Field calibration method for triaxial digital MEMS (Micro Electro Mechanical System) accelerometer

Abstract

The invention discloses a three-axis digital MEMS accelerometer field calibration method, which comprises the following steps: carrying out levelness calibration on a box-shaped object by using a bubble level, fixing an accelerometer on a specific surface of the box-shaped object, sequentially placing the box-shaped object at six different positions, recording output data of the accelerometer, and carrying out calibration on the box-shaped object according to the output data. The method comprises the following steps: calculating a theoretical acceleration value and an actual output value, obtaining an average output value of each position through mean filtering processing, calculating a zero deviation and a scale factor of the accelerometer according to the theoretical acceleration value and the actual output value, establishing an output error mathematical model, and finally, verifying a calibration result to ensure the accuracy of field calibration.

Description

Technical Field

[0001] The invention relates to the technical field of accelerometer calibration, and in particular to an on-site calibration method for a three-axis digital MEMS accelerometer. Background Art

[0002] In the fields of modern agricultural machinery, drones, and autonomous driving, accurate measurement of attitude angles is the key to achieving precise navigation and automatic control. As an inertial measurement unit, three-axis digital MEMS accelerometers are widely used in monitoring the attitude of objects to obtain attitude information such as the yaw angle, pitch angle, and roll angle of objects. In order to obtain high-precision measurement results, accelerometers need to be strictly calibrated before use to compensate for various errors and improve the accuracy of attitude measurement.

[0003] At present, the calibration of inertial measurement units usually relies on high-precision three-axis turntables. However, high-precision three-axis turntables are expensive, bulky, and complicated to operate. They are mainly used in laboratories. Ordinary users and general laboratories often lack the conditions for such precision instruments. Therefore, it is difficult to accurately calibrate three-axis accelerometers in practical applications. The main errors of accelerometers include zero deviation, scale factor error, installation error, and inter-axis non-orthogonality error. Among them, zero deviation and scale factor error are the main factors affecting the accuracy of accelerometers. Traditional calibration methods require cumbersome operations and expensive equipment, and cannot meet the needs of rapid on-site calibration. Based on this, a three-axis digital MEMS accelerometer on-site calibration method is needed. Summary of the invention

[0004] To achieve the above object, the present invention provides the following solution: a three-axis digital MEMS accelerometer on-site calibration method, comprising the following steps:

[0005] A six-position calibration platform was constructed using a bubble level and a rectangular box-shaped object. The three-axis digital MEMS accelerometer module was fixed on a certain surface of the rectangular box-shaped object so that the three sensitive axis directions of the accelerometer were parallel to the length, width, and height of the rectangular box, respectively. The bubble level was used to check the placement direction of the desktop and the box-shaped object so that the box-shaped object remained horizontal and stable during the calibration process.

[0006] Place the box-shaped object on a horizontal table in the following six different positions, record the output data of the accelerometer at each position, and process the output data through mean filtering to obtain the average output value of each position sensitive axis:

[0007] Position 1: x-axis upward, y-axis forward, z-axis left, the theoretical acceleration value is (256LSB, 0, 0);

[0008] Position 2: x-axis downward, y-axis forward, z-axis right, the theoretical acceleration value is (-256LSB, 0, 0);

[0009] Position 3: x-axis points to the left, y-axis points upward, and z-axis points forward. The theoretical acceleration value is (0, 256LSB, 0).

[0010] Position 4: x-axis points to the right, y-axis points downward, and z-axis points forward. The theoretical acceleration value is (0, -256LSB, 0).

[0011] Position 5: x-axis is forward, y-axis is left, and z-axis is upward. The theoretical acceleration value is (0, 0, 256LSB);

[0012] Position 6: x-axis forward, y-axis right, z-axis downward, the theoretical acceleration value is (0, 0, -256LSB);

[0013] Place the box-shaped object at position 1 on a relatively level table that has been checked by a bubble level, keep it still for 2 to 3 minutes, and record the x-axis output value X after mean filtering. +1g Then keep the accelerometer module still at position 2 and record the output value X of the x-axis. -1g ;

[0014] Calculate the zero deviation B of the x-axis x :

[0015]

[0016] Similarly, keep the accelerometer module stationary at position 3 and position 4, and record the output value Y of the y-axis respectively. +1g and Y -1g , calculate the zero deviation B of the y-axis y :

[0017]

