MEMS Accelerometer and Its Dynamic Compensation Method

Abstract

The present invention discloses a MEMS acceleration sensor and its dynamic compensation method. The dynamic compensation method includes the following steps: S1. Obtain the measured acceleration a through the MEMS acceleration sensing module; S2. The single-chip microcomputer uses the mixed-radix DIF-FFT algorithm to calculate the characteristic frequency f of the measured acceleration a; S3. The single-chip microcomputer obtains the proportionality factor K corresponding to the characteristic frequency f according to the frequency response fitting curve A ; S4. The single-chip microcomputer outputs the final acceleration a after compensation according to the proportionality factor K A The present invention can not only improve the real-time output accuracy of the MEMS acceleration sensor, but also improve the processing speed of dynamic compensation c ​

Description

Technical Field

[0001] The present invention relates to the technical field of acceleration sensors, and in particular to a MEMS acceleration sensor and its dynamic compensation method. Background Art

[0002] A MEMS (Micro-Electro-Mechanical System) acceleration sensor is a new type of acceleration sensor produced by using MEMS and integrated circuit processing technologies. Compared with traditional acceleration sensors, MEMS acceleration sensors have the advantages of small size, low cost, high sensitivity, good stability, etc., and have been widely used in fields such as vibration monitoring of electromechanical equipment, medical equipment, and aerospace testing. The working principle of a MEMS acceleration sensor is to convert the acceleration signal applied from the outside into an electrical signal through a microcomputer structure, so as to measure the magnitude of the acceleration. However, due to the non-ideal dynamic characteristics of the microcomputer structure of the MEMS acceleration sensor, there is a certain dynamic error in the measurement structure of the acceleration.

[0003] Currently, the correction of the measurement results of MEMS acceleration sensors is mainly carried out by exporting the measurement results from the sensor and then processing them with computer software. This method has obvious after-the-fact nature and cannot be applied to occasions where accurate acceleration monitoring results need to be obtained in real time. Summary of the Invention

[0004] The technical problem to be solved by the present invention is: to solve the technical problem that the existing MEMS acceleration sensors cannot output accurate monitoring results in real time. The present invention provides a MEMS acceleration sensor and its dynamic compensation method, which can not only improve the real-time output accuracy of the MEMS acceleration sensor, but also improve the processing speed of dynamic compensation.

[0005] The technical solution adopted by the present invention to solve its technical problems is: a dynamic compensation method for a MEMS acceleration sensor suitable for mine use, including the following steps:

[0006] S1. Obtain the measured acceleration a through a MEMS acceleration sensing module;

[0007] S2. The single-chip microcomputer uses the mixed-radix DIF-FFT algorithm to calculate the characteristic frequency f of the measured acceleration a;

[0008] S3. The single-chip microcomputer obtains the proportionality factor K corresponding to the characteristic frequency f according to the frequency response fitting curve A ;

[0009] S4. The single-chip microcomputer outputs the final acceleration a after compensation according to the proportionality factor K A c ​

[0010] Further, the frequency response fitting curve is used to represent the relationship between the characteristic frequency f of the measured acceleration a and the scale factor K A The relational expression of the frequency response fitting curve is: K A = c0 + c1×cos(w·f) + c2×sin(w·f), where c0, c1, c2, and w are all coefficients.

[0011] Further, in step S2, the calculation process of the characteristic frequency f of the measured acceleration a includes:

[0012] Set the sample point number threshold N0;

[0013] Obtain all sample points x(n) of the measured acceleration a measured by the MEMS acceleration sensing module within the sampling time t, where n = 0, 1,..., N - 1, and N represents the total number of sample points;

[0014] Judge whether the total number N of the sample points reaches the sample point number threshold N0. If so, process the N sample points x(n) using the mixed - radix DIF - FFT algorithm to obtain the characteristic frequency f of the measured acceleration a.

