A method and system for single-axis jump correction of a MEMS sensor

Abstract

This invention relates to the field of MEMS sensors and discloses a method and system for correcting single-axis abrupt changes in MEMS sensors, including the following steps: S1, obtaining the local gravitational acceleration reference value g at the deployment location of the target MEMS sensor; In this invention, through a dual judgment mechanism of modulus deviation and inter-axis correlation coefficient, it is possible to accurately distinguish between interference abrupt changes and real motion. Modulus deviation is used to initially screen abnormal data, while inter-axis correlation coefficient further verifies the authenticity of the abrupt change, effectively avoiding the problem of incorrect correction caused by relying on only a single indicator in the prior art. Furthermore, based on the physical law of conservation of the three-axis acceleration vector sum under static measurement, the abrupt axis is corrected using stable data of the other two axes, and an inter-axis coupling correction factor is introduced to compensate for manufacturing process deviations. The corrected data has high accuracy, and the total modulus deviation can be stably controlled within a preset threshold range, significantly improving the data accuracy of MEMS sensors in outdoor static measurement scenarios.

Description

Technical Field

[0001] This invention relates to the field of MEMS sensors, and in particular to a method and system for correcting uniaxial abrupt changes in MEMS sensors. Background Technology

[0002] MEMS sensors are widely used in outdoor static measurements (such as geological monitoring, building settlement observation, environmental monitoring, etc.). Their core relies on the stability and accuracy of triaxial data to reflect the static physical state of the measured object. However, in practical applications, sensors are easily affected by external environmental interference such as temperature and humidity fluctuations, air pressure changes, or defects in their own manufacturing process such as inter-axis coupling deviations, resulting in sudden and drastic changes in single-axis data, while the data of the other two axes remain basically stable.

[0003] Existing uniaxial mutation correction techniques have significant drawbacks:

[0004] The core physical laws of static measurement are not fully utilized. Data smoothing is done only through filtering algorithms (such as moving average), which cannot distinguish between axis data changes caused by real motion and single-axis abrupt changes caused by interference, which can easily lead to incorrect corrections.

[0005] The method ignores the three-axis physical orthogonality of MEMS sensors. The real motion of a single axis will inevitably cause related changes in other axes through rigid body kinematic coupling. However, the existing method does not verify the authenticity of the abrupt change through inter-axis correlation. Thirdly, it does not consider the law of conservation of the three-axis acceleration vector sum under static conditions. It judges the abrupt change only based on the fluctuation amplitude of single-axis data, which has weak anti-interference ability and the correction accuracy is difficult to meet the high-precision requirements of outdoor static measurement. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method and system for correcting single-axis abrupt changes in MEMS sensors, thus solving the above problems.

[0007] To achieve the above objectives, the present invention provides a method for correcting single-axis abrupt changes in a MEMS sensor, comprising the following steps:

[0008] S1. Obtain the local gravitational acceleration reference value g at the deployment location of the target MEMS sensor;

[0009] S2. Real-time acquisition of triaxial acceleration data sequence of the target MEMS sensor, and calculation of the total modulus and modulus deviation of the triaxial acceleration data;

[0010] S3. Determine whether the modulus deviation exceeds a preset deviation threshold. If it does, preliminarily identify a possible uniaxial abrupt change.

[0011] S4. Locate the abrupt change axis and calculate the correlation coefficient between the abrupt change axis and the acceleration data of the other two axes;

[0012] S5. Based on the correlation coefficient, determine that the single-axis mutation is an independent interference mutation. Based on the physical laws of static measurement, use the stable data of the other two axes to correct the mutation axis data.

[0013] S6. Output the corrected triaxial acceleration data to complete the correction of single-axis abrupt changes.

[0014] Preferably, the method for obtaining the local gravitational acceleration reference value g at the deployment location of the target MEMS sensor includes:

[0015] S11. Collect the latitude of the target MEMS sensor deployment location. and altitude ;

[0016] S12. Calculate the local gravitational acceleration using a geographical model of gravitational acceleration. The calculation formula is as follows:

[0017]

[0018] in, Latitude This refers to altitude.

