Accelerometer driving and control method and system based on photoelectric signal

Abstract

The invention discloses an accelerometer driving and control method and system based on photoelectric signals, and relates to the technical field of accelerometers. Comprising the steps of obtaining a to-be-processed signal, identifying removed data, enhancing the noise suppression capability, purifying a useful signal, collecting noise with stable iteration characteristics, removing noise with stable characteristics, selecting algorithms, judging the calculation time of a time domain convolution algorithm and a moving average filtering algorithm when same data is processed according to the size of the data volume, selecting an algorithm with shorter time, and carrying out calculation according to the calculation time of the time domain convolution algorithm and the moving average filtering algorithm. The algorithm iteration is simplified, and the stopping condition of the algorithm iteration process is simplified, so that the algorithm finishes circulation on the premise of meeting the precision requirement and the number of iterations. According to the method, the filtering parameters can be automatically adjusted according to the current environmental condition, so that the optimal noise suppression effect is achieved, the calculation burden is relieved, the real-time response capability is improved, meanwhile, according to the size of the data size, it is judged that an algorithm with shorter time is selected when the same data is processed, and the time needed by data processing is shortened.

Description

Technical Field

[0001] The present invention relates to the technical field of accelerometers, and in particular to an accelerometer driving and controlling method and system based on photoelectric signals. Background Art

[0002] With the increasing demand for precise motion detection in modern industry, aerospace, automotive safety and consumer electronics, traditional accelerometers (such as piezoelectric, capacitive, etc.) have gradually shown limitations in sensitivity, stability and durability. In order to solve these problems, accelerometer driving and control methods and systems based on photoelectric signals have emerged. This new type of accelerometer uses optical principles to measure the acceleration changes of objects, providing a new approach different from traditional mechanical or electronic sensors.

[0003] The invention discloses an accelerometer control method, an accelerometer control device and an intelligent terminal with application publication number CN107389980A. The accelerometer control method comprises: obtaining the current vibration condition and reporting frequency of the intelligent terminal when the accelerometer is in operation, wherein the reporting frequency is the frequency at which the driving layer of the intelligent terminal reports the output data of the accelerometer to the application layer of the intelligent terminal, and the vibration condition indicates whether the intelligent terminal is in a vibration state; determining the target filter control parameters of the accelerometer based on preset filter parameter adjustment information according to the current vibration condition and reporting frequency, wherein the filter parameter adjustment information comprises the corresponding relationship information among the vibration condition, the reporting frequency and the filter control parameters; adjusting the current filter control parameters of the accelerometer based on the target filter control parameters, thereby realizing dynamic adjustment of the filter control parameters of the accelerometer.

[0004] The algorithm processing real-time response capability of the existing photoelectric signal accelerometer drive and control system is poor. Complex signals will cause data processing delays, affecting the real-time response capability of the system. In application scenarios that require fast feedback (such as high-speed motion detection), this delay may cause the control system performance to degrade. In addition, the existing photoelectric signal accelerometer drive and control system has poor anti-noise interference processing. Although photoelectric technology has good resistance to electromagnetic interference, in actual applications, environmental noise (such as mechanical vibration, etc.) will still affect the measurement accuracy. The existing control software lacks an effective mechanism to dynamically adjust to adapt to these changes. Summary of the invention

[0005] The object of the present invention is to provide an accelerometer driving and controlling method and system based on photoelectric signals to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: a method and system for driving and controlling an accelerometer based on a photoelectric signal, the method comprising: Obtain the signal to be processed and the original signal collected by the photoelectric sensor; Identify and remove data, identify and process the input original signal and remove irrelevant data; Enhance noise suppression capability, estimate the system status through noisy data, and adjust the filter parameters according to the current environmental conditions to achieve the best noise suppression effect; Pure useful signal, using the reference signal to estimate and remove the noise components in the main signal to obtain a pure useful signal; Collect iterative characteristic stable noise, continuously detect relatively stable noise, and enter it into the specific noise database after reaching a certain frequency; Remove noise with stable characteristics. Use an algorithm to detect relatively stable noise in the original signal, compare it with a specific noise database, and remove the noise in the specific noise database. Algorithm selection: Based on the amount of data, determine the calculation time of the time domain convolution algorithm and the moving average filter algorithm when processing the same data, and select the algorithm with shorter calculation time. The specific steps are as follows: S1: Set the threshold. When the data volume is lower than the threshold, select the sliding average filtering algorithm. S2: Set a threshold. When the data volume is higher than the threshold, select a time domain convolution algorithm to convert the time domain signal into a frequency domain signal. S3: Use the time domain convolution algorithm to process the frequency domain signal. The algorithm formula is as follows: Among them, Y(k) is the frequency domain representation of the output signal, X(k) is the frequency domain representation of the input signal, and H(k) is the frequency domain representation of the filter; S4: After the calculation is completed, the frequency domain signal is reversely converted into a time domain signal; Simplify the algorithm iteration and the stopping conditions of the algorithm iteration process so that the algorithm can end the loop while meeting the accuracy requirements and the number of iterations.