[0015] Further, the process of processing the N sample points x(n) using the mixed - radix DIF - FFT algorithm includes:

[0016] Decompose the total number N of sample points into subsequence n1 and subsequence n2,

[0017] Obtain the radix A of the subsequence n1 and the radix B of the subsequence n2 respectively;

[0018] Determine the operation mode of the mixed - radix DIF - FFT algorithm according to the radix A and the radix B;

[0019] Process the N sample points x(n) according to the determined operation mode.

[0020] Further, the sample point number threshold is set to N0 = 1024, then the total number N of sample points = n1 * n2 = 1024 = 2 * 8 3 , the radix A of the subsequence n1 = 2, the radix B of the subsequence n2 = 8, and the operation mode of the mixed - radix DIF - FFT algorithm includes 1 - level radix - 2 butterfly operation and 3 - level radix - 8 butterfly operation.

[0021] Further, the discrete - time domain Fourier transform formula of the sample point x(n) is where, represents the rotation factor, The radix - 2 butterfly operation includes:

[0022] The discrete-time Fourier transform of the sample point x(n) is decomposed into two parts according to whether n is even or odd:

[0023]

[0024] Let the even n = 2r1 and the odd n = 2r1 + 1, where r1 = 0, 1, 2,..., (N / 2) - 1; then

[0025]

[0026] Thus, the operation result of the radix-2 butterfly operation for N sample points is obtained.

[0027] Furthermore, the radix-8 butterfly operation includes: let k = 8r2 + l, where r2 = 0, 1, 2,..., (N / 8) - 1; l = 0, 1, 2, 3, 4, 5, 6, 7; the radix-8 butterfly operation formula is:

[0028]

[0029] Through multi-level operations, the operation result of the radix-8 butterfly operation for N sample points is obtained.

[0030] Furthermore, the final acceleration

[0031] The present invention also provides a MEMS acceleration sensor, which adopts the dynamic compensation method described above, and includes:

[0032] A MEMS acceleration sensing module for collecting and measuring the acceleration a;

[0033] A single-chip microcomputer, which is communicatively connected to the MEMS acceleration sensing module, for calculating the characteristic frequency f of the measured acceleration a; obtaining the proportionality factor K corresponding to the characteristic frequency f according to the frequency response fitting curve A ; according to the proportionality factor K A Outputting the compensated final acceleration a c ;

[0034] A power supply module, to which the acceleration sensing module and the single-chip microcomputer are both connected.

[0035] The beneficial effect of the present invention is that the MEMS acceleration sensor and its dynamic compensation method of the present invention integrate the MEMS acceleration sensing module and the single-chip microcomputer together, and the single-chip microcomputer dynamically compensates the output value of the MEMS acceleration sensing module in real time; based on the frequency response model, the proportionality factor K between the measured acceleration and the actual acceleration is proposed A, which can quickly compensate the measured values; by adopting the mixed - radix DIF - FFT algorithm, the processing efficiency of the single - chip microcomputer can be significantly improved. The present invention can not only improve the real - time output accuracy of the MEMS acceleration sensor, but also improve the processing speed of dynamic compensation. Description of the Drawings

[0036] The present invention will be further described below in conjunction with the drawings and embodiments.

[0037] Figure 1 is the flowchart of the dynamic compensation method of the present invention.

[0038] Figure 2 is the specific process diagram of the dynamic compensation method of the present invention.

[0039] Figure 3 is the schematic diagram of the frequency response fitting curve of the present invention.

[0040] Figure 4 is the schematic diagram of the amplitude - frequency characteristic curve of the MEMS acceleration sensing module of the present invention.

[0041] Figure 5 is the experimental result curve of the embodiment and the comparative example of the present invention.

[0042] Figure 6 is the schematic structural diagram of the MEMS acceleration sensor of the present invention.

[0043] In the figure: 1. MEMS acceleration sensing module; 2. Single - chip microcomputer; 3. Power supply module. Detailed Embodiment

[0044] Now, the present invention will be further described in detail with reference to the drawings. These drawings are all simplified schematic diagrams, only illustrating the basic structure of the present invention in a schematic way, so they only show the components related to the present invention.

[0045] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, so it cannot be understood as a limitation of the present invention. In addition, the features defined as "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present invention, unless otherwise specified, the meaning of "a plurality" is two or more.