[0018] Keep the accelerometer module stationary at position 5 and position 6, and record the output value Z of the z-axis respectively. +1g and Z -1g , calculate the zero position deviation B of the z-axis z :

[0019]

[0020] For the x-axis, calculate the x-axis scale factor S based on the output values ​​of position 1 and position 2 x :

[0021]

[0022] For the y-axis, calculate the scale factor S of the y-axis based on the output values ​​of position 3 and position 4 y :

[0023]

[0024] For the z-axis, calculate the scale factor S of the z-axis based on the output values ​​of position 5 and position 6 z :

[0025]

[0026] According to the output error mathematical model of the accelerometer, the relationship between the output value of the sensor and the true acceleration value is expressed as:

[0027]

[0028] Among them, M x 、M y 、M z is the sensor measurement value, D x , D y , D z is the true value of acceleration, B x , B y , B z is zero bias, S x , S y , S z is the scale factor;

[0029] Use the calibration model to compensate the measured value output by the accelerometer in real time, eliminate the zero bias and proportional factor errors, and obtain the corrected true acceleration value:

[0030]

[0031] The accuracy of the on-site calibration method was verified by comparing the proportional factor and zero bias value obtained by on-site calibration with the precise calibration results of the three-axis turntable. The deviation of each axis was recorded without calibration. It was confirmed that the output error after on-site calibration was reduced to an acceptable range, and the proportional factor deviation was within ±10 -3 The zero offset value error is less than 1LSB.

[0032] Further preferably, in the six-position calibration process, each surface of the rectangular box-shaped object used is manufactured by high-precision machining, and its flatness and verticality meet the accuracy requirements of accelerometer calibration. During the placement of the rectangular box-shaped object, the horizontality of the box-shaped object is calibrated by a bubble level to ensure that the horizontality error of each placement does not exceed 0.1 degree.

[0033] Further preferably, the theoretical output value of the accelerometer at each position is calculated based on the gravitational acceleration and the sensitivity of the accelerometer, where the sensitivity is 256LSB / g. When the sensitive axis of the accelerometer is parallel to the direction of gravity, the theoretical output value is ±256LSB. To ensure the accuracy of calibration, the theoretical values ​​of the six positions are: the x-axis output value of position 1 and position 2 is ±256LSB; the y-axis output value of position 3 and position 4 is ±256LSB; the z-axis output value of position 5 and position 6 is ±256LSB. According to the deviation between these theoretical values ​​and the actual output value of the accelerometer, the zero deviation and the proportional factor are calculated.

[0034] Further preferably, the proportional factor S x , S y , S z and zero deviation B x , B y , B z In the calculation process, the actual output values ​​of the accelerometer at six positions are used, combined with the least squares method to optimize the calibration coefficients, which represents the output data matrix of the six positions as M 6×4 , the theoretical acceleration matrix is ​​expressed as D 6×3 , by establishing equation D 6×3 =M 6×4 ·K 4×3 , using the least squares formula The calibration coefficient matrix to be calculated is used to determine the scale factor and zero deviation of the accelerometer.

[0035] Further preferably, the installation error and inter-axis non-orthogonal error established by the calibration model further express the mathematical model of the sensor output error as: M=S·D+B+E, wherein E is the error term under the combined influence of the installation error and the non-orthogonal error. The calibration model is based on the influence of the sensor's zero position deviation, proportional factor and installation error on the calibration result. During the calibration process, the installation error and inter-axis non-orthogonal error are uniformly classified by reasonably simplifying the error model.

[0036] Further preferably, after the on-site calibration is completed, the accuracy of the on-site calibration method is verified by comparing it with the precise calibration result of the three-axis turntable. The comparison process includes: converting the proportional factor S obtained by the on-site calibration into x , S y , S z and zero deviation B x , B y , B zCompare with the reference value obtained by the three-axis turntable calibration, and calculate the deviations ΔS and ΔB, where ΔS = |Ssite-Sturntable|, ΔB = |Bsite-Bturntable|. The proportional factor deviation is required to be within the range of ±10^-3, and the zero bias error shall not exceed 1LSB.

[0037] Further preferably, the field calibration method includes periodically reviewing the calibration results to adapt to environmental changes and sensor performance drift. During the long-term use of the sensor, the calibration process is repeated regularly, the scale factor and zero deviation of each calibration are recorded, and the calibration stability is evaluated using the following formula:

[0038]

[0039] Where C is the stability evaluation value, n is the number of measurements within the review period, and D i，测量 and D i，理论 are the actual measured value and the theoretical value respectively.