[0007] The identification and removal of data is to perform signal preprocessing on the input original signal, remove the non-correlated parts through detrending processing, cut out the irrelevant parts in the signal, and reduce the amount of calculation.

[0008] The enhanced noise suppression capability sets an initial state estimate and its covariance matrix, predicts the state at the current moment based on the state estimate at the previous moment, corrects the prediction value through signal data, adjusts the filter parameters, and dynamically suppresses noise.

[0009] The steps of pure useful signal are as follows: L1: collect the main signal, which is composed of useful signal and noise signal; L2: Set the reference signal. The reference signal calculation formula is as follows: Where r is the i-th reference signal, h(t) is the filter used to adjust the i-th reference signal, y(t) is the adjusted i-th reference signal, and Fori is the value range of the index variable i; L3: All adjusted reference signals are weighted and summed to obtain an estimate of the noise signal. The formula is as follows: Wherein, w is the weight coefficient of the i-th adjusted reference signal, and n(t) is the estimated value of the noise signal; L4: Useful signal extraction, subtract the estimated noise signal from the main signal to obtain the estimated value of the pure useful signal. The calculation formula is as follows: Among them, S(t) is the estimated value of the pure useful signal, x(t) is the main signal, and n(t) is the estimated value of the noise signal.

[0010] The method collects stable noise with iterative characteristics and continuously detects noise with relatively stable characteristics. When the same noise with relatively stable characteristics is detected a set number of times, it is entered into a specific noise database for spectrum subtraction data model calculation to reduce the calculation burden.

[0011] The algorithm is used to remove the noise with relatively stable characteristics in the main signal. The algorithm formula is as follows: Wherein, S(f) represents the estimated value of the spectrum of the pure useful signal, X(f) represents the spectrum of the main signal, α represents the adjustment factor, and N(f) represents the estimated value of the noise spectrum.

[0012] When the data volume is lower than the threshold, the sliding average filtering algorithm is selected, and the algorithm formula is as follows: Where: y(n) is the output signal, that is, the signal after sliding average filtering, M1 represents the average coefficient of the signal samples in the sliding window, ∑ M-1 K=0 represents a summation operation, x represents the input signal, and (nk) represents the time index or sample index of the input signal.

[0013] When the data volume is higher than the threshold, the time domain convolution algorithm is selected to convert the time domain signal to the frequency domain signal. The conversion formula is as follows: Where: X(k) is the frequency domain representation of the input signal x(n), x(n) is the original input signal in the time domain, and e is a complex exponential function.

[0014] The frequency domain signal is reversely converted into the time domain signal, and the conversion formula is as follows: Among them, x(n) is the final time domain signal, X(k) is the processed frequency domain signal, e is the conjugate form of the complex exponential function, and 1 / N is the normalization factor.

[0015] Compared with the prior art, the present invention has the following beneficial effects: The accelerometer driving and control method and system based on photoelectric signals enhances the noise suppression capability and automatically adjusts the filtering parameters according to the current environmental conditions to achieve the best noise suppression effect. The noise with relatively stable characteristics is continuously detected. After reaching a certain frequency, it is recorded in a specific noise database. The noise with relatively stable characteristics in the original signal is detected by an algorithm, and compared with the specific noise database. The noise in the specific noise database is removed, the calculation burden is reduced, the real-time response capability is improved, and data processing delays caused by complex signals are avoided. According to the amount of data, the calculation time of the time domain convolution algorithm and the sliding average filtering algorithm when processing the same data is judged, and the algorithm with shorter time is selected. When the amount of data is lower than the threshold, the sliding average filtering algorithm is selected. When the amount of data is higher than the threshold, the time domain convolution algorithm is selected to convert the time domain signal into the frequency domain signal. The frequency domain signal is processed by the time domain convolution algorithm. After the calculation is completed, the frequency domain signal is reversely converted into the time domain signal to reduce the time required for data processing. At the same time, the stopping conditions of the algorithm iteration process are simplified so that the algorithm ends the loop under the premise of meeting the accuracy requirements and the number of iterations, further reducing the time required for data processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the principle structure of the present invention; Figure 2 This is a schematic diagram of the data removal principle structure of the present invention; Figure 3 This is a schematic diagram of the purification signal principle structure of the present invention; Figure 4 It is a schematic diagram of the principle structure of removing relatively stable noise in the present invention; Figure 5 It is a schematic diagram of the noise suppression principle structure of the present invention; Figure 6 It is a schematic diagram of the noise processing principle structure of the present invention; Figure 7 It is a simplified schematic diagram of the algorithm principle structure of the present invention. DETAILED DESCRIPTION

[0017] 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.