[0046] In the description of the present invention, it should be noted that unless otherwise clearly specified and defined, the terms "installation", "connection", and "coupling" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, and it can be the communication inside two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0047] As Figures 1 to 3 shown, the dynamic compensation method of the MEMS acceleration sensor of the present invention includes the following steps: S1. Obtain the measured acceleration a through the MEMS acceleration sensor. S2. The single-chip microcomputer uses the mixed-radix DIF-FFT algorithm to calculate the characteristic frequency f of the measured acceleration a. S3. The single-chip microcomputer obtains the proportionality factor K corresponding to the characteristic frequency f according to the frequency response fitting curve A . S4. The single-chip microcomputer outputs the final compensated acceleration a according to the proportionality factor K A . c .

[0048] It should be noted that the MEMS acceleration sensor of the present invention includes a MEMS acceleration sensing module and a single-chip microcomputer. The acceleration data collected by the MESM acceleration sensing module is directly sent to the single-chip microcomputer for real-time processing, and finally outputs the acceleration value after dynamic compensation. Thus, in practical applications, the staff does not need to import the acceleration data measured by the acceleration sensor into the host computer for data compensation, which can significantly improve work efficiency.

[0049] There is a certain proportional relationship between the measured acceleration a output by the MEMS acceleration sensing module and the actually applied acceleration. Let the proportionality factor be K A . And the proportionality factor K A has a corresponding relationship with the frequency response of the MEMS acceleration transmission module.

[0050] The frequency response includes the response of the microcomputer structure and the filtering response in the signal chain. The microcomputer structure of the MEMS acceleration sensing module can be equivalent to a spring-mass-damper vibration system. When the microcomputer structure is affected by acceleration, the differential equation of the displacement of the mass block is: where m represents mass, y represents displacement, k represents stiffness, c represents damping coefficient, and a1 represents the applied acceleration.

[0051] Performing Laplace transform (s-transform) on the cover differential equation can obtain the second-order transfer function of the input and output of the microcomputer structure: where ω0 represents the resonant frequency, ξ represents the damping ratio coefficient, and s represents the complex domain variable.

[0052] The filtering response in the signal chain includes analog filtering and digital filtering. The MEMS acceleration sensing module uses an analog low-pass filter to reduce the response gain near the resonance frequency. The transfer function of the analog first-order low-pass filter signal response after the s-transform is: where ω c represents the cut-off frequency of the analog low-pass filter. A digital filter can also be set in the MEMS acceleration sensing module to configure the ODR (data output rate) and the cut-off frequency of the digital signal. The transfer function of the digital filter is: where ω d represents the cut-off frequency of the digital filter.

[0053] Combining the above several formulas, the frequency response model of the MEMS acceleration sensing module can be obtained as: Using this frequency response model, the attenuation degree of the acceleration signal at each frequency point of the MEMS acceleration sensing module can be predicted.

[0054] For example, assuming ω0 = 5500Hz, ω c = 2400Hz, ω d = 1000Hz, ξ = 0.0294, according to the frequency response model, the amplitude-frequency characteristic curve of the MEMS acceleration sensing module can be obtained (as shown in Figure 4 ). It can be seen from the figure that at 1000Hz, the acceleration measurement value output by the MEMS acceleration sensing module is attenuated by about 30%. Thus, it can be seen that there is a corresponding relationship between the attenuation degree of the measurement value of the MEMS acceleration sensing module and the characteristic frequency of the acceleration. Based on this, the present invention proposes to calculate the characteristic frequency f of the measured acceleration a, obtain the proportionality factor K corresponding to the characteristic frequency f according to the frequency response fitting curve A , and then output the final acceleration a A after compensation according to the proportionality factor K c compensation method.

