Signal processing method and apparatus

Abstract

The application discloses a signal processing method and device, and relates to the field of electric data processing; wherein, the method comprises: collecting acceleration data of a first device by using an acceleration sensor to obtain first output data; the first device is a device where the acceleration sensor is located; performing wavelet threshold denoising processing on the first output data to obtain denoising data; and performing synchronous adaptive filtering on the denoising data based on the rotation speed of the first device to obtain second output data. The signal processing method and device provided in the application embodiment can improve the electromagnetic interference suppression effect of a sensor, thereby improving the measurement accuracy and reliability of the sensor.

Description

Technical Field

[0001] This application relates to the field of electrical data processing, and more particularly to a signal processing method and apparatus. Background Technology

[0002] With the increasing integration and standardization of operating frequencies in electronic devices, the electromagnetic environment during sensor data acquisition is becoming increasingly complex. Because the sensitive elements and signal processing circuits inside sensors are highly susceptible to external electromagnetic interference (EMI), leading to increased output signal noise, decreased accuracy, and even failure, EMI suppression is crucial for accelerating data processing. Current technologies can eliminate EMI by placing a shielding layer around the core sensor components (such as the sensor chip). However, effective suppression methods for EMI in the sensor's output data are lacking. Furthermore, since shielding layers inevitably have gaps (such as seams, openings, and cable penetrations), electromagnetic leakage is highly likely at these locations. Therefore, the EMI suppression effect is poor, resulting in lower sensor measurement accuracy and reliability.

[0003] Therefore, overcoming the aforementioned technical problems and defects has become a key issue that needs to be addressed. Summary of the Invention

[0004] To address the aforementioned technical problems, this application provides a signal processing method and apparatus that can improve electromagnetic interference suppression, thereby enhancing the measurement accuracy and reliability of sensors.

[0005] According to one aspect of this application, a signal processing method is provided, the method comprising:

[0006] Acceleration data from a first device is collected using an accelerometer to obtain first output data; the first device is the device where the accelerometer is located.

[0007] The first output data is subjected to wavelet threshold denoising to obtain denoised data.

[0008] Based on the rotational speed of the first device, the noise reduction data is subjected to synchronous adaptive filtering to obtain the second output data.

[0009] In the above scheme, the step of performing wavelet threshold denoising on the first output data to obtain denoised data includes:

[0010] The first output data is subjected to N-level db4 wavelet decomposition to obtain low-frequency approximation coefficients and high-frequency detail coefficients;

[0011] The low-frequency approximation coefficients and the high-frequency detail coefficients are subjected to inverse wavelet transform to generate denoised data.

[0012] The method in the above scheme further includes:

[0013] The number of decomposition levels N is determined based on information entropy; or, it is determined based on the statistical characteristics of noise.

[0014] The determination of the decomposition level N based on information entropy includes:

[0015] Calculate the information entropy of the high-frequency coefficients in the current layer;

[0016] Based on the information entropy ratio of the current layer's high-frequency coefficients to the previous layer's high-frequency coefficients, the entropy ratio of the current layer's high-frequency coefficients is obtained;

[0017] Determine whether the entropy ratio of the i-th layer satisfies the following formula:

[0018] ;

[0019] If satisfied, stop the calculation and let N = i-1; if not satisfied, continue to check if the next level is satisfied.

[0020] The determination of the decomposition level N based on noise statistical characteristics includes:

[0021] Assume that the high-frequency coefficients of the j-th layer follow a noise distribution;

[0022] Calculate the Kolmogorov-Smirnov statistic D for the high-frequency coefficients of the j-th layer. j ;

[0023] Judge D j If the value is higher than a preset significance level, the decomposition stops and N=j is set; otherwise, the decomposition continues.

[0024] In the above scheme, the step of synchronously adaptively filtering the noise reduction data based on the rotation speed of the first device to obtain the second output data includes:

[0025] Real-time rotational speed of the first device is collected.

[0026] Based on the real-time rotational speed, the cutoff frequency of the low-pass filter is dynamically adjusted to obtain the dynamic cutoff frequency; as the real-time rotational speed increases, the corresponding cutoff frequency also increases.

[0027] Based on the dynamic cutoff frequency and the real-time rotational speed sampling frequency, a bilinear transformation is performed, and the filter coefficients are dynamically updated based on the transformation results; when the dynamic cutoff frequency increases, the corresponding filter coefficients increase accordingly.

[0028] Based on the dynamically updated filter coefficients, a differential operation is performed on the denoised data to obtain the second output data.

[0029] In the above scheme, before performing synchronous adaptive filtering on the noise reduction data based on the rotational speed of the first device, the method further includes:

[0030] A self-heating mechanical model is established based on the relationship between ambient temperature, continuous working time and the self-heating change of the accelerometer.

[0031] Determine whether the current state of the first device is stationary or in motion;

[0032] When the current state is stationary, based on the current ambient temperature and the current continuous working time, the predicted temperature of the accelerometer is determined using the self-heating mechanical model, and the gain corresponding to the predicted temperature is determined based on the preset temperature-gain mapping relationship to obtain the target gain coefficient; or, when the current state is in motion, the gain determined in the previous stationary state is used as the target gain coefficient.

[0033] Temperature compensation is performed on the noise reduction data based on the target gain coefficient.

[0034] In the above scheme, the accelerometer includes a multi-layer nested isolation structure; the multi-layer nested isolation structure includes, from the inside to the outside, a sensor core, a first isolation layer, a first shielding layer, a second isolation layer and a second shielding layer, and a shielding cavity is formed between the first shielding layer and the second shielding layer, and the distance between the first shielding layer and the second shielding layer is configured according to the target frequency;

[0035] The method further includes:

[0036] Before using the accelerometer to collect acceleration data from the first device, the multi-layered nested isolation structure is used to perform initial suppression of external electromagnetic interference.

[0037] In the above scheme, the initial suppression of external electromagnetic interference using the multi-layer nested isolation structure includes:

[0038] The second shielding layer is used to perform the first reflection and absorption of the received external electromagnetic interference, and the induced interference current is discharged through single-point grounding to obtain the first interference signal;

[0039] Using the shielded cavity where the second isolation layer is located as a dielectric layer, the first interference signal is spatially attenuated and abnormal resonance is removed to obtain the second interference signal;

[0040] The first shielding layer is used to perform a second reflection and absorption process on the second interference signal to obtain a third interference signal;

[0041] The first isolation layer is used for vibration isolation and electrical insulation.

[0042] According to another aspect of this application, a signal processing apparatus is provided, the apparatus comprising:

[0043] An accelerometer is used to collect acceleration data from a first device to obtain first output data; the first device is the device where the accelerometer is located.

[0044] The processing unit is used to perform wavelet threshold denoising on the first output data to obtain denoised data; and to perform synchronous adaptive filtering on the denoised data based on the rotation speed of the first device to obtain second output data.