[0018] In the present application, for ease of understanding, the method steps used do not need to be executed in the order of the steps in this embodiment during actual operation. In other embodiments, these steps may be performed simultaneously or in a different order.

[0019] Embodiment 1: like Figure 1-Figure 7 As shown, the present invention provides a technical solution: a method and system for driving and controlling an accelerometer based on a photoelectric signal, the method comprising: Obtain the signal to be processed and the original signal collected by the photoelectric sensor; Identify and remove data, identify and process the input original signal and remove irrelevant data; Enhance noise suppression capability, estimate the system status through noisy data, and adjust the filter parameters according to the current environmental conditions to achieve the best noise suppression effect; Pure useful signal, using the reference signal to estimate and remove the noise components in the main signal to obtain a pure useful signal; Collect iterative characteristic stable noise, continuously detect relatively stable noise, and enter it into the specific noise database after reaching a certain frequency; Remove noise with stable characteristics. Use an algorithm to detect relatively stable noise in the original signal, compare it with a specific noise database, and remove the noise in the specific noise database. Algorithm selection: Based on the amount of data, determine the calculation time of the time domain convolution algorithm and the moving average filter algorithm when processing the same data, and select the algorithm with shorter calculation time. The specific steps are as follows: S1: Set the threshold. When the data volume is lower than the threshold, select the sliding average filtering algorithm. S2: Set a threshold. When the data volume is higher than the threshold, select a time domain convolution algorithm to convert the time domain signal into a frequency domain signal. S3: Use the time domain convolution algorithm to process the frequency domain signal. The algorithm formula is as follows: Among them, Y(k) is the frequency domain representation of the output signal, X(k) is the frequency domain representation of the input signal, and H(k) is the frequency domain representation of the filter; Bring in data: N=4; X(0) = 1 (X(0) can be any real number, let’s assume it is 1 as an example); X(1)=-2+2j (derived from X(3)=X(1)); H(k)=1; Now, we can calculate Y(k): Y(0)=X(0)×H(0); =1×1; =1; Y(1)=X(1)×H(1); =(-2+2j)×1; =-2+2j; Y(2)=X(2)×H(2); =(-6j)×1; =-6j; Y(3)=X(3)×H(3); =(-2-2j)×1; =-2-2j; Therefore, a frequency domain representation of the output signal is obtained.

[0020] S4: After the calculation is completed, the frequency domain signal is reversely converted into a time domain signal; Simplify the algorithm iteration and the stopping conditions of the algorithm iteration process so that the algorithm can end the loop while meeting the accuracy requirements and the number of iterations.

[0021] The identification and removal of data is to perform signal preprocessing on the input original signal, remove the non-correlated parts through detrending processing, cut out the irrelevant parts in the signal, and reduce the amount of calculation.

[0022] It should be noted that the input raw signal is preprocessed to identify and remove non-relevant data. The linear or nonlinear trend components in the signal are eliminated through detrending processing, and then these trend components are subtracted from the original signal. The signal is analyzed to determine and trim off parts that are obviously irrelevant to the target analysis, such as silent segments, background noise, or invalid data within a specific time period. This step helps to reduce the amount and complexity of data for subsequent calculations. In addition, the window function is applied at both ends of the signal, which makes the signal edge transition to zero more smoothly, thereby effectively reducing spectral leakage, significantly reducing unnecessary computational burden, and improving the quality of subsequent signal processing and analysis, ensuring that the final result is more accurate and reliable, which can greatly improve work efficiency and analysis accuracy.

[0023] The enhanced noise suppression capability sets an initial state estimate and its covariance matrix, predicts the state at the current moment based on the state estimate at the previous moment, corrects the prediction value through signal data, adjusts the filter parameters, and dynamically suppresses noise.

[0024] It should be noted that based on the state estimation value at the previous moment, the state at the current moment is predicted by the system model, and the predicted value is corrected by the actual measurement data to obtain a more accurate state estimation. The specific steps are: set the initial state and its covariance matrix, and then at each time step, use the state transition model to predict the current state and its covariance, and then calculate the gain and update the state estimate and covariance based on the actual measurement value. In addition, the noise level in the current environment is evaluated based on the real-time collected data, and the process noise covariance and measurement noise covariance are dynamically adjusted to optimize the filtering effect and achieve efficient noise suppression, thereby reducing noise interference and improving signal quality.