[0055] It should be noted that the frequency response fitting curve is obtained through experiments in advance. For example, set the vibration table acceleration to 4g n , sweep the frequency in the range of 0 - ~1KHz at intervals of 50Hz, and the MEMS acceleration sensing module only collects data without performing dynamic error compensation processing. By fitting the measurement data of the MEMS acceleration sensing module, the relationship curve between the proportionality factor K A between the output result of the MEMS acceleration sensing module and the actual acceleration and the frequency can be obtained (as shown in the figure). For example, the relational expression of the frequency response fitting curve is: K A= c0 + c1×cos(w·f) + c2×sin(w·f), where c0, c1, c2, and w are all coefficients. For example, c0 = 0.6691, c1 = 0.3298, c2 = 0.004807, and w = 0.001562. This curve shows that the higher the frequency of the acceleration, the more severely the measured value of the acceleration sensor decays.

[0056] Thus, after the microcontroller calculates the characteristic frequency f of the measured acceleration a, the characteristic frequency f can be substituted into the frequency response fitting curve to obtain the corresponding scale factor K. A . Then, according to the calculation formula of the final acceleration the compensated acceleration value is obtained.

[0057] In step S2, the calculation process of the characteristic frequency f of the measured acceleration a includes: setting the sample point quantity threshold N0, obtaining all sample points x(n) of the measured acceleration a measured by the MEMS acceleration sensing module within the sampling time t, where n = 0, 1,..., N - 1, and N represents the total number of sample points. It is judged whether the total number of sample points N reaches the sample point quantity threshold N0. If so, the mixed-radix DIF-FFT algorithm is used to process the N sample points x(n) to obtain the characteristic frequency f of the measured acceleration a.

[0058] It should be noted that the MEMS acceleration sensing module is collecting acceleration data in real time, and thousands of acceleration data can be collected within 1 second. To improve the processing efficiency of the microcontroller, ensure the real-time output of the MEMS acceleration sensor, and at the same time, reduce the memory requirement of the microcontroller during the operation process, the present invention sets the sample point quantity threshold N0 to 1024. That is to say, when the acceleration data collected by the MEMS acceleration sensing module within the sampling time t reaches 1024, the microcontroller performs arithmetic processing on these 1024 acceleration data. After processing, the microcontroller automatically deletes these 1024 acceleration data. When the number of sampling points reaches 1024 again, the new 1024 sample points are processed again, and are deleted after processing, and so on in a loop.

[0059] For example, the present invention uses the mixed-radix DIF-FFT algorithm to process N sample points. The processing process includes: decomposing the total number of sample points N into subsequence n1 and subsequence n2, and respectively obtaining the radix A of subsequence n1 and the radix B of subsequence n2; determining the operation mode of the mixed-radix DIF-FFT algorithm according to radix A and radix B; processing the N sample points according to the determined operation mode. It should be noted that the DIF-FFT algorithm is the abbreviation of the decimation-in-frequency fast Fourier transform algorithm. For example, when N reaches 1024, the sequence N can be decomposed into n1*n2, that is, N = n1*n2 = 1024 = 2 * 8. 3, the cardinality A of subsequence n1 is 2, the cardinality B of subsequence n2 is 8, and the operation mode of the mixed-radix DIF-FFT algorithm includes 1-level radix-2 butterfly operations and 3-level radix-8 butterfly operations. Compared with the pure radix-2 DIF-FFT algorithm, radix-4 DIF-FFT algorithm, and radix-8 DIF-FFT algorithm, the mixed-radix DIF-FFT algorithm can significantly reduce the number of iterative levels of operations, reduce the number of operations of multipliers and adders, and improve the real-time operation speed.

[0060] For example, the discrete-time domain Fourier transform formula of the sample point x(n) is where represents the rotation factor, After performing the discrete-time domain Fourier transform on the acceleration measurement value, it can be transformed into a frequency domain signal.

[0061] The radix-2 butterfly operation includes: decomposing the discrete-time domain Fourier transform of the sample point x(n) into two parts according to whether n is even or odd: Let the even n = 2r1, the odd n = 2r1 + 1, r1 = 0, 1, 2,..., (N / 2) - 1; then

[0062]

[0063] Thus, the operation result of the radix-2 butterfly operation on N sample points is obtained.