[0040] Further preferably, the calibration method also includes repeating the calibration test under different environmental conditions to verify the environmental adaptability of the calibration method. After completing the initial calibration, the accelerometer is placed under different environmental temperature and humidity conditions and repeated the test at least three times. After each test, the calculated proportional factor and zero deviation are compared with the results under the initial environmental conditions to evaluate the impact of environmental factors on the calibration results.

[0041] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0042] The three-axis digital MEMS accelerometer on-site calibration method proposed in the present invention is based on a common bubble level and a box-shaped object with a rectangular structure, and adopts a six-position calibration method, which has the advantages of being convenient, fast, simple, easy to operate, accurate and effective;

[0043] The present invention can complete calibration with simple tools, the operation process is intuitive, no complicated equipment configuration is required, the calibration time is significantly shortened, and it is particularly suitable for rapid implementation on site. By utilizing the geometric characteristics of the rectangular box-shaped object, the operator only needs to place the box-shaped object in sequence and record the output data, thereby simplifying the cumbersome steps in the traditional calibration method and reducing the difficulty of operation. At the same time, the method can effectively compensate for the zero deviation and proportional factor error of the accelerometer. In practical applications, the output error of the accelerometer calibrated on site can be controlled within an acceptable range, thereby ensuring the reliability of attitude measurement. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0045] Figure 1 It is a flow chart of the steps of the present invention. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0048] Example

[0049] like Figure 1 As shown, the on-site calibration method of the three-axis digital MEMS accelerometer in this embodiment includes:

[0050] The main errors of the three-axis digital MEMS accelerometer include zero deviation, scale factor error, installation error and inter-axis non-orthogonality error. Among them, zero error and scale factor error are the main factors affecting the accuracy of the accelerometer, and they change with time and environment. Installation error and inter-axis non-orthogonality error will not change significantly over a long period of time, and the impact can be reduced by improving the installation process and manufacturing process. Based on the above characteristics, the mathematical model of the sensor output error is expressed as:

[0051]

[0052] Among them, M x 、M y 、M z is the sensor measurement value, D x , D y , D z is the true value of acceleration, B x , B y , B z is zero bias, S x , S y , S z is the scaling factor, k ij(i=x,y,z;j=x,y,z;i≠j) is the sensor installation error coefficient;

[0053] The calibration of the acceleration sensor is to determine the zero bias, proportional factor, and installation error coefficient in the mathematical model. There are 12 unknown parameters in the model. With the improvement of installation and manufacturing processes, the impact of installation errors is very small, which is negligible compared to the errors caused by proportional factor and zero bias. Therefore, simple on-site calibration can only calibrate the proportional factor and zero bias, and the error mathematical model of the sensor is simplified to:

[0054]

[0055] The acceleration sensor used in the present invention is a three-axis digital accelerometer ADXL345, which is mainly used in miniature inertial devices for consumer electronics, has a sensitivity of 3.9mg / LSB, the least significant bit, and a data output frequency of up to 3200Hz. The sensor uses an integrated memory management system with an optional 32-level first-out FIFO buffer to store data, which greatly reduces the host processor load and system power consumption;

[0056] The calibration of the accelerometer is usually carried out in a gravity field using a multi-position tumble calibration method based on gravity. Commonly used calibrations are 6-position, 10-position, 12-position and 24-position. The unknowns in the sensor calibration coefficient matrix are solved according to the theoretical acceleration of the determined position and the measured value of the sensor. In order to simplify the difficulty of on-site calibration, the present invention adopts a six-position calibration method based on a rectangular parallelepiped and sets the position direction of each sensitive axis of the sensor as shown in Table 1. When the ADXL345 is set to full resolution mode, its sensitivity is 256LSB / g. When the sensitive axis is parallel to the direction of gravity, its theoretical value should output 256LSB or -256LSB.