[0025] The steps of pure useful signal are as follows: L1: collect the main signal, which is composed of useful signal and noise signal; L2: Set the reference signal. The reference signal calculation formula is as follows: Bringing in the data to perform the calculation, there are two simple discrete-time reference signals and filters: r1(n)={1,0,0,0} (a unit pulse signal); r2(n)={1,1,1,1} (a constant signal); h(n) = {21,21} (a simple averaging filter); For r1(n) and h(n), we can calculate: y1(0)=1×21=21; y1(1)=0×21+1×21=21; y1(2)=0 for n>1; Where r is the i-th reference signal, h(t) is the filter used to adjust the i-th reference signal, y(t) is the adjusted i-th reference signal, and Fori is the value range of the index variable i; L3: All adjusted reference signals are weighted and summed to obtain an estimate of the noise signal. The formula is as follows:

[0026] Bring in data, w1=1, y1(0)=21; y1(1)=21; n(0)=w1y1(0); =1×2; =2; n(1)=w1y1(1); =1×2; =2; Wherein, w is the weight coefficient of the i-th adjusted reference signal, and n(t) is the estimated value of the noise signal; L4: Useful signal extraction, subtract the estimated noise signal from the main signal to obtain the estimated value of the pure useful signal. The calculation formula is as follows:

[0027] Among them, S(t) is the estimated value of the pure useful signal, x(t) is the main signal, and n(t) is the estimated value of the noise signal; Bring in data, n(0)=2; n(1)=2; x(0)=5; x(1)=7; Now, we can calculate s(t): s(0)=x(0)-n(0); =5-2; =3; s(1)=x(1)-n(1); =7-2; =5; Therefore, at time points t=0 and t=1, the estimated values ​​of the pure useful signal are s(0)=3 and s(1)=5 respectively.

[0028] The method collects stable noise with iterative characteristics and continuously detects noise with relatively stable characteristics. When the same noise with relatively stable characteristics is detected a set number of times, it is entered into a specific noise database for spectrum subtraction data model calculation to reduce the calculation burden.

[0029] The algorithm is used to remove the noise with relatively stable characteristics in the main signal. The algorithm formula is as follows:

[0030] Among them, S(f) represents the estimated value of the spectrum of the pure useful signal, which is the result after processing, X(f) represents the spectrum of the main signal, α represents the adjustment factor, which is a scalar value used to control the intensity of spectrum subtraction, and N(f) represents the estimated value of the noise spectrum; Bring in the following data: The value of the spectrum X(f) of the main signal at a certain frequency point f0 is X(f0)=10; The estimated value N(f) of the noise spectrum at the same frequency point f0 is N(f0)=2; Adjustment factor α=0.5.

[0031] Substituting these values ​​into the formula, we get: S(f0)=max{10-0.5×2}; S(f0)=max{10-1}; S(f0)=max{9}; S(f0)=9; Therefore, at the frequency point f0, the estimated value S(f0) of the spectrum of the pure useful signal is 9.

[0032] When the data volume is lower than the threshold, the sliding average filtering algorithm is selected, and the algorithm formula is as follows: Where: y(n) is the output signal, that is, the signal after sliding average filtering, 1 / M represents the average coefficient of the signal samples in the sliding window, ∑ M-1 K=0 represents a summation operation, x represents an input signal, and (nk) represents a time index or a sample index of the input signal; Suppose we have the following sample values ​​of the input signal x(n): x(0)=1; x(1)=2; x(2)=3; x(3)=4; x(4)=5; And we choose a sliding window size M=3; However, to keep the example clear, we start counting from n=2 and continue counting until n=4; When n=2: y(2)=1 / 3[x(2)+x(1)+x(0)]; y(2)=1 / 3[3+2+1]; y(2)=1 / 3×6; y(2)=2; When n=3: y(3)=1 / 3[x(3)+x(2)+x(1)]; y(3)=1 / 3[4+3+2]; y(3)=1 / 3×9; y(3)=3; When n=4: y(4)=1 / 3[x(4)+x(3)+x(2)]; y(4)=1 / 3[5+4+3]; y(4)=1 / 3×12; y(4)=4; Thus, the output signal is obtained.