[0019] Preferably, the method for real-time acquisition of triaxial acceleration data sequences from the target MEMS sensor and calculation of the total modulus and modulus deviation of the triaxial acceleration data includes:

[0020] S21. Acquire real-time triaxial acceleration data of the MEMS sensor at a preset sampling period T. Forming a time series , , ,in, , This represents the number of sampling points;

[0021] S22. Calculate the total triaxial acceleration modulus at each time point ti. The specific counting method is as follows:

[0022]

[0023] in, , , These are the time series that were formed;

[0024] S23. Calculate the overall modulus deviation. That is, the absolute value of the difference between the current modulus and the reference gravitational acceleration.

[0025] Preferably, the method for determining whether the modulus deviation exceeds a preset deviation threshold, and if it does, preliminarily identifying a suspected uniaxial abrupt change, includes:

[0026] Preset deviation threshold The threshold is determined based on the static measurement accuracy of the MEMS sensor, specifically between 0.05g and 0.1g.

[0027] ,like > And only the change in data on a single axis compared to the previous time point. > The changes in the other two axes are all less than If so, it is preliminarily determined that the axis is a mutation axis, and there is a suspicion of uniaxial mutation.

[0028] Preferably, the method for locating the abrupt change axis and calculating the correlation coefficient between the abrupt change axis and the acceleration data of the other two axes includes:

[0029] S41. Let the axis of the initial mutation be the X-axis, and the other two axes be the Y-axis and Z-axis.

[0030] S42. Extract the data within the mutation time window and calculate the Pearson correlation coefficients between the mutation axis X and Y axes, and between X and Z axes, respectively. , The specific formula is as follows:

[0031] ;

[0032] in, The covariance of the X-axis and Y-axis data. The covariance of the X-axis and Z-axis data. , , The variances of the X, Y, and Z axis data are respectively used. The average of the Pearson correlation coefficients between the two axes is taken as the indicator of inter-axis correlation. .

[0033] Preferably, the method for determining whether the uniaxial mutation is an independent interference mutation based on the correlation coefficient includes:

[0034] S51, Preset correlation coefficient threshold ;

[0035] S52, if < Therefore, the uniaxial mutation is ultimately determined to be an independent interference mutation and requires correction; if ≥ If the change is due to actual motion, it is determined to be a change in axis data and no correction is needed.

[0036] Preferably, the method for correcting the abrupt change axis data using stable data from the other two axes based on the physical laws of static measurement includes:

[0037] S5011. Core physical law based on static measurement: As a rigid body, the sum of the three-axis acceleration vectors of a MEMS sensor under static conditions is equal to the local gravitational acceleration. ,Right now ,in, This is the data after X-axis correction. , For the other two axes, stabilize the data;

[0038] S5012, Introducing an inter-axis coupling correction factor Its value ranges from 0.95 to 1.05, so the formula for correcting the abrupt change axis is:

[0039]

[0040] in, The sign of the X-axis data at the moment before the mutation, if the corrected If the value exceeds the sensor's reasonable measurement range, a linear interpolation method is used to combine the stable data from the previous three time points with the correction formula results to smoothly transition to the correction value and avoid data jumps.

[0041] Preferably, the method for outputting the corrected triaxial acceleration data includes:

[0042] S61. Modify the mutation axis data Replace the original mutation data ,and , The corrected triaxial acceleration data are composed of ( ,and , );

[0043] S62. Verify the corrected total modulus deviation. ,like ≤ Then output the data; if it still exceeds the threshold, repeat the correction steps until the requirements are met.

[0044] A single-axis mutation correction system for a MEMS sensor, comprising:

[0045] The reference value acquisition module is used to acquire the local gravitational acceleration reference value at the deployment location of the target MEMS sensor. ;

[0046] The data acquisition and calculation module is used to acquire the triaxial acceleration data sequence of the target MEMS sensor in real time and calculate the total modulus and modulus deviation of the triaxial acceleration data.