[0064] The radix-8 butterfly operation includes: let k = 8r2 + l, r2 = 0, 1, 2,..., (N / 8) - 1; l = 0, 1, 2, 3, 4, 5, 6, 7; the radix-8 butterfly operation formula is:

[0065]

[0066] Substitute l into the formula, and after multi-level operations, the operation result of the radix-8 butterfly operation on N sample points is obtained. Each time the radix-8 butterfly operation processes 8 sample points. Although the complexity increases, the number of iterative levels of operations will be significantly reduced, and the operation speed is significantly improved. Integrate the results of 1-level radix-2 butterfly operations and 3-level radix-8 butterfly operations to obtain the final operation result. Take the frequency point with the largest amplitude in the final operation result as the characteristic frequency f.

[0067] Next, the technical effects of this method are illustrated through comparative experiments.

[0068] Example

[0069] Set the acceleration of the vibration table to 4g n , in the frequency sweep mode of 0 - 1KHz, the single-chip microcomputer uses the mixed-radix DIF-FFT algorithm for real-time dynamic compensation processing and outputs the final acceleration value.

[0070] Comparative example

[0071] Set the acceleration of the vibration table to 4g n , in the sweep frequency mode of 0 - 1KHz, without real-time dynamic compensation processing, directly output the acceleration value.

[0072] As Figure 5 shown, the abscissa represents the frequency, the ordinate represents the output acceleration, Curve 1 represents the output result of the embodiment, and Curve 2 represents the output result of the comparative example. It can be seen from the figure that after dynamic compensation, the result output by the MEMS sensor fluctuates stably around 4g n , and the error is within an acceptable range. Without dynamic compensation, as the frequency increases, the output acceleration shows a gradually decreasing trend, and the deviation from the actual value of 4g n becomes larger and larger. It can be proved from this that by adopting the dynamic compensation method of the present invention, the real-time output accuracy of the MEMS acceleration sensor can be significantly improved.

[0073] As Figure 6 shown, the present invention also provides a MEMS acceleration sensor, adopting the dynamic compensation method, including: a MEMS acceleration sensing module 1, a single-chip microcomputer 2, and a power supply module 3. The MEMS acceleration sensing module 1 is used to collect and measure the acceleration a. The single-chip microcomputer 2 is communicatively connected to the MEMS acceleration sensing module 1 and is used to calculate the characteristic frequency f of the measured acceleration a; obtain the proportionality factor K corresponding to the characteristic frequency f according to the frequency response fitting curve A ; output the final compensated acceleration a according to the proportionality factor K A . Both the acceleration sensing module 1 and the single-chip microcomputer 2 are connected to the power supply module 3. c . The measured acceleration a collected by the MEMS acceleration sensing module 1 can be transmitted to the single-chip microcomputer 2 through the serial port. When the MEMS acceleration sensing module 1 collects new data, it outputs a high-level pulse signal to the single-chip microcomputer 2 through the DRDY pin. After the single-chip microcomputer 2 recognizes the pulse signal, it reads the acceleration data sent by the serial port. After the single-chip microcomputer 2 performs dynamic compensation on the acceleration data, it outputs it to other devices such as the host computer and the display through the serial port. The measurement range of the MEMS acceleration sensing module 1 is 0 - 16g

[0074] , and the single-chip microcomputer 2 uses a microprocessor of the STM32L series with ultra-low power consumption and floating-point operation. The computing power can reach 100DMIPS, which can effectively execute complex algorithms and improve the processing efficiency. n

[0075] In summary, for the MEMS acceleration sensor and its dynamic compensation method of the present invention, by integrating the MEMS acceleration sensing module and the single-chip microcomputer, the single-chip microcomputer dynamically compensates the output value of the MEMS acceleration sensing module in real time; based on the frequency response model, the proportional factor K between the measured acceleration and the actual acceleration is proposed. A , which can quickly compensate the measured value; by adopting the mixed-radix DIF-FFT algorithm, the processing efficiency of the single-chip microcomputer can be significantly improved. The present invention can not only improve the real-time output accuracy of the MEMS acceleration sensor, but also improve the processing speed of dynamic compensation.