[0045] In the above scheme, the accelerometer includes a multi-layer nested isolation structure; the multi-layer nested isolation structure includes, from the inside out, a sensor core, a first isolation layer, a first shielding layer, a second isolation layer, and a second shielding layer, with a shielding cavity formed between the first shielding layer and the second shielding layer, and the distance between the first shielding layer and the second shielding layer is configured according to the target frequency; the multi-layer nested isolation structure is used to perform initial suppression processing on external electromagnetic interference before the accelerometer acquires acceleration data from the first device.

[0046] In the above scheme, the first isolation layer includes a sleeve gasket, a limiting ring, and an isolation plate. The sleeve gasket is located below the sensor core, the limiting ring is sleeved around the sensor core, and the isolation plate is located at the upper end of the sensor core and between the sensor core and the first shielding layer. The first isolation layer is tightly attached to and fixed to the outer wall of the sensor core.

[0047] The first shielding layer includes a sleeve and a pressure ring. The sleeve is sleeved outside the first isolation layer, and the pressure ring is located above the sensor core and inside the sleeve. The isolation sheet is located between the sensor core and the pressure ring. The pressure ring and the sleeve form a receiving cavity to accommodate the first isolation layer.

[0048] The second isolation layer includes an isolation cylinder, an upper isolation plate, and a lower isolation plate. The upper isolation plate and the lower isolation plate are located on the upper and lower sides of the isolation cylinder, respectively, and together with the isolation cylinder, form a receiving cavity for accommodating the first shielding layer.

[0049] The second shielding layer includes an outer shell and a cover plate, the outer shell and the cover plate forming an internal receiving cavity, and the second isolation layer is located within the internal receiving cavity;

[0050] The gaps on the surface of the first shielding layer and the gaps on the surface of the second shielding layer are staggered.

[0051] The signal processing method and apparatus provided in this application achieve preliminary noise reduction and preprocessing of the output data through wavelet threshold denoising, thereby effectively filtering out random interference in the output signal, providing a stable input signal for subsequent adaptive filtering, and ensuring the accuracy and reliability of the adaptive filtering results. Furthermore, by introducing an adaptive filtering algorithm, the threshold can be dynamically adjusted according to the signal characteristics to ensure that the filter passband is aligned with the useful signal frequency band, avoiding signal distortion, thereby improving the sensor measurement accuracy while ensuring the electromagnetic interference suppression effect.

[0052] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0053] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application, and do not constitute an undue limitation of this application.

[0054] Figure 1 This is a schematic flowchart of a signal processing method provided in an embodiment of this application;

[0055] Figure 2 This is a schematic diagram of the wavelet threshold denoising process in the signal processing method of this application embodiment;

[0056] Figure 3 This is a schematic flowchart of temperature compensation in the signal processing method of this application embodiment;

[0057] Figure 4 This is a flowchart illustrating the synchronous adaptive filtering process in the signal processing method of this application embodiment;

[0058] Figure 5 This is a schematic diagram of a signal processing device provided in an embodiment of this application;

[0059] Figure 6 This is a schematic diagram of a multi-layered nested isolation structure in the signal processing device of this application embodiment. Detailed Implementation

[0060] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0061] This application provides a signal processing method, such as... Figure 1As shown, the method may include steps S101 to S103. Steps S101 to S103 will be described in detail below with reference to specific embodiments.

[0062] S101: Acceleration data of the first device is collected using an accelerometer to obtain first output data; the first device is the device where the accelerometer is located.

[0063] In practical applications, an acceleration sensor can be installed in the first device and used to collect the acceleration status and specific acceleration parameters during the movement of the first device.

[0064] In practical applications, in order to further improve the accuracy and reliability of the accelerometer detection data, before performing interference suppression processing on the accelerometer output data using interference signal processing, the structure of the accelerometer can be improved first to perform preliminary suppression processing on external electromagnetic interference using physical structure.

[0065] Based on this, in one embodiment, the accelerometer may include a multi-layer nested isolation structure; the multi-layer nested isolation structure includes, from the inside out, a sensor core, a first isolation layer, a first shielding layer, a second isolation layer, and a second shielding layer, wherein the first shielding layer and the second shielding layer form a shielding cavity, and the distance between the first shielding layer and the second shielding layer is configured according to the target frequency; the method may further include:

[0066] Before using the accelerometer to collect acceleration data from the first device, the multi-layered nested isolation structure is used to perform initial suppression of external electromagnetic interference.

[0067] In practical applications, the initial interference suppression processing using a multi-layered nested isolation structure can serve as interference suppression at the physical or hardware level, while the interference suppression processing of the output data using a composite algorithm can serve as interference suppression at the algorithm or software level. Through the combination of the two, a dual interference suppression effect is achieved, effectively improving the effectiveness and accuracy of interference suppression.

[0068] In one embodiment, the initial suppression of external electromagnetic interference using the multi-layered nested isolation structure may include:

[0069] The second shielding layer is used to perform the first reflection and absorption of the received external electromagnetic interference, and the induced interference current is discharged through single-point grounding to obtain the first interference signal;

[0070] Using the shielded cavity where the second isolation layer is located as a dielectric layer, the first interference signal is spatially attenuated and abnormal resonance is removed to obtain the second interference signal;

[0071] The first shielding layer is used to perform a second reflection and absorption process on the second interference signal to obtain a third interference signal;

[0072] The first isolation layer is used for vibration isolation and electrical insulation.

[0073] In practical applications, the second shielding layer can also be called the outer shielding layer. As the first line of defense, the second shielding layer can remove most of the interference energy through reflection and absorption, and conduct away the induced interference current through grounding.

[0074] In practical applications, the second isolation layer can also be called the intermediate isolation layer; the second isolation layer, located between the two shielding layers, utilizes the distance between the two shielding layers to spatially attenuate the residual interference leaking from the outer shielding layer.

[0075] In practical applications, the distance between the two shielding layers can be configured according to the target frequency. For example, the distance can be configured to be one-quarter of the wavelength or greater than one-twentieth of the wavelength. By configuring the distance between the two shielding layers according to the target frequency, the resonance conditions in the shielding cavity can be effectively disrupted, thereby preventing harmful resonances from occurring in the shielding cavity and achieving effective interference suppression. Furthermore, due to the existence of the distance between the two shielding layers, when the electromagnetic waves leaking from the outer layer reach the inner layer, their energy has been greatly diffused and attenuated. Moreover, the signals leaking from the two shielding layers through small holes, gaps, etc., are regarded as two mismatched antennas. The two leakage events can partially cancel each other out in phase, thereby significantly reducing the total energy of the interference signal penetrating to the innermost layer.

[0076] In practical applications, the first shielding layer can also be called the inner shielding layer. As the second line of defense, the first shielding layer can reflect and absorb residual interference that has penetrated in. In practical applications, there may be defects such as seams, openings, and pores on the surface of the shielding layer. Here, the defect positions between the two shielding layers can be staggered. In this way, because the defects of the two shielding layers are staggered, electromagnetic waves cannot penetrate directly through a straight path and must undergo multiple reflections and attenuation. The leakage path is effectively blocked, which greatly improves the overall shielding effectiveness.

[0077] Here, the first shielding layer and the first isolation layer together achieve the first level of defense in the suppression process, suppressing vibration and common-mode interference, and completing the first electromagnetic shielding effect.