[0047] The preliminary identification module is used to determine whether the modulus deviation exceeds a preset deviation threshold. If it does, it is preliminarily identified as having a suspected uniaxial abrupt change.

[0048] A precise determination module is used to locate the abrupt change axis and calculate the correlation coefficient between the abrupt change axis and the acceleration data of the other two axes. Based on the correlation coefficient, the module determines that the single-axis abrupt change is an independent interference abrupt change.

[0049] The data correction module is used to correct the abrupt change axis data based on the physical laws of static measurement and using stable data from the other two axes.

[0050] The data output module is used to output the corrected triaxial acceleration data to correct for single-axis abrupt changes.

[0051] Preferably, the benchmark value acquisition module, data acquisition and calculation module, preliminary identification module, precise judgment module, data correction module and data output module are all implemented by computer programs and run on the same hardware platform.

[0052] Beneficial effects

[0053] This invention provides a method and system for correcting single-axis abrupt changes in MEMS sensors. Compared with existing technologies, it has the following advantages:

[0054] In this invention, a dual judgment mechanism of modulus deviation and inter-axis correlation coefficient can accurately distinguish between interference mutations and real motion. Modulus deviation is used to initially screen abnormal data, while inter-axis correlation coefficient further verifies the authenticity of mutations. This effectively avoids the problem of miscorrection caused by relying on only a single indicator in the prior art. Furthermore, based on the physical law of conservation of triaxial acceleration vector sum under static measurement, stable data of the other two axes are used to correct the mutation axis, and an inter-axis coupling correction factor is introduced to compensate for manufacturing process deviations. The corrected data has high accuracy, and the total modulus deviation can be stably controlled within the preset threshold range, significantly improving the data accuracy of MEMS sensors in outdoor static measurement scenarios.

[0055] 2. In this invention, after locating the initially identified mutation axis, a time window containing stable data before and after the mutation is extracted. The correlation coefficient between the mutation axis and the data of the other two axes is calculated, and the type of mutation is determined by the magnitude of the correlation coefficient. If the correlation coefficient is close to 0 or negative, it indicates that the mutation axis is not significantly related to the other two axes, and it can be ultimately determined to be an independent mutation caused by interference, which requires correction. If the correlation coefficient reaches or exceeds a preset threshold, it is determined to be a normal data change caused by real motion, which does not require correction, thereby ensuring the real-time performance and accuracy of the monitoring data. Attached Figure Description

[0056] Figure 1 This is a flowchart of a method and system for correcting single-axis mutations in MEMS sensors proposed in this invention.

[0057] Figure 2 This is a schematic diagram of the time window for a single-axis mutation correction method and system for MEMS sensors proposed in this invention;

[0058] Figure 3 This is a curve showing the comparison of the module length before and after correction of a single-axis abrupt change correction method and system for MEMS sensors proposed in this invention. Detailed Implementation

[0059] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0060] Please see Figures 1-3 The present invention provides the following technical solutions, specifically including the following embodiments:

[0061] Example 1:

[0062] A method for correcting uniaxial abrupt changes in a MEMS sensor includes the following steps:

[0063] S1. Obtain the local gravitational acceleration reference value g at the deployment location of the target MEMS sensor. The method for obtaining the local gravitational acceleration reference value g at the deployment location of the target MEMS sensor includes:

[0064] S11. Collect the latitude of the target MEMS sensor deployment location. and altitude ;

[0065] S12. Calculate the local gravitational acceleration using a geographical model of gravitational acceleration. The calculation formula is as follows:

[0066]

[0067] in, Latitude Altitude;

[0068] S2. Real-time acquisition of triaxial acceleration data sequences from the target MEMS sensor, and calculation of the total modulus and modulus deviation of the triaxial acceleration data. The methods for real-time acquisition of triaxial acceleration data sequences from the target MEMS sensor and calculation of the total modulus and modulus deviation of the triaxial acceleration data include:

