance. The relevant flags and thresholds are configured through the following API.

[0058] / *** @brief Configure Accelerometer module length constraint related parameters * acc_constrain_flag: Enable the accelerometer module length constraint flag, 1: enabled, 0: disabled * max_thresh: Maximum value of accelerometer module length deviation, in g, default value 0.02f (20mg) * min_thresh: Minimum value of accelerometer module length deviation, in g, default value 0.006f (6mg) * / void Put_AccBiasCfg(constunsignedchar acc_constrain_flag, constfloatmin_thresh, constfloatmax_thresh) / *** @brief: Get the zero-bias estimation status flags * @return: 0: Normal, 1: Accumulation modulus too large, 2: Correction * / unsigned int Get Accstate(void); illustrate: When the IMU's accelerometer magnitude is found to be too large, the `acc_constrain_flag` flag can be enabled to initiate accelerometer magnitude constraint functionality in the IMU's static state. The range of `max_thresh` and `min_thresh` for magnitude correction is determined by configuration. When the accelerometer magnitude difference is less than `min_thresh`, no constraint is applied; when the magnitude difference is greater than `max_thresh`, an accelerometer fault is reported; when the accelerometer magnitude difference is within the range of `min_thresh`, ... According to the present invention, the accelerometer modulus constraint function can be realized under any static attitude. Before and after the correction, each vector only changes in magnitude, while its direction remains unchanged, and it is the same as the magnitude of the gravitational acceleration 1g.

[0059] According to the present invention, an accelerometer modulus anomaly is monitored and analyzed in real time through a self-test mechanism after the IMU is powered on. Adaptive constraints are applied to the detected modulus anomalies, thereby effectively improving the stability and reliability of the device during long-term use.

Claims

1. A constraint-based method for suppressing IMU accelerometer modulus length deviation, characterized in that, Includes the following steps: Step 1. Start-up: Power on the IMU and output the raw data of the linear acceleration (acc_x, acc_y, acc_z) of the accelerometer along the XYZ axes and the angular velocity (gyro_x, gyro_y, gyro_z) of the gyroscope. Configure the parameter file to set the accelerometer module length constraint flag and the maximum and minimum out-of-tolerance thresholds for the accelerometer module length. Step 2. If the algorithm entry point detects that the accelerometer modulus constraint flag is invalid, the accelerometer directly sends the raw data from the accelerometer and gyroscope. If the accelerometer constraint flag is valid, the accelerometer module length constraint flag is triggered, and the accelerometer module length difference is calculated. The accelerometer modulus is equal to the square root of the sum of the squares of the accelerometer readings from the three axes. ， norm: modulus, acc_x, acc_y, and acc_z represent the acceleration measurements along the X, Y, and Z axes of the accelerometer, respectively. The actual modulus at rest is equal to g (the Earth's gravitational acceleration constant), and the difference between the two is the modulus difference. Calculate the current acceleration modulus : If it deviates from 1g, normalization is performed: ， ， Step 3. If the accelerometer modulus difference is greater than the maximum out-of-tolerance threshold, report a fault directly; If the accelerometer modulus is less than the minimum out-of-tolerance threshold, the accelerometer zero bias is directly set to zero. Output the original linear accelerations (acc_x, acc_y, acc_z) of the XYZ axes from the accelerometer. Step 4. When the calculated accelerometer modulus difference is between the minimum and maximum out-of-tolerance thresholds, proceed to IMU dynamic / static judgment: When the IMU is in a dynamic state, it is not processed and is ignored; When the IMU is in a static state, the accelerometer module length is constrained: Step 5: Normalize the accelerometer output, calculate the zero bias values ​​that need to be corrected for each of the three axes, and finally output the accelerometer measurement values ​​after correction (zero bias values).

2. The constraint-based IMU accelerometer modulus length out-of-tolerance suppression method according to claim 1, characterized in that, If the IMU module length exceeds the tolerance by 10mg and fails the calibration test, installation angle calibration is required. At this time, power on the IMU to trigger the module length constraint flag to start.

3. The constraint-based IMU accelerometer modulus length out-of-tolerance suppression method according to claim 2, characterized in that, The minimum exceedance threshold is set to 6 mg, and the maximum exceedance threshold is set to 20 mg. The threshold values ​​of 20mg and 6mg are common and within the range of experience, but can be freely configured according to the project indicators.

4. The constraint-based IMU accelerometer modulus length out-of-tolerance suppression method according to claim 2, characterized in that, The installation angle calibration is performed under static conditions during dynamic and static testing, triggering the module length constraint normalization program, estimating the three-axis correction zero bias, and outputting the corrected data.

5. The constraint-based IMU accelerometer modulus length out-of-tolerance suppression method according to claim 1, characterized in that, Step 5, the corrected data includes: Corrected accelerometer output vector , Acceleration-corrected XYZ axis observations , , , Corrected pitch angle values ​​compared to the original values , , Corrected value of roll angle compared to original value , , raw accelerometer output vector The actual accelerometer vector It is the zero bias of the accelerometer. .

6. The constraint-based IMU accelerometer modulus length out-of-tolerance suppression method according to claim 5, characterized in that, The algorithm and proof process of the constraint-based IMU accelerometer modulus length out-of-tolerance suppression method are as follows: accelerometer output vector after settlement correction : ， in, This is the original accelerometer output vector, in g, representing the Earth's gravitational acceleration. It is the actual accelerometer vector, in units of g. It is the accelerometer zero bias, measured in g. ， g is the acceleration due to gravity on Earth. norm is the accelerometer modulus. Substitute into the formulas for calculating pitch and roll: Pitch calculation: ， in, , , These are the XYZ axis observations after acceleration correction. , These are the corrected and original values ​​of the pitch angle, respectively. Roll calculation: , These are the corrected and original values ​​for the roll angle, respectively. = 1.000g， It is the corrected accelerometer module length.

7. The constraint-based IMU accelerometer modulus length out-of-tolerance suppression method according to claim 1, characterized in that, In step 1, by enabling the configuration parameter file (i.e. enabling: accelerometer modulus constraint function; minimum modulus threshold; maximum modulus threshold), the accelerometer modulus constraint function can be flexibly controlled by dynamic and static judgments, providing more accurate initial values ​​for subsequent user estimation of zero bias, and shortening the initialization time and convergence time.

8. The constraint-based IMU accelerometer modulus length out-of-tolerance suppression method according to claim 1, characterized in that, In step 1, the configuration parameter file is set in the sensor configuration or data processing to ensure that the measured values ​​are within a safe or effective range.

9. The constraint-based IMU accelerometer modulus length out-of-tolerance suppression method according to claim 1, characterized in that, This invention is applicable to situations where the accelerometer modulus is found to be out of tolerance and cannot proceed to the installation angle calibration. It is mainly used in static coarse calibration of IMU installation angle.

Images