Pulse signal wireless transmission and calibration method based on cloud edge cooperation

Abstract

The invention discloses a pulse signal wireless transmission and calibration method based on cloud-edge cooperation, and relates to the technical field of signal transmission, and the method comprises the steps: an edge node collects reference noise in a transmission environment, generates estimated noise through an adaptive filter, carries out the denoising of a received pulse signal with noise, and carries out the amplitude distortion detection and preliminary correction; edge nodes exchange timestamps through a self-organizing network to realize initial time synchronization, and a cloud absolute time signal is combined, a local clock is finely adjusted by adopting a PI control algorithm, and time unification of the whole network is realized; the edge nodes detect the actual feature points of the pulse signals through a sliding window, align the actual feature points with the standard feature points, adjust window parameters, and perform local calibration in combination with local initial calibration parameters; the cloud end issues initial calibration parameters to the edge nodes, the edge nodes carry out secondary calibration and upload local optimization parameters, the cloud end carries out weighted fusion to generate global optimal parameters, iterative optimization is carried out until calibration errors are stable, and the real-time performance and accuracy of signals are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of signal transmission technology, and specifically to a pulse signal wireless transmission and calibration method based on cloud-edge collaboration. Background Art

[0002] In the battery management system (BMS) and motor control system of new energy vehicles, wireless transmission and calibration technology of pulse signals are crucial for real-time monitoring and control. At present, the mainstream pulse signal transmission methods are mainly divided into two categories: cloud-based centralized processing and edge node independent processing. The cloud-based centralized processing method uses a remote server to denoise, synchronize and calibrate the signal. Although it can utilize the powerful computing power of the cloud, the long data transmission link leads to high latency and poor real-time performance, which makes it difficult to meet the needs of high-speed dynamic detection of new energy vehicles. The edge node independent processing method relies on local algorithms for signal processing. Although the response speed is fast, it is limited by the insufficient coordination ability between nodes, making it difficult to achieve high-precision time synchronization and global parameter optimization, resulting in large signal calibration errors.

[0003] In the existing technology, noise elimination at edge nodes usually uses filters with fixed parameters, which cannot adapt to the dynamic noise changes in the complex electromagnetic environment of new energy vehicles, resulting in unstable denoising effect and affecting the accuracy of pulse signals. Time synchronization between nodes mainly relies on a single clock synchronization protocol, such as NTP or PTP. In an environment where wireless signals are susceptible to interference, synchronization accuracy is difficult to guarantee, which in turn affects the timing consistency of signal transmission. These problems are particularly prominent in the high-frequency pulse signal detection scenarios of new energy vehicles, which may lead to misjudgment of battery status or inaccurate motor control, affecting system safety and reliability.

[0004] In addition, existing calibration methods usually adopt centralized parameter distribution or local optimization strategies, lacking an iterative optimization mechanism for cloud-edge collaboration. The parameters distributed by the cloud are based on global statistics and are difficult to adapt to the local characteristics of different edge nodes; and the local optimization of edge nodes lacks a global perspective and is prone to falling into local optimality. This fragmented processing method leads to low calibration efficiency and slow error convergence, and cannot meet the requirements of new energy vehicle detection scenarios for high-precision, low-latency pulse signal transmission. Summary of the Invention

[0005] (1) Technical problems solved In response to the shortcomings of the existing technology, the present invention provides a pulse signal wireless transmission and calibration method based on cloud-edge collaboration. Through adaptive filtering denoising, dynamic time synchronization and cloud-edge collaborative iterative optimization, it solves the problems of high pulse signal transmission noise, low synchronization accuracy and poor calibration efficiency in the complex electromagnetic environment of new energy vehicles, and significantly improves the real-time and accuracy of the signal.

[0006] (2) Technical solution To achieve the above objectives, the present invention is implemented through the following technical solutions: a pulse signal wireless transmission and calibration method based on cloud-edge collaboration, comprising: The edge node collects reference noise in the transmission environment, generates estimated noise through an adaptive filter, denoises the received noisy pulse signal, and performs amplitude distortion detection and preliminary correction. Edge nodes achieve preliminary time synchronization by exchanging timestamps through a self-organizing network. Combined with the absolute time signal from the cloud, the PI control algorithm is used to fine-tune the local clock to achieve unified time across the entire network. The edge node dynamically detects the actual feature points of the pulse signal through a sliding window, aligns them with the pre-stored standard feature points, adjusts the window parameters, and performs local calibration based on the local initial calibration parameters; The cloud sends initial calibration parameters to the edge node, the edge node performs secondary calibration and uploads local optimization parameters, and the cloud performs weighted fusion to generate the global optimal parameters, and iterates the optimization until the calibration error stabilizes.