[0078] First, vibration isolation and electrical insulation are achieved using the first isolation layer. Specifically, the first isolation layer acts as a mechanical low-pass filter, absorbing and attenuating external high-frequency, large-amplitude mechanical vibration energy through its own damping characteristics, preventing it from being transmitted to the core sensing unit, avoiding vibration noise from drowning out the real acceleration signal, and achieving the effect of vibration suppression. At the same time, the electrical insulation cuts off the ground loop formed between external devices through the sensor housing, preventing the common-mode current caused by the ground potential difference from flowing into the sensor, fundamentally eliminating an important source of low-frequency interference, and achieving the effect of suppressing ground loop interference.

[0079] Second, electromagnetic shielding is achieved using a first shielding layer. Specifically, by setting up a first shielding layer, a specific "cover" is installed outside the first isolation layer to surround it, so that external electromagnetic interference cannot enter the first isolation layer and internal electromagnetic signals cannot leak out. In this way, the first shielding layer serves as the first line of electromagnetic defense. By utilizing the wave impedance mismatch characteristics of metal, electromagnetic waves radiated in space are reflected on the surface. When electromagnetic waves that partially penetrate the surface propagate inside the shielding material, they will generate heat energy (Joule heating) due to induced current, thus being consumed and attenuated, achieving the effects of reflection and absorption.

[0080] Here, the second shielding layer and the second isolation layer together realize the second defense layer in the suppression process, achieving structural decoupling and ultimate electromagnetic shielding.

[0081] First, the inner module carrying the core circuit is mechanically fixed and electrically isolated from the external conductive shell by an insulating support (such as plastic or ceramic) using a second isolation layer, achieving structural decoupling and insulation. The decoupling design prevents stress and minor deformations caused by temperature changes, impacts, or residual vibrations from being directly transmitted to the core chip, improving long-term stability and blocking stress and vibration transmission. The insulating layer forms a dielectric layer between the first and second shielding layers, providing an attenuation cavity. Residual electromagnetic waves penetrating the first shielding layer undergo spatial diffusion and attenuation within this cavity, with their field strength decreasing inversely with the square of the distance. Simultaneously, the reasonable spacing design disrupts cavity resonances that may form at specific frequencies, preventing interference from being amplified in the interlayer. Furthermore, the insulation provided by the second isolation layer ensures the "purity" of the internal circuit reference ground, preventing direct connection to the shell ground that may introduce noise, thus reinforcing the effect of the first isolation layer in cutting off ground loops.

[0082] Secondly, the second shielding layer, the outermost metal shell of the sensor, is used as a conductive shell to provide final protection for the entire sensor. Specifically, the second shielding layer, as the most robust barrier and structural main body, performs final and efficient reflection and absorption of interference that penetrates the inner defense line. Furthermore, due to its large size and relatively better conductivity, its intrinsic shielding effectiveness is also high. At the same time, the entire sensor structure can achieve good single-point grounding through this conductive shell, providing the final and only discharge path for all collected interference energy to be conducted to the ground.

[0083] In practical applications, the first shielding layer can be made by wrapping the sensor core with insulating material to eliminate vibration damping and electrical insulation, thereby achieving mechanical vibration isolation and electrical isolation; the first shielding layer can also be called the inner shielding layer.

[0084] In summary, when interference signals pass through this double-layer shielding structure, they are greatly attenuated through a series of actions involving "reflection-absorption-guide-distance attenuation-re-reflection-re-absorption," thus achieving excellent electromagnetic interference suppression. This makes it suitable not only for general measuring equipment but also for extremely demanding environments such as military, aerospace, medical, and high-precision measuring equipment.

[0085] In practical applications, the embodiments of this application introduce a multi-layer nested isolation structure, which enables the physical structure to suppress interference before the sensor outputs data, and the composite algorithm to suppress electromagnetic interference in the output data after the data is output, thereby achieving a dual suppression of electromagnetic interference, improving the electromagnetic interference suppression effect, and ensuring the measurement accuracy and reliability of the sensor.

[0086] S102: Perform wavelet threshold denoising on the first output data to obtain denoised data.

[0087] In one embodiment, such as Figure 2 As shown, the step of performing wavelet threshold denoising on the first output data to obtain denoised data, i.e., S102, may include:

[0088] S201: Perform N-level db4 wavelet decomposition on the first output data to obtain low-frequency approximation coefficients and high-frequency detail coefficients;

[0089] S202: Perform inverse wavelet transform on the low-frequency approximation coefficients and the high-frequency detail coefficients to generate noise-reduced data.

[0090] In practical applications, wavelet thresholding denoising first involves signal decomposition, then threshold calculation and quantization of the decomposed signal, and finally signal reconstruction.

[0091] In practical applications, when decomposing a signal, N-level wavelet decomposition can be performed to decompose the original signal into low-frequency approximation coefficients and high-frequency detail coefficients.

[0092] In practical applications, during threshold calculation and quantization, the noise standard deviation estimate can be expressed as:

[0093] ;

[0094] in, Indicates the standard deviation of noise. Represents the high-frequency coefficients of the first layer. Here, represents the Gaussian noise figure. It can be configured to 0.6;

[0095] The threshold for each layer can be determined using the following formula:

[0096] ;

[0097] in, Let represent the threshold of the i-th level wavelet decomposition, where i ≤ N; Determine whether the wavelet coefficients are mainly contributed by the useful signal or by noise;

[0098] After determining the threshold for each layer, the wavelet coefficients are subjected to soft thresholding; the soft thresholding function can be expressed by the following formula:

[0099] ;

[0100] in, This represents the wavelet coefficients after processing with the soft thresholding function. Represents the wavelet coefficients of the i-th layer. It is a sign function used to preserve the sign of the original coefficients. , ,when , ;

[0101] By using a soft thresholding function, when the absolute value of the original wavelet coefficients is greater than or equal to the corresponding threshold, they are shrunk (the threshold is subtracted); when the absolute value of the original wavelet coefficients is less than the corresponding threshold, they are set to 0. This can effectively preserve the characteristics of the useful signal while removing noise.

[0102] In practical applications, during signal reconstruction, low-frequency coefficients and high-frequency coefficients processed by soft thresholding can be used to perform inverse wavelet transform to output a denoised signal.

[0103] In practical applications, after the sensor's own double-layer shielding structure has been used to initially suppress electromagnetic interference, the residual interference in the output signal usually manifests as low-frequency drift, weak high-frequency noise, or non-stationary transient pulses. In order to accurately determine the appropriate number of wavelet decomposition layers to avoid over-decomposition leading to distortion of the effective signal or under-decomposition leading to noise residue, the number of wavelet decomposition layers N can be determined based on information entropy or noise statistical characteristics.

[0104] Based on this, in one embodiment, the method may further include:

[0105] The number of decomposition layers N can be determined based on information entropy; or, the number of decomposition layers N can be determined based on the statistical characteristics of noise.