[0033] When the data volume is higher than the threshold, the time domain convolution algorithm is selected to convert the time domain signal to the frequency domain signal. The conversion formula is as follows: Where: X(k) is the frequency domain representation of the input signal x(n), x(n) is the original input signal in the time domain, and e is a complex exponential function; Substitute the sample value of the time domain signal x(n): x(0)=1; x(1)=2; x(2)=3; x(3)=4; Sample size N=4; Calculate the sample value of the frequency domain signal X(k); When k=0: X(0)=x(0)+x(1)+x(2)+x(3); X(0)=1+2+3+4; X(0)=10; (Note: When k = 0, the complex exponential term is 1, so the sum is actually the sum of all sample values ​​of the time domain signal.) When k=1: X(1)=1×(1)+2×(-j)+3×(-1)+4×(j); X(1)=1-2j-3+4j; X(1)=-2+2j; (Note: Here we use properties of complex exponential functions, specifically e / 2=-j and e / 2=j.) Similarly, calculate the value of X(k) when k=2 and k=3. Due to the symmetry of DFT (for real input signals), X(Nk) is the conjugate complex number of X(k) (that is, the real part is the same and the imaginary part has the opposite sign). Therefore, if X(1) has been calculated, then X(3) is the conjugate complex number of X(1). Similarly, X(2) will be a pure imaginary number (for real input signals and DFT of even length), and the absolute value of its imaginary part is related to X(0) and X(N / 2).

[0034] The calculation results of X(2) and X(3) are given (using the same complex exponential function properties): When k=2: X(2)=-6j; When k=3: ; here represents the complex conjugate of X(1).

[0035] The frequency domain signal is reversely converted into the time domain signal, and the conversion formula is as follows: Among them, x(n) is the final time domain signal, X(k) is the processed frequency domain signal, e is the conjugate form of the complex exponential function, and 1 / N is the normalization factor; Bring in data, Y(0)=1; Y(1)=-2+2j; Y(2)=-6j; Y(3)=-2-2j; N = 4 (because there are four frequency domain samples); Now, plug this data into the IDFT formula to calculate the time domain signal x(n): For n=0: x(0)=1 / 4[Y(0)+Y(1)+Y(2)+Y(3)]; x(0)=1 / 4[1+(-2+2j)+(-6j)+(-2-2j)]; x(0)=1 / 4[-3-6j]; x(0)=-3 / 4-3 / 2j; For n=1: x(1)=1 / 4[Y(0)×e+Y(1)×ej+Y(2)×e+Y(3)×e]; x(1)=1 / 4[1+(-2+2j)×i+(-6j)×(-1)+(-2-2j)×(-i)]; (Here i=-1 is the imaginary unit); x(1)=1 / 4[1-2i+2j+6j+2i+2j]; x(1)=1 / 4[1+8j]; x(1)=1 / 4+2j; For n=2: x(2)=1 / 4[1-2-6j+2]; = -(3 / 2)j; For n=3: x(3)=1 / 4[1+2i-6j-2-2i]; = -3 / 4 - (3 / 2) j; The final time domain signal is calculated.

[0036] Embodiment 2: like Figure 1-Figure 7 As shown, the present invention provides a technical solution: a method and system for driving and controlling an accelerometer based on a photoelectric signal, the method comprising: Obtain the signal to be processed and the original signal collected by the photoelectric sensor; Identify and remove data, identify and process the input original signal and remove irrelevant data; Enhance noise suppression capability, estimate the system status through noisy data, and adjust the filter parameters according to the current environmental conditions to achieve the best noise suppression effect; Pure useful signal, using the reference signal to estimate and remove the noise components in the main signal to obtain a pure useful signal; Collect iterative characteristic stable noise, continuously detect relatively stable noise, and enter it into the specific noise database after reaching a certain frequency; Remove noise with stable characteristics. Use an algorithm to detect relatively stable noise in the original signal, compare it with a specific noise database, and remove the noise in the specific noise database. Algorithm selection: Based on the amount of data, determine the calculation time of the time domain convolution algorithm and the moving average filter algorithm when processing the same data, and select the algorithm with shorter calculation time. The specific steps are as follows: S1: Select the time domain convolution algorithm to convert the time domain signal into the frequency domain signal; S2: Use the time domain convolution algorithm to process the frequency domain signal. The algorithm formula is as follows: Among them, Y(k) is the frequency domain representation of the output signal, X(k) is the frequency domain representation of the input signal, and H(k) is the frequency domain representation of the filter; Bring in data: N=4; x(0) = 1 (x(0) can be any real number, let’s assume it is 1 as an example); x(1)=-2+2j (derived from x(3)=x(1)); H(k)=1; Now, we can calculate Y(k): Y(0)=x(0)×H(0); =1×1; =1; Y(1)=x(1)×H(1); =(-2+2j)×1; =-2+2j; Y(2)=x(2)×H(2); =(-6j)×1; =-6j; Y(3)=x(3)×H(3); =(-2-2j)×1; =-2-2j; Therefore, a frequency domain representation of the output signal is obtained.