[0057] Table 1 Accelerometer axis orientation and acceleration output

[0058]

[0059] The accelerometer zero deviation calibration needs to be performed in a static state. The sensor module is placed on a certain surface of a rectangular box-shaped object, and the three sensitive axes of the sensor are parallel to the length, width and height of the rectangular box. The box-shaped object is placed on a relatively horizontal table that has been checked by a bubble level according to position 1. Keep it static for 2 to 3 minutes, and record the x-axis output after mean filtering, which is recorded as X +1g ; Then place it in position 2 to get X -1g , then the zero position deviation of the x-axis is B x Depend on

[0060]

[0061] Calculate it;

[0062] Similarly, keep the accelerometer module stationary at position 3 and position 4, and record the output value Y of the y-axis respectively. +1g and Y -1g , calculate the zero deviation B of the y-axis y :

[0063]

[0064] Keep the accelerometer module stationary at position 5 and position 6, and record the output value Z of the z-axis respectively. +1g and Z -1g , calculate the zero position deviation B of the z-axis z :

[0065]

[0066] The zero bias and proportional factor of each sensitive axis of the accelerometer are obtained through 6 position tests (Table 2). In order to reduce the processing load of the single chip microcomputer, the data is rounded, where M is the actual output of the sensitive axis of the accelerometer, and the calibration process is recorded as on-site calibration;

[0067] Table 2 Simple on-site calibration of accelerometer zero bias and scale factor:

[0068]

[0069] In order to verify the accuracy of on-site calibration, the sensor module was placed on the inner frame of the SGT320E three-axis multifunctional turntable developed by the Beijing Institute of Aeronautical Precision Mechanics. The MCU used Atmega16, the serial port baud rate was set to 110592, and the turntable was set to work in "position mode". The "bit rate" of the turntable used the default value of 10.000 (°) / s, and the "bit acceleration" used the default value of 10.000 (°) / s 2 ;

[0070] At least 20 groups of raw data of xyz and three sensitive axes were collected at 6 positions, and the average value was taken, and 2 significant digits were retained after the decimal point. The raw data at 6 positions were shown in Table 3.

[0071] In order to solve the 12 unknown parameters in the sensor mathematical model, the sensor model 2 is transformed equivalently [D x D y D z ]=[M x M y M z 1]

[0072]

[0073] It is recorded as: D = M·K, where: D is the known gravity acceleration at 6 positions; M is the original output of the accelerometer at 6 positions; K is the calibration coefficient matrix to be determined, a total of 12 unknown parameters;

[0074] Table 3 Laboratory turntable calibration accelerometer 6 position output

[0075]

[0076]

[0077] There is D under position 1 (1) =

[25600] ,M (1) =[277.583.978.791]

[0078] Similarly, from the theoretical output value of the accelerometer in Table 1 and the actual measured value in Table 3, we can get D (2) , D (3) , ..., D (4) and

[0079] Expressed in matrix form:

[0080] D 6×3 =[D (1) , D (2) , D (3) , D (5) , D (6) , D (7) ] T

[0081]

[0082] Formula Expressed as: D 6×3 =M 6×4 ·K 4×3 , we get 18 equations, and we need to solve 12 unknowns. Therefore, we can get the optimal solution by using the least squares method. The expression of the coefficient matrix solution is as follows:

[0083]

[0084] The calibration coefficient matrix is ​​obtained in MATLAB:

[0085]

[0086] The output value of the accelerometer after calibration and compensation can be obtained from the following calibration and compensation solution matrix:

[0087]

[0088] The scale factor S obtained by calibrating the turntable i (i = x, y, z) and zero bias value B i (i = x, y, z) is compared with the data in Table 2 as shown in Table 4. It is not difficult to find that the accelerometer scale factor and zero bias value obtained by the simple on-site calibration I and the turntable accurate calibration are very close, and the scale factor deviation is within ±10 -3 The maximum deviation of the zero offset value of the x-axis is 1.018LSB, which is about 0.00397g, and the zero offset value deviations of the y-axis and z-axis are both less than 0.5LSB;

[0089] Table 4 Comparison of turntable calibration and two on-site calibration results:

[0090]

[0091] The purpose of simple on-site calibration I is to facilitate the fast processing of the single-chip microcomputer. Therefore, the data in Table 2 are rounded off. If two valid digits are retained after the decimal point for simple calibration (recorded as simple on-site calibration II), the scale factors of the three axes calculated using the data in Table 3 are closer to the scale factors after the turntable calibration (see Table 4). The scale factor deviation can be reduced by another order of magnitude within ±10 -4 The zero offset error of the x-axis is also reduced to 0.568LSB, i.e. 0.00222g. Although the zero offset deviations of the y-axis and z-axis increase slightly, they are both less than 0.1LSB.