[0069] S21. Acquire real-time triaxial acceleration data of the MEMS sensor at a preset sampling period T. Forming a time series , , ,in, , This represents the number of sampling points;

[0070] S22. Calculate the total triaxial acceleration modulus at each time point ti. The specific counting method is as follows:

[0071]

[0072] in, , , These are the time series that were formed;

[0073] S23. Calculate the overall modulus deviation. That is, the absolute value of the difference between the current modulus and the reference gravitational acceleration;

[0074] S3. Determine whether the modulus deviation exceeds a preset deviation threshold. If it does, a preliminary suspicion of a uniaxial abrupt change is identified. The methods for determining whether the modulus deviation exceeds the preset deviation threshold and, if it does, a preliminary suspicion of a uniaxial abrupt change are included:

[0075] Preset deviation threshold The threshold is determined based on the static measurement accuracy of the MEMS sensor, specifically between 0.05g and 0.1g.

[0076] ,like > And only the change in data on a single axis compared to the previous time point. > The changes in the other two axes are all less than If so, it is preliminarily determined that the axis is a mutation axis and there is a suspicion of uniaxial mutation;

[0077] S4. Locate the abrupt change axis and calculate the correlation coefficient between the abrupt change axis and the acceleration data of the other two axes;

[0078] S5. Based on the correlation coefficient, the method for determining whether a single-axis abrupt change is an independent disturbance abrupt change, and using the stable data of the other two axes to correct the abrupt axis data based on the physical laws of static measurement, to locate the abrupt axis, and to calculate the correlation coefficient between the abrupt axis and the acceleration data of the other two axes includes:

[0079] S41. Let the axis of the initial mutation be the X-axis, and the other two axes be the Y-axis and Z-axis.

[0080] S42. Extract the data within the mutation time window and calculate the Pearson correlation coefficients between the mutation axis X and Y axes, and between X and Z axes, respectively. , The specific formula is as follows:

[0081] ;

[0082] in, The covariance of the X-axis and Y-axis data. The covariance of the X-axis and Z-axis data. , , The variances of the X, Y, and Z axis data are respectively used. The average of the Pearson correlation coefficients between the two axes is taken as the indicator of inter-axis correlation. Methods for determining whether a uniaxial mutation is an independent interference mutation based on the correlation coefficient include:

[0083] S51, Preset correlation coefficient threshold ;

[0084] S52, if < Therefore, the uniaxial mutation is ultimately determined to be an independent interference mutation and requires correction; if ≥ This is determined to be a change in axis data caused by actual motion, and no correction is required;

[0085] Based on the physical laws of static measurement, methods for correcting abrupt change axis data using stable data from the other two axes include:

[0086] S5011. Core physical law based on static measurement: As a rigid body, the sum of the three-axis acceleration vectors of a MEMS sensor under static conditions is equal to the local gravitational acceleration. ,Right now ,in, This is the data after X-axis correction. , For the other two axes, stabilize the data;

[0087] S5012, Introducing an inter-axis coupling correction factor Its value ranges from 0.95 to 1.05, so the formula for correcting the abrupt change axis is:

[0088]

[0089] in, The sign of the X-axis data at the moment before the mutation, if the corrected If the value exceeds the sensor's reasonable measurement range, a linear interpolation method is used to combine the stable data from the previous three time points with the correction formula results to smoothly transition to the correction value and avoid data jumps.

[0090] S6. Output the corrected triaxial acceleration data to correct for single-axis abrupt changes. Methods for outputting the corrected triaxial acceleration data include:

[0091] S61. Modify the mutation axis data Replace the original mutation data ,and , The corrected triaxial acceleration data are composed of ( ,and , );

[0092] S62. Verify the corrected total modulus deviation. ,like ≤ Then output the data; if it still exceeds the threshold, repeat the correction steps until the requirements are met.