[0037] S4: After the calculation is completed, the frequency domain signal is reversely converted into a time domain signal; Simplify algorithm iteration and the stopping conditions of the algorithm iteration process so that the algorithm can end the loop while meeting the accuracy requirements and the number of iterations.

[0038] The identification and removal of data is to perform signal preprocessing on the input original signal, remove the non-correlated parts through detrending processing, cut out the irrelevant parts in the signal, and reduce the amount of calculation.

[0039] It should be noted that the input raw signal is preprocessed to identify and remove non-relevant data. The linear or nonlinear trend components in the signal are eliminated through detrending processing, and then these trend components are subtracted from the original signal. The signal is analyzed to determine and trim off parts that are obviously irrelevant to the target analysis, such as silent segments, background noise, or invalid data within a specific time period. This step helps to reduce the amount and complexity of data for subsequent calculations. In addition, the window function is applied at both ends of the signal, which makes the signal edge transition to zero more smoothly, thereby effectively reducing spectral leakage, significantly reducing unnecessary computational burden, and improving the quality of subsequent signal processing and analysis, ensuring that the final result is more accurate and reliable, which can greatly improve work efficiency and analysis accuracy.

[0040] The enhanced noise suppression capability sets an initial state estimate and its covariance matrix, predicts the state at the current moment based on the state estimate at the previous moment, corrects the prediction value through signal data, adjusts the filter parameters, and dynamically suppresses noise.

[0041] It should be noted that based on the state estimation value at the previous moment, the state at the current moment is predicted by the system model, and the predicted value is corrected by the actual measurement data to obtain a more accurate state estimation. The specific steps are: set the initial state and its covariance matrix, and then at each time step, use the state transition model to predict the current state and its covariance, and then calculate the gain and update the state estimate and covariance based on the actual measurement value. In addition, the noise level in the current environment is evaluated based on the real-time collected data, and the process noise covariance and measurement noise covariance are dynamically adjusted to optimize the filtering effect and achieve efficient noise suppression, thereby reducing noise interference and improving signal quality.

[0042] The steps of pure useful signal are as follows: L1: collect the main signal, which is composed of useful signal and noise signal; L2: Set the reference signal. The reference signal calculation formula is as follows: Bringing in the data to perform the calculation, there are two simple discrete-time reference signals and filters: r1(n)={1,0,0,0} (a unit pulse signal); r2(n)={1,1,1,1} (a constant signal); h(n) = {21,21} (a simple averaging filter); For r1(n) and h(n), we can calculate: y1(0)=1×21=21; y1(1)=0×21+1×21=21; y1(2)=0 for n>1; Where r is the i-th reference signal, h(t) is the filter used to adjust the i-th reference signal, y(t) is the adjusted i-th reference signal, and Fori is the value range of the index variable i; L3: All adjusted reference signals are weighted and summed to obtain an estimate of the noise signal. The formula is as follows:

[0043] Bring in data, w1=1, y1(0)=21; y1(1)=21; n(0)=w1y1(0); =1×2; =2; n(1)=w1y1(1); =1×2; =2; Wherein, w is the weight coefficient of the i-th adjusted reference signal, and n(t) is the estimated value of the noise signal; L4: Useful signal extraction, subtract the estimated noise signal from the main signal to obtain the estimated value of the pure useful signal. The calculation formula is as follows:

[0044] Among them, S(t) is the estimated value of the pure useful signal, x(t) is the main signal, and n(t) is the estimated value of the noise signal; Bring in data, n(0)=2; n(1)=2; x(0)=5; x(1)=7; Now, we can calculate s(t): s(0)=x(0)-n(0); =5-2; =3; s(1)=x(1)-n(1); =7-2; =5; Therefore, at time points t=0 and t=1, the estimated values ​​of the pure useful signal are s(0)=3 and s(1)=5 respectively.

[0045] The method collects stable noise with iterative characteristics and continuously detects noise with relatively stable characteristics. When the same noise with relatively stable characteristics is detected a set number of times, it is entered into a specific noise database for spectrum subtraction data model calculation to reduce the calculation burden.