[0076] Inspired by the ideal embodiments of the present invention described above, through the above description, relevant staff can make various changes and modifications completely within the scope without departing from the technical idea of the present invention. The technical scope of the present invention is not limited to the content in the specification, and its technical scope must be determined according to the scope of the claims.

Claims

1. A dynamic compensation method for a MEMS acceleration sensor, characterized in that It includes the following steps: S1. Obtain the measured acceleration a through the MEMS acceleration sensing module; S2. The single-chip microcomputer uses the mixed-radix DIF-FFT algorithm to calculate the characteristic frequency f of the measured acceleration a. Specifically, it includes: Set the sample point number threshold N0; Obtain all sample points x(n) of the measured acceleration a measured by the MEMS acceleration sensing module within the sampling time t, where n = 0, 1,..., N - 1, and N represents the total number of sample points; Judge whether the total number of sample points N reaches the sample point number threshold N0. If so, use the mixed-radix DIF-FFT algorithm to process the N sample points x(n) to obtain the characteristic frequency f of the measured acceleration a; The process of using the mixed-radix DIF-FFT algorithm to process the N sample points x(n) includes: Decompose the total number of sample points N into subsequence n1 and subsequence n2; Obtain the radix A of the subsequence n1 and the radix B of the subsequence n2 respectively; Determine the operation mode of the mixed-radix DIF-FFT algorithm according to the radix A and the radix B; Process the N sample points x(n) according to the determined operation mode; The threshold of the number of sample points is set to N0 = 1024, and the total number of sample points N = n1 * n2 = 1024 = 2 * 8 3 , the radix A of the subsequence n1 is 2, the radix B of the subsequence n2 is 8, and the operation mode of the mixed-radix DIF-FFT algorithm includes 1-level radix-2 butterfly operations and 3-level radix-8 butterfly operations; S3. The single-chip microcomputer obtains the proportionality factor K corresponding to the characteristic frequency f according to the frequency response fitting curve A ; The frequency response fitting curve is used to represent the relationship between the characteristic frequency f of the measured acceleration a and the proportionality factor K A ; The relational expression of the frequency response fitting curve is as follows: K A = c0 + c1×cos(w·f) + c2×sin(w·f), where c0, c1, c2, and w are all coefficients; S4. The single-chip microcomputer outputs the final compensated acceleration a according to the scale factor K A c .​ 2. The dynamic compensation method according to claim 1, wherein The discrete-time Fourier transform formula of the sample point x(n) is where represents the rotation factor, k = 0, 1,..., N - 1; the radix-2 butterfly operation includes: Decompose the discrete-time domain Fourier transform of the sample point x(n) into two parts according to whether n is even or odd: Let the even n = 2r1, the odd n = 2r1 + 1, where r1 = 0, 1, 2,..., (N / 2) - 1; then Thus, the operation result of the radix-2 butterfly operation on the N sample points is obtained.

3. The dynamic compensation method according to claim 2, characterized in that, The radix-8 butterfly operation includes: let k = 8r2 + l, where r2 = 0, 1, 2,..., (N / 8) - 1; l = 0, 1, 2, 3, 4, 5, 6, 7; the radix-8 butterfly operation formula is: Through multi-stage operations, the operation result of the radix-8 butterfly operation on the N sample points is obtained.

4. The dynamic compensation method according to claim 1, wherein The final acceleration 5. A MEMS acceleration sensor, characterized in that, Adopt the dynamic compensation method described in any one of claims 1-4, including: The MEMS acceleration sensing module (1) is used to collect the measured acceleration a; A single-chip microcomputer (2), which is communicatively connected to the MEMS acceleration sensing module (1), is used to calculate the characteristic frequency f of the measured acceleration a; and obtain the proportionality factor K corresponding to the characteristic frequency f according to the frequency response fitting curve A ; according to the proportionality factor K A output the final compensated acceleration a c ; The power supply module (3), and the acceleration sensing module (1) and the single-chip microcomputer (2) are both connected to the power supply module (3).

Images