[0093] A single-axis mutation correction system for a MEMS sensor, comprising:

[0094] The reference value acquisition module is used to acquire the local gravitational acceleration reference value at the deployment location of the target MEMS sensor. ;

[0095] The data acquisition and calculation module is used to acquire the triaxial acceleration data sequence of the target MEMS sensor in real time and calculate the total modulus and modulus deviation of the triaxial acceleration data.

[0096] The preliminary identification module is used to determine whether the modulus deviation exceeds the preset deviation threshold. If it does, it is preliminarily identified as having a suspected uniaxial mutation.

[0097] The precise determination module is used to locate the abrupt change axis and calculate the correlation coefficient between the abrupt change axis and the acceleration data of the other two axes. Based on the correlation coefficient, it determines whether a single-axis abrupt change is an independent interference abrupt change.

[0098] The data correction module is used to correct the abrupt change axis data based on the physical laws of static measurement and using stable data from the other two axes.

[0099] The data output module is used to output the corrected triaxial acceleration data to correct for single-axis abrupt changes.

[0100] The benchmark acquisition module, data acquisition and calculation module, preliminary identification module, precise judgment module, data correction module, and data output module are all implemented through computer programs and run on the same hardware platform.

[0101] Example 2:

[0102] Based on Example 1, the local gravitational acceleration reference value is obtained:

[0103] First, key geographic information of the deployment location is collected through the sensor's accompanying positioning module (such as GPS / BeiDou), specifically including latitude and altitude. Taking a typical outdoor monitoring scenario as an example, assuming the sensor is deployed in an area at latitude 39.9° and altitude 50 meters, the local gravitational acceleration is calculated using a gravitational acceleration geographic model. The calculation formula is as follows:

[0104]

[0105] in, Latitude This refers to altitude.

[0106] Substituting the above scenario parameters into the formula, the calculated local gravitational acceleration baseline value is approximately 9.8000 m / s². 2 This value will serve as the benchmark for determining whether the total triaxial acceleration modulus is normal in subsequent steps. It is directly related to the subsequent modulus deviation calculation and the verification of the correction effect in step S6. Therefore, the accuracy of the benchmark value directly affects the reliability of the entire correction process.

[0107] Example 3:

[0108] This embodiment corresponds to the continuous steps of acquiring triaxial data, calculating the modulus deviation, and initially identifying abrupt changes. It also fully replicates the logical judgment process for initially identifying uniaxial abrupt changes. First, two core thresholds need to be preset. The setting of these thresholds must be combined with the sensor's own performance to avoid misjudgments or omissions: The first is the modulus deviation threshold, determined based on the sensor's static measurement accuracy. In this embodiment, the sensor's static accuracy is ±30mg, so the modulus deviation threshold is set to 60mg. The second is the uniaxial abrupt change amplitude threshold, taken as 10% of the sensor's full scale. In this embodiment, the sensor's full scale is ±1633mg, so this threshold is set to 163.3mg. Furthermore, the sampling period is set to 50ms to ensure the continuity of data acquisition and meet the needs of outdoor real-time monitoring. First, the real-time acquired triaxial acceleration data and the historical data from the previous moment are input for the first judgment step: calculating the total modulus of the triaxial acceleration at the current moment and calculating the modulus deviation compared to the baseline value (999.32mg) obtained in Embodiment 1. In this embodiment, data was collected at a certain moment: the X-axis acceleration was 607.44 mg, compared to 404.74 mg at the previous moment, a change of 202.70 mg, far exceeding the single-axis abrupt change threshold of 163.3 mg; while the Y-axis and Z-axis data fluctuated very little, with changes of only 0.61 mg and 0.85 mg respectively, both within the threshold range. Further calculation showed that the total modulus of the three axes at the current moment was approximately 1097.02 mg, with a modulus deviation of 97.70 mg, far exceeding the preset threshold of 60 mg, satisfying the dual conditions of "modulus deviation exceeding the threshold" and "only single-axis fluctuation exceeding the threshold". Therefore, the X-axis was initially identified as the abrupt change axis, with a suspected single-axis abrupt change, and a step to accurately determine the type of abrupt change is required.