[0007] Furthermore, the edge node receives the noisy pulse signal through the wireless receiving module, and at the same time collects the reference noise from the same source as the noise in the pulse signal; the initial parameters of the adaptive filter are randomly set within a preset range, and the reference noise is filtered to generate estimated noise; the actual noise in the noisy pulse signal is extracted, the mean square error between the estimated noise and the actual noise is calculated, and the filter parameters are adjusted by the gradient descent method until the error is less than the preset error threshold.

[0008] Furthermore, the noisy pulse signal and the estimated noise are time-aligned and then a point-by-point subtraction operation is performed to obtain a denoised pulse signal; the amplitude distortion area of ​​the denoised pulse signal is detected, and a linear correction algorithm is used to perform preliminary amplitude correction so that the signal amplitude falls within the amplitude range of the standard pulse.

[0009] Furthermore, each edge node records its own local time through a local timing module; broadcasts a timestamp data packet containing the node ID and local time to all adjacent nodes in the network through a self-organizing network at a fixed period; receives the timestamp data packet of the adjacent node, calculates the preliminary deviation from its own local time, and uses it as the basis for correction after deducting the transmission time; collects the deviation values ​​of multiple adjacent nodes, removes the abnormal values ​​and takes the average value as the local clock correction value, adjusts the crystal oscillator frequency to correct the clock in real time; repeats the timestamp exchange and clock correction until the time difference between adjacent nodes is less than the preset time difference threshold.

[0010] Furthermore, the cloud server acts as the time master node and broadcasts a synchronization signal containing the absolute time and the broadcast time mark to all edge nodes at a fixed period. After receiving the synchronization signal broadcast by the cloud, the edge node parses the absolute time, calculates its deviation from the local current time, and uses the PI control algorithm to fine-tune the local clock phase.

[0011] Furthermore, the characteristic parameters of the standard pulse waveform are pre-stored in the edge node, including the peak position, the starting point of the rising edge and the end point of the falling edge; the initial sliding window size and the starting position are set according to the standard pulse duration and the sampling rate; the sliding window is moved to detect the actual pulse feature points, and the position deviation from the standard feature points is calculated, and the feature points are aligned by adjusting the window position.

[0012] Furthermore, if the actual pulse length deviates from the standard length by more than a preset length threshold, the window size is dynamically adjusted to ensure that the waveform is fully contained. Based on the aligned feature points, the local initial calibration parameters are called for phase and timing compensation to correct the phase deviation and time mark of the signal.

[0013] Furthermore, the cloud sends initial calibration parameters to each edge node through an encrypted channel. The parameters include phase calibration coefficient, timing delay correction value and amplitude compensation factor; the edge node uses the initial calibration parameters to perform secondary calibration on the locally calibrated pulse signal and calculates the root mean square error between the calibrated signal and the standard template.

[0014] Furthermore, the edge node adjusts the calibration parameters locally based on the mean square error through the particle swarm optimization algorithm, obtains the local optimization parameters and the corresponding error values, and uploads them to the cloud; the cloud assigns weights according to the node data volume and reliability, and uses the weighted least squares method to fuse the local optimization parameters to generate the global optimal parameters; the global optimal parameters are broadcast to the edge node to replace the local parameters, and the iterative optimization is repeated until the change in the global calibration error is less than the preset change threshold.

[0015] An electronic device includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, a pulse signal wireless transmission and calibration method based on cloud-edge collaboration is implemented.

[0016] A computer-readable storage medium stores a computer program, which, when executed, implements a pulse signal wireless transmission and calibration method based on cloud-edge collaboration.

[0017] (3) Beneficial effects The present invention provides a pulse signal wireless transmission and calibration method based on cloud-edge collaboration, which has the following beneficial effects: (1) By generating estimated noise through adaptive filters and performing time alignment and point-by-point subtraction denoising, the homologous noise in the pulse signal is effectively eliminated, and the signal-to-noise ratio of the signal is significantly improved. Through amplitude distortion detection and preliminary correction, the signal distortion is reduced, the initial quality of the pulse signal is ensured, and a reliable foundation is laid for subsequent time synchronization and calibration, thereby improving the accuracy and stability of the overall system.

[0018] (2) Through the synergy of timestamp exchange in the self-organizing network and the absolute time synchronization signal in the cloud, high-precision time synchronization is achieved between edge nodes. The PI control algorithm is used to fine-tune the local clock and control the time deviation of the entire network within ±5μs, which significantly improves the timing consistency of pulse signal transmission. It not only enhances the system's anti-interference ability, but also provides a reliable time reference for subsequent signal calibration, ensuring the stability and synchronization accuracy of the entire cloud-edge collaborative network.

[0019] (3) By dynamically detecting the actual pulse feature points through a sliding window and aligning them with the pre-stored standard parameters, accurate phase and timing compensation of the pulse signal is achieved. The window position and size can be adaptively adjusted to effectively eliminate the feature point deviation and ensure the accuracy and consistency of the pulse waveform. Combined with the local initial calibration parameters, the phase and timing accuracy of the signal are further optimized, laying a reliable foundation for subsequent global calibration and significantly improving the overall calibration efficiency and signal quality of the system.