[0046] The algorithm is used to remove the noise with relatively stable characteristics in the main signal. The algorithm formula is as follows: Where S(f) represents the estimated value of the spectrum of the pure useful signal, X(f) represents the spectrum of the main signal, α represents the adjustment factor, and N(f) represents the estimated value of the noise spectrum; Bring in the following data: The value of the spectrum X(f) of the main signal at a certain frequency point f0 is X(f0)=10; The estimated value N(f) of the noise spectrum at the same frequency point f0 is N(f0)=2; Adjustment factor α=0.5.

[0047] Substituting these values ​​into the formula, we get: S(f0)=max{10-0.5×2}; S(f0)=max{10-1}; S(f0)=max{9}; S(f0)=9; Therefore, at the frequency point f0, the estimated value S(f0) of the spectrum of the pure useful signal is 9.

[0048] The time domain convolution algorithm is selected to convert the time domain signal into the frequency domain signal. The conversion formula is as follows: Where: X(k) is the frequency domain representation of the input signal x(n), x(n) is the original input signal in the time domain, and e is a complex exponential function; Substitute the sample value of the time domain signal x(n): x(0)=1; x(1)=2; x(2)=3; x(3)=4; Sample size N=4; Calculate the sample value of the frequency domain signal X(k); When k=0: X(0)=x(0)+x(1)+x(2)+x(3); X(0)=1+2+3+4; X(0)=10; (Note: When k = 0, the complex exponential term is 1, so the sum is actually the sum of all sample values ​​of the time domain signal.) When k=1: X(1)=1×(1)+2×(-j)+3×(-1)+4×(j); X(1)=1-2j-3+4j; X(1)=-2+2j; (Note: Here we use properties of complex exponential functions, specifically e / 2=-j and e / 2=j.) Similarly, calculate the value of X(k) when k=2 and k=3. Due to the symmetry of DFT (for real input signals), X(Nk) is the conjugate complex number of X(k) (that is, the real part is the same and the imaginary part has the opposite sign). Therefore, if X(1) has been calculated, then X(3) is the conjugate complex number of X(1). Similarly, X(2) will be a pure imaginary number (for real input signals and DFT of even length), and the absolute value of its imaginary part is related to X(0) and X(N / 2).

[0049] The calculation results of X(2) and X(3) are given (using the same complex exponential function properties): When k=2: X(2)=-6j; When k=3: ; here represents the complex conjugate of X(1).

[0050] The frequency domain signal is reversely converted into the time domain signal, and the conversion formula is as follows: Among them, x(n) is the final time domain signal, X(k) is the processed frequency domain signal, e is the conjugate form of the complex exponential function, and 1 / N is the normalization factor; Bring in data, Y(0)=1; Y(1)=-2+2j; Y(2)=-6j; Y(3)=-2-2j; N = 4 (because there are four frequency domain samples); Now, plug this data into the IDFT formula to calculate the time domain signal x(n): For n=0: x(0)=1 / 4[Y(0)+Y(1)+Y(2)+Y(3)]; x(0)=1 / 4[1+(-2+2j)+(-6j)+(-2-2j)]; x(0)=1 / 4[-3-6j]; x(0)=-3 / 4-3 / 2j; For n=1: x(1)=1 / 4[Y(0)×e+Y(1)×ej+Y(2)×e+Y(3)×e]; x(1)=1 / 4[1+(-2+2j)×i+(-6j)×(-1)+(-2-2j)×(-i)]; (Here i=-1 is the imaginary unit); x(1)=1 / 4[1-2i+2j+6j+2i+2j]; x(1)=1 / 4[1+8j]; x(1)=1 / 4+2j; For n=2: x(2)=1 / 4[1-2-6j+2]; = -(3 / 2)j; For n=3: x(3)=1 / 4[1+2i-6j-2-2i]; = -3 / 4 - (3 / 2) j; The final time domain signal is calculated.

[0051] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is limited by the attached embodiments and their equivalents.