[0109] Example 4:

[0110] This embodiment addresses the core step of accurately determining the type of mutation. Since changes in axis data caused by actual motion inevitably trigger related changes in other axes through rigid body coupling, and interfering mutations are mostly independent, the authenticity of the mutation needs to be verified through inter-axis correlation coefficients. First, a time window is selected, the length of which must cover the stable data before and after the mutation to ensure the representativeness of the calculated correlation. In this embodiment, the sampling period is 50ms, so 8 sampling periods (400ms) are selected as the window length. The window range covers the stable data for the 3 periods before the mutation, the data at the moment of the mutation, and the transitional data for the 5 periods after the mutation. Within the selected time window, acceleration data for the X-axis (the initially determined mutation axis), Y-axis, and Z-axis are extracted, and the Pearson correlation coefficients between the X-axis and Y-axis, and between the X-axis and Z-axis are calculated. The core purpose is to determine whether there is a correlation between the mutation axis and other stable axes. The Pearson correlation coefficient ranges from [-1, 1], with values ​​closer to 0 indicating a weaker correlation. This embodiment presets a correlation coefficient threshold of 0.3; values ​​below this threshold are considered to have no significant correlation. The correlation coefficients between the X and Y axes were calculated to be 0.7081 and -0.7101, with an average of -0.0010, which are far below the threshold of 0.3. This indicates that the abrupt change in the X axis is not significantly related to the other two axes. It was ultimately confirmed that the abrupt change was an independent change caused by interference and requires further correction. If the correlation coefficient is higher than the threshold, it is determined to be real motion, and the original data can be output directly.

[0111] Example 5:

[0112] This embodiment corresponds to the final stage of correcting abrupt axis data, verifying the effect, and outputting data. The correction effect can be intuitively reflected through the modulus comparison curve. The core basis of the correction is the physical law of "conservation of the sum of the three-axis acceleration vectors" under static measurement, that is, the total modulus of the three axes after correction needs to return to near the local gravitational acceleration reference value. At the same time, in order to compensate for the inter-axis coupling deviation caused by sensor manufacturing defects, a coupling correction factor obtained during factory calibration is introduced (the factor value in this embodiment is 0.98). In order to ensure the continuity of the corrected data, the sign of the X-axis data before the abrupt change needs to be referenced (in this embodiment, the X-axis data before the abrupt change is positive, so the corrected data remains positive). Based on the above principle, the core correction formula is used to correct the X-axis abrupt change data (the formula is simplified to: corrected X-axis data = coupling correction factor × √(reference gravitational acceleration)). 2 -Y-axis data 2 -Z-axis data 2(×Pre-mutation data symbol). Substituting the known parameters into the formula, the corrected X-axis data is calculated to be approximately 397.07 mg. After correction, the corrected triaxial total modulus is recalculated to be approximately 996.06 mg, with a modulus deviation of only 3.26 mg, which has fallen within the preset 60 mg threshold, meeting the accuracy requirements. This effect is completely consistent with expectations. Before correction, the deviation from the baseline value (g=999.32 mg) was significant at the mutation time, exceeding the upper limit of the threshold; while after correction, the deviation quickly returned to near the baseline value after the mutation time and stabilized within the threshold range. Finally, the corrected triaxial acceleration data is output, completing the entire correction process for the single-axis mutation. If the deviation after correction still exceeds the threshold, the correction steps can be repeated until the requirements are met, ensuring the reliability of the output data. This effect is consistent with... Figure 3 Their curve characteristics are completely identical. Figure 3 The dashed line (before correction) deviates significantly from the baseline value (g=999.32mg) at the abrupt change, falling below the lower threshold. The solid line (after correction) quickly returns to near the baseline value after the abrupt change and stabilizes within the threshold range. The final output is the corrected triaxial acceleration data, completing the entire correction process for uniaxial abrupt changes. If the deviation still exceeds the threshold after correction, the correction steps can be repeated until the requirements are met, ensuring the reliability of the output data.