[0106] In one embodiment, determining the decomposition level N based on information entropy may include:

[0107] Calculate the information entropy of the high-frequency coefficients in the current layer;

[0108] Based on the information entropy ratio of the current layer's high-frequency coefficients to the previous layer's high-frequency coefficients, the entropy ratio of the current layer's high-frequency coefficients is obtained;

[0109] Determine whether the entropy ratio of the i-th layer satisfies the following formula:

[0110] ;

[0111] If satisfied, stop the calculation and set N=i-1; if not satisfied, continue to check if the next layer is satisfied.

[0112] In practical applications, the information entropy of the high-frequency coefficients of the current layer can be calculated using the following formula:

[0113] ;

[0114] in, This represents the energy entropy of the i-th layer. This represents the k-th high-frequency coefficient in the i-th level decomposition. This represents the energy of the k-th high-frequency coefficient in the i-th layer. Let represent the total energy of all high-frequency coefficients in the i-th layer.

[0115] In one embodiment, determining the number of decomposition layers N based on noise statistical characteristics may include:

[0116] Assume that the high-frequency coefficients of the j-th layer follow a noise distribution;

[0117] Calculate the Kolmogorov-Smirnov statistic D for the high-frequency coefficients of the j-th layer. j ;

[0118] Judge D jIf the value is higher than a preset significance level, the decomposition stops and N=j is set; otherwise, the decomposition continues.

[0119] In practical applications, if we assume that the high-frequency coefficients of the j-th layer follow a noise distribution, we can assume that the high-frequency coefficients of the j-th layer follow a Gaussian distribution.

[0120] In practical applications, the significance level value can be pre-configured according to the actual application scenario. For example, it can be pre-configured as 0.01.

[0121] S103: Based on the rotation speed of the first device, perform synchronous adaptive filtering on the noise reduction data to obtain the second output data.

[0122] Here, a composite algorithm combining wavelet threshold denoising followed by synchronous adaptive filtering is used to achieve interference suppression processing of sensor output data at the software level.

[0123] In practical applications, in order to reduce the interference of temperature drift in synchronous adaptive filtering, temperature compensation can be performed on the noise reduction data before synchronous adaptive filtering.

[0124] Therefore, in one embodiment, before performing synchronous adaptive filtering on the noise reduction data based on the rotational speed of the first device, the method further includes temperature compensation on the noise reduction data.

[0125] In one embodiment, such as Figure 3 As shown, temperature compensation for the noise reduction data may include:

[0126] S301: Based on the relationship between ambient temperature, continuous working time and the self-heating change of the accelerometer, a self-heating mechanical model is established;

[0127] S302: Determine whether the current state of the first device is stationary or in motion;

[0128] S303: When the current state is stationary, based on the current ambient temperature and the current continuous working time, the predicted temperature of the accelerometer is determined using the self-heating mechanical model, and the gain corresponding to the predicted temperature is determined based on the preset temperature-gain mapping relationship to obtain the target gain coefficient.

[0129] S304: When the current state is in motion, use the gain determined in the previous static state as the target gain coefficient;

[0130] S305: Based on the target gain coefficient, perform temperature compensation on the noise reduction data.

[0131] In practical applications, a temperature-gain mapping relationship can be pre-constructed based on the temperature and performance error data of the same sample under different ambient temperatures.

[0132] In practical applications, by pre-constructing a temperature-gain mapping relationship, the influence of ambient temperature on the sensor's own temperature can be introduced, and the differences in the degree of influence of different ambient temperatures on the sensor's own performance can be determined. This can effectively correct the systematic errors caused by the temperature effect on the sensor's own performance and improve the accuracy of the sensor's output data.

[0133] In practical applications, as a low-frequency, slowly changing interference, temperature drift causes the signal baseline to fluctuate along the entire time axis. By first compensating for temperature and then using adaptive filtering, we can avoid the situation where temperature drift becomes part of the baseline drift due to adaptive filtering and is misjudged as part of the signal itself, thus preventing filtering deviation or convergence difficulties. This ensures the accuracy of subsequent adaptive filtering and improves the accuracy and reliability of the sensor output data.

[0134] In practical applications, the motion state of the first device can be determined by the data from the accelerometer.

[0135] In practical applications, the acceleration output is stable when the first device is stationary, while the output in motion contains the true dynamic acceleration. However, the true dynamic acceleration can easily be misinterpreted as zero-point drift or sensitivity change of the sensor. This embodiment of the application determines the target gain using a self-heating mechanical model when the first device is stationary, which enables fine calibration and obtains a high-precision reference value. This provides an accurate basis for compensation in motion. Furthermore, by using the target gain from the most recent stationary state in motion, the accuracy of temperature compensation can be avoided due to the introduction of dynamic errors. It can also effectively suppress parameter fluctuations during dynamic processes, ensuring the smoothness and continuity of the system output data, thereby guaranteeing the stability and continuity of sensor data output.

[0136] In practical applications, after temperature compensation is performed on the denoised data to eliminate the interference of temperature drift on adaptive filtering, synchronous adaptive filtering can be performed on the temperature-compensated denoised data.

[0137] In practical applications, the cutoff frequency of the filter can be dynamically adjusted based on the real-time rotational speed of the first device, thereby enabling synchronous adaptive filtering.

[0138] Based on this, in one embodiment, such as Figure 4 As shown, the step of synchronously adaptively filtering the noise reduction data based on the rotation speed of the first device to obtain the second output data, i.e., S103, may include:

[0139] S401: Real-time acquisition of the real-time rotational speed of the first device;

[0140] S402: Based on the real-time rotational speed, the cutoff frequency of the low-pass filter is dynamically adjusted to obtain a dynamic cutoff frequency; when the real-time rotational speed increases, the corresponding cutoff frequency increases accordingly.

[0141] S403: Based on the dynamic cutoff frequency and the real-time rotational speed sampling frequency, perform a bilinear transformation and dynamically update the filter coefficients based on the transformation result; when the dynamic cutoff frequency increases, the corresponding filter coefficients increase accordingly.

[0142] S404: Based on the dynamically updated filter coefficients, perform a differential operation on the noise-reduced data to obtain the second output data.

[0143] In practical applications, a speed-cutoff frequency correlation can be established in advance. After obtaining the real-time speed, the corresponding cutoff frequency can be obtained based on this correlation.

[0144] In practical applications, based on the dynamic cutoff frequency and the real-time rotational speed sampling frequency, a bilinear transformation is performed to convert the continuous s-domain transfer function into a discrete z-domain transfer function.

[0145] In practical applications, the transfer function in the s-domain It can be represented as:

[0146] ;

[0147] in, Indicates the dynamic cutoff frequency. denoted by the cutoff angular frequency, and s denotes the Laplace operator;

[0148] The bilinear transformation formula is expressed as:

[0149] ;

[0150] ;

[0151] in, Indicates the sampling period. Indicates the sampling frequency;

[0152] Substituting the bilinear transformation formula In this process, we obtain the difference equation:

[0153] ;

[0154] ;

[0155] ;

[0156] ;

[0157] in, , , These are the filter coefficients.

[0158] In practical applications, the updated filter coefficients are used to perform differential operations. This can be understood as follows: at each sampling moment, the dynamically updated filter coefficients are calculated based on the real-time rotational speed obtained from the sampling. Then, the updated filter coefficients are updated into the differential equation, so that the output signal is the one that has filtered out high-frequency noise and retains only the pure signal related to the current rotational speed, thus achieving the noise reduction effect.