Claims

1. A method for driving and controlling an accelerometer based on a photoelectric signal, characterized in that: The method comprises: Obtain the signal to be processed and the original signal collected by the photoelectric sensor; Identify and remove data, identify and process the input original signal and remove irrelevant data; Enhance noise suppression capability, estimate the system status through noisy data, and adjust the filter parameters according to the current environmental conditions to achieve the best noise suppression effect; Pure useful signal, using the reference signal to estimate and remove the noise components in the main signal to obtain a pure useful signal; Collect iterative characteristic stable noise, continuously detect relatively stable noise, and enter it into the specific noise database after reaching a certain frequency; Remove noise with stable characteristics. Use an algorithm to detect relatively stable noise in the original signal, compare it with a specific noise database, and remove the noise in the specific noise database. Algorithm selection: Based on the amount of data, determine the calculation time of the time domain convolution algorithm and the moving average filter algorithm when processing the same data, and select the algorithm with shorter calculation time. The specific steps are as follows: S1: Set the threshold. When the data volume is lower than the threshold, select the sliding average filtering algorithm. S2: Set a threshold. When the data volume is higher than the threshold, select a time domain convolution algorithm to convert the time domain signal into a frequency domain signal. S3: Use the time domain convolution algorithm to process the frequency domain signal. The algorithm formula is as follows: Among them, Y(k) is the frequency domain representation of the output signal, X(k) is the frequency domain representation of the input signal, and H(k) is the frequency domain representation of the filter; S4: After the calculation is completed, the frequency domain signal is reversely converted into a time domain signal; Simplify algorithm iteration and the stopping conditions of the algorithm iteration process so that the algorithm can end the loop while meeting the accuracy requirements and the number of iterations.

2. The method for driving and controlling an accelerometer based on a photoelectric signal according to claim 1, characterized in that: The identification and removal of data is to perform signal preprocessing on the input original signal, remove the non-correlated parts through detrending processing, cut out the irrelevant parts in the signal, and reduce the amount of calculation.

3. The method for driving and controlling an accelerometer based on a photoelectric signal according to claim 1, characterized in that: The enhanced noise suppression capability sets an initial state estimate and its covariance matrix, predicts the state at the current moment based on the state estimate at the previous moment, corrects the prediction value through signal data, adjusts the filter parameters, and dynamically suppresses noise.

4. The method for driving and controlling an accelerometer based on a photoelectric signal according to claim 1, characterized in that: The steps of pure useful signal are as follows: L1: collect the main signal, which is composed of useful signal and noise signal; L2: Set the reference signal. The reference signal calculation formula is as follows: Where r is the i-th reference signal, h(t) is the filter used to adjust the i-th reference signal, y(t) is the adjusted i-th reference signal, and Fori is the value range of the index variable i; L3: All adjusted reference signals are weighted and summed to obtain an estimate of the noise signal. The formula is as follows: Wherein, w is the weight coefficient of the i-th adjusted reference signal, and n(t) is the estimated value of the noise signal; L4: Useful signal extraction, subtract the estimated noise signal from the main signal to obtain the estimated value of the pure useful signal. The calculation formula is as follows: Among them, S(t) is the estimated value of the pure useful signal, x(t) is the main signal, and n(t) is the estimated value of the noise signal.

5. The method for driving and controlling an accelerometer based on a photoelectric signal according to claim 1, characterized in that: The method collects stable noise with iterative characteristics and continuously detects noise with relatively stable characteristics. When the same noise with relatively stable characteristics is detected a set number of times, it is entered into a specific noise database for spectrum subtraction data model calculation to reduce the calculation burden.

6. The method for driving and controlling an accelerometer based on a photoelectric signal according to claim 1, characterized in that: The algorithm is used to remove the noise with relatively stable characteristics in the main signal. The algorithm formula is as follows: Wherein, S(f) represents the estimated value of the spectrum of the pure useful signal, X(f) represents the spectrum of the main signal, α represents the adjustment factor, and N(f) represents the estimated value of the noise spectrum.

7. The method for driving and controlling an accelerometer based on a photoelectric signal according to claim 1, characterized in that: When the data volume is lower than the threshold, the sliding average filtering algorithm is selected, and the algorithm formula is as follows: Where: y(n) is the output signal, that is, the signal after sliding average filtering, M1 represents the average coefficient of the signal samples in the sliding window, ∑ M-1 K=0 represents a summation operation, x represents the input signal, and (nk) represents the time index or sample index of the input signal.

8. The method for driving and controlling an accelerometer based on a photoelectric signal according to claim 1, characterized in that: When the data volume is higher than the threshold, the time domain convolution algorithm is selected to convert the time domain signal to the frequency domain signal. The conversion formula is as follows: Where: X(k) is the frequency domain representation of the input signal x(n), x(n) is the original input signal in the time domain, and e is a complex exponential function.

9. The method for driving and controlling an accelerometer based on a photoelectric signal according to claim 1, characterized in that: The frequency domain signal is reversely converted into the time domain signal, and the conversion formula is as follows: Among them, x(n) is the final time domain signal, X(k) is the processed frequency domain signal, e is the conjugate form of the complex exponential function, and 1 / N is the normalization factor.

10. An accelerometer drive and control system based on photoelectric signals, characterized in that: An accelerometer driving and control method based on photoelectric signals as described in any one of claims 1 to 9 is used.

Images