[0113] Based on the above embodiments, the following conclusions can be drawn:

[0114] 1. This invention, through a dual judgment mechanism of modulus deviation and inter-axis correlation coefficient, can accurately distinguish between interference mutations and real motion. In the embodiment, by setting a reasonable modulus deviation threshold (0.05g~0.1g) and a single-axis mutation amplitude threshold (5%~10% of the sensor's full scale), the mutation axis was successfully identified initially. The authenticity of the mutation was further verified by calculating the correlation coefficient, avoiding the erroneous correction caused by misjudgment in the prior art and ensuring the scientific nature and accuracy of the correction process.

[0115] 2. Based on the core physical laws of static measurement (conservation of triaxial acceleration vector sum) and the interaxial coupling correction factor (0.95~1.05), this invention can efficiently correct abrupt axis data. The corrected triaxial total modulus deviation is significantly reduced and returns to near the local gravitational acceleration reference value, meeting the preset accuracy requirements (such as modulus deviation ≤0.05g). This shows that the correction method can not only effectively restore the accuracy of the data, but also ensure the continuity and stability of the data, and avoid introducing new errors due to correction.

[0116] By accurately identifying interference mutations, false alarms and unnecessary emergency responses caused by misjudgment are reduced, thus lowering resource waste and maintenance costs. At the same time, the stability of the corrected data is improved, reducing frequent calibrations caused by data mutations. This effectively corrects uniaxial mutations in sensor data, significantly improving the accuracy and reliability of measurement data and providing more accurate data for monitoring and analysis in related fields.

[0117] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for correcting uniaxial abrupt changes in a MEMS sensor, characterized in that: Includes the following steps: S1. Obtain the local gravitational acceleration reference value g at the deployment location of the target MEMS sensor; S2. Real-time acquisition of triaxial acceleration data sequence of the target MEMS sensor, and calculation of the total modulus and modulus deviation of the triaxial acceleration data; S3. Determine whether the modulus deviation exceeds a preset deviation threshold. If it does, preliminarily identify a possible uniaxial abrupt change. Step S3 specifically includes: Preset deviation threshold The threshold is determined based on the static measurement accuracy of the MEMS sensor, specifically between 0.05g and 0.1g. ,like > And only the change in data on a single axis compared to the previous time point. > The changes in the other two axes are all less than If so, it is preliminarily determined that the axis is a mutation axis and there is a suspicion of uniaxial mutation; S4. Locate the abrupt change axis and calculate the correlation coefficient between the abrupt change axis and the acceleration data of the other two axes; S5. Based on the correlation coefficient, determine that the single-axis mutation is an independent interference mutation. Based on the physical laws of static measurement, use the stable data of the other two axes to correct the mutation axis data. S6. Output the corrected triaxial acceleration data to complete the correction of single-axis abrupt changes.

2. The method for correcting single-axis abrupt changes in a MEMS sensor according to claim 1, characterized in that: The method for obtaining the local gravitational acceleration reference value g at the deployment location of the target MEMS sensor includes: S11. Collect the latitude of the target MEMS sensor deployment location. and altitude ; S12. Calculate the local gravitational acceleration using a geographical model of gravitational acceleration. The calculation formula is as follows: in, Latitude This refers to altitude.

3. The method for correcting single-axis abrupt changes in a MEMS sensor according to claim 1, characterized in that: The method for real-time acquisition of triaxial acceleration data sequences from the target MEMS sensor and calculation of the total modulus and modulus deviation of the triaxial acceleration data includes: S21. Acquire real-time triaxial acceleration data of the MEMS sensor at a preset sampling period T. Forming a time series , , ,in, , This represents the number of sampling points; S22. Calculate the total triaxial acceleration modulus at each time point ti. The specific counting method is as follows: in, , , These are the time series that were formed; S23. Calculate the overall modulus deviation. That is, the absolute value of the difference between the current modulus and the reference gravitational acceleration.