[0159] In practical applications, by dynamically adjusting the cutoff frequency and filter coefficients based on the real-time rotation speed of the first device, the threshold can be dynamically adjusted according to the signal characteristics to ensure that the filter passband is aligned with the useful signal frequency band and avoid signal distortion. This ensures the electromagnetic interference suppression effect while improving the sensor measurement accuracy, achieving a balance between noise suppression and dynamic response.

[0160] In summary, the signal processing method provided in this application achieves preliminary noise reduction and preprocessing of the output data through wavelet threshold denoising, thereby effectively filtering out random interference in the output signal and providing a stable input signal for subsequent adaptive filtering, ensuring the accuracy and reliability of the adaptive filtering results. Furthermore, by introducing an adaptive filtering algorithm, the threshold can be dynamically adjusted according to the signal characteristics to ensure that the filter passband is aligned with the useful signal frequency band, avoiding signal distortion, thereby improving the sensor measurement accuracy while ensuring electromagnetic interference suppression. Furthermore, by setting a multi-layer nested isolation structure in the sensor, the cavity between the two shielding layers causes the propagation attenuation of the passing interference signal, achieving electromagnetic shielding, mechanical vibration isolation, and electrical isolation of the sensitive element, thereby achieving efficient suppression of broadband electromagnetic interference, ensuring the measurement accuracy and reliability of the sensor in harsh electromagnetic environments, and achieving hardware and software interference suppression through physical structure and composite algorithms respectively, realizing a dual suppression effect of electromagnetic interference, improving the electromagnetic interference suppression effect, and ensuring the measurement accuracy and reliability of the sensor.

[0161] To implement the signal processing method of this application, embodiments of this application also provide a signal processing apparatus, such as... Figure 5 As shown, the device may include an acceleration sensor 501 and a processing unit 502. The various parts of the device will be described in detail below with reference to specific embodiments.

[0162] Accelerometer 501 is used to collect acceleration data from the first device and obtain first output data; the first device is the device where the acceleration sensor is located.

[0163] The processing unit 502 is used to perform wavelet threshold denoising processing on the first output data to obtain denoised data; and to perform synchronous adaptive filtering on the denoised data based on the rotation speed of the first device to obtain second output data.

[0164] In one embodiment, the accelerometer 501 includes a multi-layered nested isolation structure; such as Figure 6 As shown, the multi-layer nested isolation structure includes, from the inside out, a sensor core 1, a first isolation layer, a first shielding layer, a second isolation layer, and a second shielding layer. The first shielding layer and the second shielding layer form a shielding cavity. The distance between the first shielding layer and the second shielding layer is configured according to the target frequency. The multi-layer nested isolation structure is used to perform initial suppression of external electromagnetic interference before the accelerometer collects acceleration data from the first device.

[0165] In practical applications, the accelerometer 501 uses a multi-layer nested isolation structure to achieve a first-order suppression effect and outputs the acceleration data after the first interference suppression, i.e., the first output data.

[0166] In practical applications, the first isolation layer can also be called the inner isolation layer, the second isolation layer can also be called the outer isolation layer, the first shielding layer can also be called the inner shielding layer, and the second shielding layer can also be called the outer shielding layer. This application does not limit these, as long as the function can be achieved.

[0167] In one embodiment, such as Figure 6 As shown, the first isolation layer includes a sleeve gasket 11, a limiting ring 12, and an isolation plate 13. The sleeve gasket 11 is located below the sensor core 1, the limiting ring 12 is sleeved around the sensor core 1, and the isolation plate 13 is disposed at the upper end of the sensor core 1 and located between the sensor core 1 and the first shielding layer. The first isolation layer is tightly attached to and fixed to the outer wall of the sensor core 1.

[0168] The first shielding layer includes a sleeve 21 and a pressure ring 22. The sleeve 21 is sleeved on the outside of the first isolation layer, and the pressure ring 22 is disposed above the sensor core 1 and located inside the sleeve 21. The isolation piece 13 is located between the sensor core 1 and the pressure ring 22. The pressure ring 22 and the sleeve 21 form a receiving cavity for accommodating the first isolation layer.

[0169] The second isolation layer includes an isolation cylinder 31, an upper isolation plate 32, and a lower isolation plate 33. The upper isolation plate 32 and the lower isolation plate 33 are located on the upper and lower sides of the isolation cylinder 31, respectively, and together with the isolation cylinder 31, form a receiving cavity for accommodating the first shielding layer.

[0170] The second shielding layer includes a housing 52 and a cover plate 51, the housing 52 and the cover plate 51 forming an internal receiving cavity, and the second isolation layer is located inside the internal receiving cavity.

[0171] In practical applications, the accelerometer may include an internal component, a first isolation layer, a first shielding layer, a second isolation layer, a circuit board 7, and an external component. The external component has an internal cavity with threads. The first isolation layer, the first shielding layer, the second isolation layer, the internal component, and the circuit board 7 are all located within the internal cavity of the external component. The internal component includes a sensor core 1, and the external component includes a housing 52, a pressure block 4, a cover plate 51, and a connector 6. The pressure block 4 is located inside the housing 52 and is threadedly connected to the housing 52 via internal threads. The cover plate 51 fits tightly with the housing 52 to form a conductive housing that accommodates the entire internal structure of the accelerometer. The conductive housing constitutes the second shielding layer of the accelerometer, i.e., the outer shielding layer. The connector 6 is embedded outside the housing 52 and fixedly connected to the housing 52. The circuit board 7 is fixedly mounted on the pressure block 4 and located between the second isolation layer and the second shielding layer. The input end of the circuit board 7 is connected to the sensor core 1 via internal wires, and the output end is connected to the input end of the connector 6 via internal wires.

[0172] In practical applications, in the first isolation layer, the sleeve gasket 11 and the isolation plate 13 are respectively tightly attached to the upper and lower surfaces of the sensor core 1, and the limiting ring 12 is fitted into the outer wall of the sensor core 1 to isolate and fix the sensor core 1; the first isolation layer is tightly attached to and fixed to the outer peripheral wall of the sensor core 1, providing basic vibration damping and electrical insulation; the sleeve gasket 11, the isolation ring 12 and the isolation plate 13 can be made of polyimide to improve the vibration isolation and electrical insulation effect.

[0173] In practical applications, in the first shielding layer, the inner bottom surface of the sleeve 21 is in contact with the sleeve gasket 11, and the side wall step of the sleeve 21 is in close contact with the limiting ring 12; the pressure ring 22 is in contact with the upper surface of the isolation sheet 13, so as to tightly wrap around the outer wall of the first isolation layer and form the first layer of electromagnetic shielding for the sensor core 1; the pressure ring 22 and the isolation sheet 13 can be made of TC4.