[0092] Indirect verification method is used in gravity field space. The principle is as follows: when the accelerometer module is tilted, the gravity acceleration components M on its three orthogonal sensitive axes i (i=x,y,z) calculate the angles of each sensitive axis relative to the horizontal plane, define the pitch angle ρ as the angle between the sensitive axis and the horizontal plane, define the roll angle γ as the angle between the sensitive axis y and the horizontal plane, then we have

[0093]

[0094] The accelerometer module is placed on the inner frame of the three-axis turntable. The inner frame of the turntable is rotated to set the accelerometer roll angle to 30.000° and keep it unchanged. Then the middle frame of the turntable is rotated according to the pitch angle shown in Table 5. The pitch angle is calculated from the original data of the accelerometer and the data compensated by the two calibration methods. The results are shown in Table 5:

[0095] Table 5 Accelerometer calibration and angle error

[0096]

[0097] It can be seen from Table 5 that the maximum angle deviation before calibration and compensation is 3.591°, the absolute error after simple on-site calibration I is reduced to 0.051°, and the angle error after turntable calibration compensation is reduced to 0.048°. It can be seen that the output error without calibration compensation is large. Both the simple calibration method and the turntable calibration can reduce the output absolute error by about 2 orders of magnitude. At the same time, it can be seen that the accuracy of the simple on-site calibration I method is roughly the same as that of the turntable calibration method. Even in some positions, the accuracy after simple on-site calibration compensation is higher, which proves that the simple on-site calibration method that ignores the installation error is simple, feasible, correct and effective.

[0098] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0099] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A three-axis digital MEMS accelerometer field calibration method, characterized in that: The following steps are involved: A six-position calibration platform was constructed using a bubble level and a rectangular box-shaped object. The three-axis digital MEMS accelerometer module was fixed on a certain surface of the rectangular box-shaped object so that the three sensitive axis directions of the accelerometer were parallel to the length, width, and height of the rectangular box, respectively. The bubble level was used to check the placement direction of the desktop and the box-shaped object so that the box-shaped object remained horizontal and stable during the calibration process. Place the box-shaped object on a horizontal table in the following six different positions, record the output data of the accelerometer at each position, and process the output data through mean filtering to obtain the average output value of each position sensitive axis: Position 1: x-axis upward, y-axis forward, z-axis left, the theoretical acceleration value is (256LSB, 0, 0); Position 2: x-axis downward, y-axis forward, z-axis right, the theoretical acceleration value is (-256LSB, 0, 0); Position 3: x-axis points to the left, y-axis points upward, and z-axis points forward. The theoretical acceleration value is (0, 256LSB, 0). Position 4: x-axis points to the right, y-axis points downward, and z-axis points forward. The theoretical acceleration value is (0, -256LSB, 0). Position 5: x-axis is forward, y-axis is left, and z-axis is upward. The theoretical acceleration value is (0, 0, 256LSB); Position 6: x-axis forward, y-axis right, z-axis downward, the theoretical acceleration value is (0, 0, -256LSB); Place the box-shaped object at position 1 on a relatively level table that has been checked by a bubble level, keep it still for 2 to 3 minutes, and record the x-axis output value X after mean filtering. +1g Then keep the accelerometer module still at position 2 and record the output value X of the x-axis. -1g ; Calculate the zero deviation B of the x-axis x : Similarly, keep the accelerometer module stationary at position 3 and position 4, and record the output value Y of the y-axis respectively. +1g and Y -1g , calculate the zero deviation B of the y-axis y : Keep the accelerometer module stationary at position 5 and position 6, and record the output value Z of the z-axis respectively. +1g and Z -1g , calculate the zero position deviation B of the z-axis z : For the x-axis, calculate the x-axis scale factor S based on the output values ​​of position 1 and position 2 x : For the y-axis, calculate the scale factor S of the y-axis based on the output values ​​of position 3 and position 4 y : For the z-axis, calculate the scale factor S of the z-axis based on the output values ​​of position 5 and position 6 z : According to the output error mathematical model of the accelerometer, the relationship between the output value of the sensor and the true acceleration value is expressed as: Among them, M x 、M y 、M z is the sensor measurement value, D x , D y , D z is the true value of acceleration, B x , B y , B z is zero bias, S x , S y , S z is the scale factor; Use the calibration model to compensate the measured value output by the accelerometer in real time, eliminate the zero bias and proportional factor errors, and obtain the corrected true acceleration value: The accuracy of the on-site calibration method is verified by comparing the proportional factor and zero offset value obtained by on-site calibration with the precise calibration results of the three-axis turntable. The deviation of each axis is recorded without calibration. It is confirmed that the output error is reduced to an acceptable range after on-site calibration. The proportional factor deviation is around ±10^-3, and the zero offset value error is less than 1LSB.