4. The method for correcting single-axis abrupt changes in a MEMS sensor according to claim 1, characterized in that: The method for locating the abrupt change axis and calculating the correlation coefficient between the abrupt change axis and the acceleration data of the other two axes includes: S41. Let the axis of the initial mutation be the X-axis, and the other two axes be the Y-axis and Z-axis. S42. Extract the data within the mutation time window and calculate the Pearson correlation coefficients between the mutation axis X and Y axes, and between X and Z axes, respectively. , The specific formula is as follows: ; in, The covariance of the X-axis and Y-axis data. The covariance of the X-axis and Z-axis data. , , The variances of the X, Y, and Z axis data are respectively used. The average of the Pearson correlation coefficients between the two axes is taken as the indicator of inter-axis correlation. .

5. The method for correcting single-axis abrupt changes in a MEMS sensor according to claim 1, characterized in that: The method for determining whether a uniaxial mutation is an independent interference mutation based on the correlation coefficient includes: S51, Preset correlation coefficient threshold ; S52, if < Therefore, the uniaxial mutation is ultimately determined to be an independent interference mutation and requires correction; if ≥ If the change is due to actual motion, it is determined to be a change in axis data and no correction is needed.

6. The method for correcting single-axis abrupt changes in a MEMS sensor according to claim 1, characterized in that: The method for correcting abrupt axis data using stable data from the other two axes based on the physical laws of static measurement includes: S5011. Core physical law based on static measurement: As a rigid body, the sum of the three-axis acceleration vectors of a MEMS sensor under static conditions is equal to the local gravitational acceleration. ,Right now ,in, This is the data after X-axis correction. , For the other two axes, stabilize the data; S5012, Introducing an inter-axis coupling correction factor Its value ranges from 0.95 to 1.05, so the formula for correcting the abrupt change axis is: in, The sign of the X-axis data at the moment before the mutation, if the corrected If the value exceeds the sensor's reasonable measurement range, a linear interpolation method is used to combine the stable data from the previous three time points with the correction formula results to smoothly transition to the correction value and avoid data jumps.

7. The method for correcting single-axis abrupt changes in a MEMS sensor according to claim 1, characterized in that: The method for outputting corrected triaxial acceleration data includes: S61. Modify the mutation axis data Replace the original mutation data ,and , The corrected triaxial acceleration data are composed of ( ,and , ); S62. Verify the corrected total modulus deviation. ,like ≤ Then output the data; if it still exceeds the threshold, repeat the correction steps until the requirements are met.

8. A single-axis abrupt change correction system for a MEMS sensor, based on the single-axis abrupt change correction method for a MEMS sensor according to any one of claims 1-7, characterized in that: include: The reference value acquisition module is used to acquire the local gravitational acceleration reference value at the deployment location of the target MEMS sensor. ; The data acquisition and calculation module is used to acquire the triaxial acceleration data sequence of the target MEMS sensor in real time and calculate the total modulus and modulus deviation of the triaxial acceleration data. The preliminary identification module is used to determine whether the modulus deviation exceeds a preset deviation threshold. If it does, it is preliminarily identified as having a suspected uniaxial abrupt change. A precise determination module is used to locate the abrupt change axis and calculate the correlation coefficient between the abrupt change axis and the acceleration data of the other two axes. Based on the correlation coefficient, the module determines that the single-axis abrupt change is an independent interference abrupt change. The data correction module is used to correct the abrupt change axis data based on the physical laws of static measurement and using stable data from the other two axes. The data output module is used to output the corrected triaxial acceleration data to correct for single-axis abrupt changes.

9. The MEMS sensor uniaxial abrupt change correction system according to claim 8, characterized in that, The benchmark value acquisition module, data acquisition and calculation module, preliminary identification module, precise judgment module, data correction module, and data output module are all implemented by computer programs and run on the same hardware platform.

Images