[0174] In practical applications, in the second isolation layer, the isolation cylinder 31 is fitted into the first shielding layer, and the upper isolation plate 32 and the lower isolation plate 33 are respectively attached to the upper and lower surfaces of the first shielding layer to form a second isolation layer, which is wrapped around the outside of the first shielding layer to mechanically decouple and electrically isolate the internal shielding layer from the external mechanism; the interior of the second shielding layer forms a cavity to accommodate the entire internal structure; the materials of the isolation cylinder 31, the upper isolation plate 32 and the lower isolation plate 33 can be polyimide.

[0175] Here, by directly wrapping the core with the first isolation layer, mechanical vibrations transmitted to the sensitive element can be effectively attenuated and primary electrical insulation can be provided. The second isolation layer decouples the inner shield from the external environment, preventing the shield from forming a ground loop due to contact with the external conductor and introducing new interference, thus achieving dual-function isolation. Furthermore, the first shield (inner shield) and the conductive shell (outer shield) form a dual electromagnetic shielding system, and the insulation layer between the two shielding layers ensures the independence between the inner and outer shields, avoiding damage to the shielding effectiveness due to poor connection at a single point, thus achieving a synergistic electromagnetic shielding effect. Furthermore, by innovatively implementing vibration isolation, electrical insulation, and electromagnetic shielding through a multi-layer nested structure, the complex problem of the intertwined influence of mechanical noise and electromagnetic noise in high-precision measurement is solved, thereby significantly improving the signal-to-noise ratio and measurement accuracy of the sensor in harsh environments.

[0176] Here, through the design of a multi-layer nested isolation structure, the interference signal undergoes a cascaded filtering process, which involves the first layer of reflection and absorption, propagation and attenuation in the cavity, and the second layer of reflection and absorption. In this way, even if a small amount of leakage signal penetrates the first layer of shielding, it will enter the isolation space between the two layers of shielding and undergo reflection and absorption again when it reaches the second layer of shielding, achieving multiple reflections and attenuation. At the same time, through this structure, the total shielding effectiveness (SE_total) is approximately equal to the sum of the shielding effectiveness of the two layers (SE1+SE2). In addition, the additional attenuation caused by the spacing makes the suppression effect on broadband interference, especially high-frequency interference, far exceed that of single-layer shielding, resulting in a significant interference suppression effect.

[0177] In practical applications, to further improve the shielding capability, the gaps between the two shielding layers can be staggered.

[0178] Based on this, in one embodiment, the gaps on the surface of the first shielding layer and the gaps on the surface of the second shielding layer are arranged in a staggered manner.

[0179] In practical applications, the surface gaps here can be gaps existing on the surface of the shielding layer, specifically, they can be openings, seams, cable penetration points, etc. Because the gaps of the two shielding layers are staggered, when electromagnetic waves leak from the gap of the second shielding layer, the gap corresponds to the complete surface of the first shielding layer, that is, the surface without gaps. In this way, the leaked electromagnetic waves will be blocked again by this complete surface and cannot directly enter the inner shielding layer. Instead, they need to be reflected and absorbed again, thereby effectively improving the shielding capability and thus improving the interference suppression capability.

[0180] In one embodiment, the processing unit 502 may specifically be used for:

[0181] The first output data is subjected to N-level db4 wavelet decomposition to obtain low-frequency approximation coefficients and high-frequency detail coefficients;

[0182] The low-frequency approximation coefficients and the high-frequency detail coefficients are subjected to inverse wavelet transform to generate denoised data.

[0183] In one embodiment, the processing unit 502 can also be used for:

[0184] The number of decomposition levels N is determined based on information entropy; or, it is determined based on the statistical characteristics of noise.

[0185] The determination of the decomposition level N based on information entropy includes:

[0186] Calculate the information entropy of the high-frequency coefficients in the current layer;

[0187] Based on the information entropy ratio of the current layer's high-frequency coefficients to the previous layer's high-frequency coefficients, the entropy ratio of the current layer's high-frequency coefficients is obtained;

[0188] Determine whether the entropy ratio of the i-th layer satisfies the following formula:

[0189] ;

[0190] If satisfied, stop the calculation and let N = i-1; if not satisfied, continue to check if the next level is satisfied.

[0191] The determination of the decomposition level N based on noise statistical characteristics includes:

[0192] Assume that the high-frequency coefficients of the j-th layer follow a noise distribution;

[0193] Calculate the Kolmogorov-Smirnov statistic D for the high-frequency coefficients of the j-th layer. j ;

[0194] Judge D j If the value is higher than a preset significance level, the decomposition stops and N=j is set; otherwise, the decomposition continues.

[0195] The processing unit 502 can be specifically used for:

[0196] Real-time rotational speed of the first device is collected.

[0197] Based on the real-time rotational speed, the cutoff frequency of the low-pass filter is dynamically adjusted to obtain the dynamic cutoff frequency; as the real-time rotational speed increases, the corresponding cutoff frequency also increases.

[0198] Based on the dynamic cutoff frequency and the real-time rotational speed sampling frequency, a bilinear transformation is performed, and the filter coefficients are dynamically updated based on the transformation results; when the dynamic cutoff frequency increases, the corresponding filter coefficients increase accordingly.

[0199] Based on the dynamically updated filter coefficients, a differential operation is performed on the denoised data to obtain the second output data.

[0200] In one embodiment, before performing synchronous adaptive filtering on the noise reduction data based on the rotational speed of the first device, the processing unit 502 may further be used to:

[0201] A self-heating mechanical model is established based on the relationship between ambient temperature, continuous working time and the self-heating change of the accelerometer.

[0202] Determine whether the current state of the first device is stationary or in motion;

[0203] When the current state is stationary, based on the current ambient temperature and the current continuous working time, the predicted temperature of the accelerometer is determined using the self-heating mechanical model, and the gain corresponding to the predicted temperature is determined based on the preset temperature-gain mapping relationship to obtain the target gain coefficient; or, when the current state is in motion, the gain determined in the previous stationary state is used as the target gain coefficient.

[0204] Temperature compensation is performed on the noise reduction data based on the target gain coefficient.

[0205] In one embodiment, the acceleration sensor 501 can also be used to: perform initial suppression of external electromagnetic interference using the multi-layer nested isolation structure.

[0206] It should be noted that the signal processing device provided in the above embodiments is only illustrated by the division of the above program modules. In practical applications, the above processing can be assigned to different program modules as needed, that is, the internal structure of the device can be divided into different program modules to complete all or part of the processing described above. In addition, the signal processing device and the signal processing method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here.

[0207] The effects of the embodiments of this application will be explained below with reference to specific experimental data.

[0208] The control group used an accelerometer and its output data that had a traditional single-layer metal shell shielding, without dedicated vibration isolation or internal circuit shielding.

[0209] The experimental group adopted the signal processing method and apparatus described in the embodiments of this application.

[0210] The electromagnetic interference suppression effect will be verified below.

[0211] Test environment: In an anechoic chamber, standard electromagnetic interference is applied to the sensor, including a radiated field immunity test with a frequency range of 30MHz-1GHz and a field strength of 10V / m; at the same time, a known reference mechanical vibration signal is applied using a vibration table.

[0212] Measurement metrics: Sensor output signal-to-noise ratio (SNR) or error voltage.