2. The on-site calibration method for a three-axis digital MEMS accelerometer according to claim 1, characterized in that: During the six-position calibration process, each surface of the rectangular box-shaped object used is manufactured by high-precision machining, and its flatness and verticality meet the accuracy requirements of accelerometer calibration. During the placement of the rectangular box-shaped object, the horizontality of the box-shaped object is calibrated by a bubble level to ensure that the horizontality error of each placement does not exceed 0.1 degrees.

3. The on-site calibration method for a three-axis digital MEMS accelerometer according to claim 1, characterized in that: The theoretical output value of the accelerometer at each position is calculated based on the gravitational acceleration and the sensitivity of the accelerometer, wherein the sensitivity is 256LSB / g. When the sensitive axis of the accelerometer is parallel to the direction of gravity, the theoretical output value is ±256LSB. To ensure the accuracy of calibration, the theoretical values ​​of the six positions are: the x-axis output value of position 1 and position 2 is ±256LSB; the y-axis output value of position 3 and position 4 is ±256LSB; the z-axis output value of position 5 and position 6 is ±256LSB. Based on the deviation between these theoretical values ​​and the actual output value of the accelerometer, the zero deviation and the proportional factor are calculated.

4. The on-site calibration method for a three-axis digital MEMS accelerometer according to claim 1, characterized in that: The scaling factor S x , S y , S z and zero deviation B x , B y , B z In the calculation process, the actual output values ​​of the accelerometer at six positions are used, combined with the least squares method to optimize the calibration coefficients, which represents the output data matrix of the six positions as M 6×4 , the theoretical acceleration matrix is ​​expressed as D 6×3 , by establishing equation D 6×3 =M 6×4 ·K 4×3 , using the least squares formula The calibration coefficient matrix to be calculated is used to determine the scale factor and zero deviation of the accelerometer.

5. The on-site calibration method for a three-axis digital MEMS accelerometer according to claim 1, characterized in that: The installation error and inter-axis non-orthogonal error established by the calibration model further express the mathematical model of the sensor output error as: M=S·D+B+E, wherein E is the error term under the combined influence of the installation error and the non-orthogonal error. The calibration model is based on the zero position deviation, proportional factor and installation error of the sensor on the calibration result. During the calibration process, the installation error and inter-axis non-orthogonal error are uniformly classified by reasonably simplifying the error model.

6. The on-site calibration method for a three-axis digital MEMS accelerometer according to claim 1, characterized in that: After the on-site calibration is completed, the accuracy of the on-site calibration method is verified by comparing it with the precise calibration result of the three-axis turntable. The comparison process includes: x , S y , S z and zero deviation B x , B y , B z Compare with the reference value obtained by the three-axis turntable calibration, and calculate the deviation ΔS and ΔB, where ΔS = |S site - S turntable |, ΔB = |B site - B turntable |, and the proportional factor deviation is required to be within ±10 -3 Within the range, the zero bias error does not exceed 1LSB.

7. The on-site calibration method for a three-axis digital MEMS accelerometer according to claim 1, characterized in that: The field calibration method includes periodic review of the calibration results to adapt to environmental changes and sensor performance drift. During the long-term use of the sensor, the calibration process is repeated regularly, the scale factor and zero deviation of each calibration are recorded, and the calibration stability is evaluated using the following formula: Where C is the stability evaluation value, n is the number of measurements within the review period, and D i，测量 and D i，理论 are the actual measured value and the theoretical value respectively.

8. The on-site calibration method for a three-axis digital MEMS accelerometer according to claim 1, characterized in that: The calibration method also includes repeating the calibration test under different environmental conditions to verify the environmental adaptability of the calibration method. After completing the initial calibration, the accelerometer is placed under different environmental temperature and humidity conditions and repeated testing is performed at least three times. After each test, the calculated proportional factor and zero deviation are compared with the results under the initial environmental conditions to evaluate the impact of environmental factors on the calibration results.

Images