[0213] Table 1 shows the comparison results of electromagnetic interference suppression effects between the experimental group and the control group:

[0214] Table 1:

[0215] frequency Interference field strength Control group output error (mV) Experimental group output error (mV) Improved inhibition effect 100 MHz 10 V / m 150 5 30 times (approximately 30 dB) 400 MHz 10 V / m 320 8 40 times (approximately 32 dB) 1 GHz 10 V / m 280 10 28 times (approximately 29 dB)

[0216] As shown in Table 1, the output error of the experimental group was much smaller than that of the control group throughout the entire test frequency band, while the average shielding effectiveness was improved by about 30dB. The interference signal was attenuated to less than 1 / 30 of its original value, and the interference suppression effect was significantly improved.

[0217] The mechanical vibration isolation effect will be verified below.

[0218] Test conditions: Apply a 1000Hz high-frequency vibration noise to the outside of the sensor.

[0219] Measurement index: The amplitude of the 1000Hz frequency component in the sensor output signal (i.e., the amplitude of vibration noise).

[0220] Table 2 shows the comparison results of the mechanical vibration isolation effect between the experimental group and the control group:

[0221] Table 2:

[0222] structure Internally measured 1000Hz noise amplitude (mg) Vibration transmission attenuation Control group (rigid fixation) 50 mg benchmark Experimental group (with vibration isolation) 2 mg 28 times (approximately 29 dB)

[0223] As shown in Table 2, the vibration isolation layer attenuates external high-frequency vibration noise by approximately 29 dB, significantly improving the signal-to-noise ratio and measurement accuracy of the sensor in vibration environments.

[0224] The overall performance effect will be verified below.

[0225] Test conditions: In a quiet environment with absolute stillness and no electromagnetic interference, the output noise (ground noise) of the sensor itself is measured, which represents the minimum acceleration signal that it can detect.

[0226] Measurement results:

[0227] Control group background noise: 100 μg / √Hz;

[0228] Experimental group background noise: approximately 20 μg / √Hz;

[0229] The test results show that, due to the effective suppression of external electromagnetic interference and mutual interference between internal circuits, the background noise of the sensor in the experimental group was reduced by 5 times, proving that the measurement sensitivity and dynamic range of the structure provided in this application embodiment have been greatly improved, and it can detect weaker real acceleration signals.

[0230] The effects of the embodiments of this application will be explained below with reference to specific experimental data.

[0231] The control group used an accelerometer and its output data that had a traditional single-layer metal shell shielding, without dedicated vibration isolation or internal circuit shielding.

[0232] The experimental group adopted the signal processing method and apparatus described in the embodiments of this application.

[0233] The electromagnetic interference suppression effect will be verified below.

[0234] Test environment: In an anechoic chamber, standard electromagnetic interference is applied to the sensor, including a radiated field immunity test with a frequency range of 30MHz-1GHz and a field strength of 10V / m; at the same time, a known reference mechanical vibration signal is applied using a vibration table.

[0235] Measurement metrics: Sensor output signal-to-noise ratio (SNR) or error voltage.

[0236] Table 1 shows the comparison results of electromagnetic interference suppression effects between the experimental group and the control group:

[0237] Table 1:

[0238] frequency Interference field strength Control group output error (mV) Experimental group output error (mV) Improved inhibition effect 100 MHz 10 V / m 150 5 30 times (approximately 30 dB) 400 MHz 10 V / m 320 8 40 times (approximately 32 dB) 1 GHz 10 V / m 280 10 28 times (approximately 29 dB)

[0239] As shown in Table 1, the output error of the experimental group was much smaller than that of the control group throughout the entire test frequency band, while the average shielding effectiveness was improved by about 30dB. The interference signal was attenuated to less than 1 / 30 of its original value, and the interference suppression effect was significantly improved.

[0240] The mechanical vibration isolation effect will be verified below.

[0241] Test conditions: Apply a 1000Hz high-frequency vibration noise to the outside of the sensor.

[0242] Measurement index: The amplitude of the 1000Hz frequency component in the sensor output signal (i.e., the amplitude of vibration noise).

[0243] Table 2 shows the comparison results of the mechanical vibration isolation effect between the experimental group and the control group:

[0244] structure Internally measured 1000Hz noise amplitude (mg) Vibration transmission attenuation Control group (rigid fixation) 50 mg benchmark Experimental group (with vibration isolation) 2 mg 28 times (approximately 29 dB)

[0245] As shown in Table 2, the vibration isolation layer attenuates external high-frequency vibration noise by approximately 29 dB, significantly improving the signal-to-noise ratio and measurement accuracy of the sensor in vibration environments.

[0246] The overall performance effect will be verified below.

[0247] Test conditions: In a quiet environment with absolute stillness and no electromagnetic interference, the output noise (ground noise) of the sensor itself is measured, which represents the minimum acceleration signal that it can detect.

[0248] Measurement results:

[0249] Control group background noise: 100 μg / √Hz;

[0250] Experimental group background noise: approximately 20 μg / √Hz;

[0251] The test results show that, due to the effective suppression of external electromagnetic interference and mutual interference between internal circuits, the background noise of the sensor in the experimental group was reduced by 5 times, proving that the measurement sensitivity and dynamic range of the structure provided in this application embodiment have been greatly improved, and it can detect weaker real acceleration signals.

[0252] In summary, the signal processing method and apparatus provided in this application achieve preliminary noise reduction and preprocessing of the output data through wavelet threshold denoising, thereby effectively filtering out random interference in the output signal and providing a stable input signal for subsequent adaptive filtering, ensuring the accuracy and reliability of the adaptive filtering results. Furthermore, by introducing an adaptive filtering algorithm, the threshold can be dynamically adjusted according to the signal characteristics to ensure that the filter passband is aligned with the useful signal frequency band, avoiding signal distortion, thereby improving the sensor measurement accuracy while ensuring the electromagnetic interference suppression effect. Furthermore, by setting a multi-layer nested isolation structure in the sensor, the cavity between the two shielding layers causes the propagation attenuation of the passing interference signal, realizing electromagnetic shielding, mechanical vibration isolation, and electrical isolation of the sensitive element, thereby achieving efficient suppression of broadband electromagnetic interference, ensuring the measurement accuracy and reliability of the sensor in harsh electromagnetic environments, and achieving hardware and software interference suppression through physical structure and composite algorithm respectively, realizing a dual suppression effect of electromagnetic interference, improving the electromagnetic interference suppression effect, and ensuring the measurement accuracy and reliability of the sensor.

[0253] It should be noted that terms such as "first" and "second" are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0254] Furthermore, the technical solutions described in the embodiments of this application can be combined arbitrarily without conflict.

[0255] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application.

Claims

1. A signal processing method, characterized by, The method comprises: Collecting acceleration data of a first device by using an acceleration sensor to obtain first output data; the first device is a device where the acceleration sensor is located; Performing wavelet threshold denoising processing on the first output data to obtain denoised data; Performing synchronous adaptive filtering on the denoised data based on the rotation speed of the first device to obtain second output data; The wavelet threshold denoising processing on the first output data to obtain denoised data comprises: performing N-layer db4 wavelet decomposition on the first output data to obtain low-frequency approximation coefficients and high-frequency detail coefficients; and performing wavelet inverse transform on the low-frequency approximation coefficients and the high-frequency detail coefficients to generate denoised data; The method further comprises: determining the decomposition layer number N based on information entropy; or determining the decomposition layer number N based on noise statistical characteristics; wherein The determination of the decomposition layer number N based on information entropy comprises: Calculating the information entropy of the high-frequency coefficients of the current layer; Obtaining the entropy ratio of the high-frequency coefficients of the current layer based on the information entropy ratio of the high-frequency coefficients of the current layer and the high-frequency coefficients of the previous layer; Determining whether the entropy ratio of the i-th layer satisfies the following formula: ； If yes, stop calculation and set N=i-1; if no, continue to determine whether the next layer satisfies the formula; The determination of the decomposition layer number N based on noise statistical characteristics comprises: Assuming that the high-frequency coefficients of the j-th layer are subject to noise distribution; calculating the Kolmogorov-Smirnov statistic D for the jth layer of high frequency coefficients j ; Determination D j whether it is higher than a preset significant level value, if higher, stop decomposition, and let N = j; if not higher, continue decomposition.

2. The method of claim 1, wherein, The synchronous adaptive filtering on the denoised data based on the rotation speed of the first device to obtain second output data comprises: Real-time collecting the real-time rotation speed of the first device; Based on the real-time rotation speed, dynamically adjusting the cutoff frequency of the low-pass filter to obtain a dynamic cutoff frequency; when the real-time rotation speed increases, the corresponding cutoff frequency also increases; Based on the dynamic cutoff frequency and the real-time rotation speed sampling frequency, performing bilinear transformation, and dynamically updating the filter coefficients based on the transformation result; when the dynamic cutoff frequency increases, the corresponding filter coefficients also increase; Based on the dynamically updated filter coefficients, performing difference operation on the denoised data to obtain second output data.

3. The method of claim 2, wherein, Before the synchronous adaptive filtering on the denoised data based on the rotation speed of the first device, the method further comprises: Establishing a self-heating mechanics model based on the relationship between the ambient temperature, the continuous working time length and the self-heating change of the acceleration sensor; Determining whether the current state of the first device is static or dynamic; When the current state is static, determining the predicted temperature of the acceleration sensor based on the current ambient temperature and the current continuous working time length by using the self-heating mechanics model, and determining the gain corresponding to the predicted temperature based on a preset temperature-gain mapping relationship to obtain a target gain coefficient; or when the current state is dynamic, using the gain determined last time when the state is static as the target gain coefficient; Performing temperature compensation on the denoised data based on the target gain coefficient.

4. The method according to any one of claims 1 to 3, characterized in that, The acceleration sensor comprises a multi-layer nested isolation structure; the multi-layer nested isolation structure comprises, from inside to outside, a sensor core, a first isolation layer, a first shielding layer, a second isolation layer and a second shielding layer, a shielding cavity is formed between the first shielding layer and the second shielding layer, and the distance between the first shielding layer and the second shielding layer is configured according to a target frequency; The method further comprises: Before collecting acceleration data of the first device by using the acceleration sensor, the multi-layer nested isolation structure is used to perform primary suppression processing on external electromagnetic interference.

5. The method of claim 4, wherein, The primary suppression processing on external electromagnetic interference by using the multi-layer nested isolation structure comprises: The second shielding layer is used to perform first reflection and absorption processing on received external electromagnetic interference, and induced interference current is led out through single-point grounding to obtain a first interference signal; The shielding cavity in which the second isolation layer is located is used as a medium layer to perform spatial attenuation and abnormal resonance removal processing on the first interference signal to obtain a second interference signal; The first shielding layer is used to perform second reflection and absorption processing on the second interference signal to obtain a third interference signal; The first isolation layer is used for vibration isolation and electrical insulation.

6. A signal processing device, characterized by The device comprises: An acceleration sensor configured to collect acceleration data of a first device to obtain first output data; the first device is a device in which the acceleration sensor is located; A processing unit configured to perform wavelet threshold denoising processing on the first output data to obtain denoised data, and perform synchronous adaptive filtering on the denoised data based on a rotation speed of the first device to obtain second output data; The processing unit is configured to perform N-layer db4 wavelet decomposition on the first output data to obtain low-frequency approximation coefficients and high-frequency detail coefficients, and perform wavelet inverse transformation on the low-frequency approximation coefficients and the high-frequency detail coefficients to generate denoised data; The processing unit is further configured to determine the decomposition layer number N based on information entropy or based on noise statistical characteristics; wherein The determination of the decomposition layer number N based on information entropy comprises: calculating information entropy of a current layer of high-frequency coefficients; obtaining an entropy ratio of the current layer of high-frequency coefficients based on a ratio of information entropy of the current layer of high-frequency coefficients to information entropy of a previous layer of high-frequency coefficients; determining whether the entropy ratio of the i-th layer satisfies the following formula: ； if yes, stopping calculation and setting N=i-1; if no, continuing to determine whether a next layer satisfies the formula; The determination of the decomposition layer number N based on noise statistical characteristics comprises: assuming that the j-th layer of high-frequency coefficients is subject to noise distribution; calculating the Kolmogorov-Smirnov statistic D for the jth layer of high frequency coefficients j ; Determination D j whether it is higher than a preset significant level value, if higher, stop decomposition, and let N = j; if not higher, continue decomposition.

7. The apparatus of claim 6, wherein, The acceleration sensor comprises a multi-layer nested isolation structure; the multi-layer nested isolation structure comprises, from inside to outside, a sensor core, a first isolation layer, a first shielding layer, a second isolation layer and a second shielding layer, a shielding cavity is formed between the first shielding layer and the second shielding layer, and the distance between the first shielding layer and the second shielding layer is configured according to a target frequency; the multi-layer nested isolation structure is used to perform primary suppression processing on external electromagnetic interference before the acceleration sensor collects acceleration data of the first device.

8. The apparatus of claim 7, wherein, The first isolation layer comprises a sleeve gasket, a limiting ring and an isolation sheet, the sleeve gasket is located below the sensor core, the limiting ring is sleeved on the periphery of the sensor core, and the isolation sheet is arranged on the upper end of the sensor core and between the sensor core and the first shielding layer; the first isolation layer is closely attached to and fixed with the outer wall of the sensor core; The first shielding layer comprises a sleeve and a pressing ring, the sleeve is sleeved outside the first isolation layer, the pressing ring is arranged above the sensor core and in the sleeve, and the isolation sheet is located between the sensor core and the pressing ring; the pressing ring and the sleeve form a containing cavity containing the first isolation layer; The second isolation layer comprises an isolation cylinder, an isolation upper plate and an isolation lower plate, the isolation upper plate and the isolation lower plate are respectively located on the upper and lower sides of the isolation cylinder, and form a containing cavity containing the first shielding layer with the isolation cylinder; The second shielding layer comprises a shell and a cover plate, the shell and the cover plate form an internal containing cavity, and the second isolation layer is located in the internal containing cavity; The surface gap of the first shielding layer is arranged in a staggered manner with the surface gap of the second shielding layer.

Images