[0020] (4) Initial calibration parameters are sent from the cloud and the edge nodes are guided to perform secondary calibration. The particle swarm optimization algorithm is combined to achieve local parameter optimization, which significantly improves the calibration accuracy. The cloud-side weighted fusion of the local optimization results of each node generates the global optimal parameters, and the calibration error is quickly converged to a stable state through iterative optimization. The advantages of cloud-edge collaboration are fully utilized to achieve high-precision, adaptive distributed calibration, while reducing bandwidth usage and ensuring the efficiency and scalability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a schematic diagram of the steps of the pulse signal wireless transmission and calibration method based on cloud-edge collaboration of the present invention; Figure 2 Schematic diagram of the adaptive denoising and preliminary correction process of the present invention; Figure 3 Schematic diagram of the local calibration process of the present invention. DETAILED DESCRIPTION

[0022] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0023] See also Figures 1 to 3 The present invention provides a pulse signal wireless transmission and calibration method based on cloud-edge collaboration, comprising the following steps: Step 1: The edge node collects reference noise in the transmission environment, generates estimated noise through an adaptive filter, denoises the received noisy pulse signal, and performs amplitude distortion detection and preliminary correction. The step 1 includes the following contents: Step 101: The edge node receives a noisy pulse signal transmitted from an external wireless network through a built-in wireless receiving module. For example, the edge node receives the external pulse signal through a radio frequency receiving circuit. At the same time, the edge node activates a noise acquisition module on the same node, which is in the same transmission environment as the wireless receiving module, such as a shared antenna or adjacent sampling channel. The edge node collects reference noise in the transmission environment during the same period as the pulse signal is received. The reference noise and the noise mixed in the pulse signal are homologous noise, i.e., they originate from the same noise source and have the same frequency characteristics and amplitude distribution characteristics. Step 102: The edge device activates a built-in adjustable adaptive filter, such as an adaptive filter based on the LMS algorithm. The initial parameters of the filter are randomly selected within a preset range. The preset range is set based on the spectral characteristics of common noise in pulse signal transmission scenarios. For example, the initial range of the filter order is set to 8-32, and the initial range of the filter coefficient is set to [-0.5, 0.5]. Step 103: Use an adjustable adaptive filter to filter the collected reference noise, output an estimated noise, and extract the actual noise from the received noisy pulse signal. Specifically, the following steps are performed: identify the pulse-free period in the signal using a pulse signal detection algorithm, intercept the sampled values ​​of the pulse-free period as the actual noise sample; and calculate the mean square error between the estimated noise and the actual noise using the mean square error formula: ,in, e is the mean square error, N is the number of sampling points of the noise sample, The actual noise n Sample values, To estimate the noise n Sample values; Step 104: Based on mean square error e, the filter parameters are automatically adjusted through the feedback adjustment mechanism: the filter coefficients are updated using the gradient descent method, and the adjustment step size is set to , μ is the preset learning rate, ranging from 0.01 to 0.1, and the filter parameters are adjusted repeatedly until e Less than or equal to the preset error threshold, such as 0.01. The error threshold is set according to the signal accuracy requirements; Step 105: Time-synchronize the received noisy pulse signal with the estimated noise, mark the sampling time based on the local clock of the edge node, and perform a point-by-point subtraction operation to subtract the sampling value of the estimated noise at the corresponding time from each sampling value of the noisy pulse signal to obtain a denoised pulse signal; Step 106: Perform amplitude distortion detection on the denoised pulse signal. Compare the amplitude sampling value of the denoised signal with the amplitude range of a preset standard pulse (e.g., standard peak amplitude A ± 5% A) to determine the amplitude distortion area, that is, the area that exceeds the amplitude range of the standard pulse. Use a linear correction algorithm to perform preliminary correction on the distortion area. The correction formula is: ,in, is the first signal after denoising k Amplitude sampling values, 、 is the correction coefficient, which is obtained by fitting the relationship between the distortion amplitude and the standard amplitude using the least squares method, so that the amplitude values ​​of the corrected signal fall within the amplitude range of the standard pulse.

[0024] When using, combine the contents of steps 101 to 106: By generating estimated noise through adaptive filters and performing time alignment and point-by-point subtraction denoising, the homologous noise in the pulse signal is effectively eliminated, significantly improving the signal-to-noise ratio of the signal. Through amplitude distortion detection and preliminary correction, signal distortion is reduced, ensuring the initial quality of the pulse signal, laying a reliable foundation for subsequent time synchronization and calibration, and improving the accuracy and stability of the overall system.

[0025] Step 2: Edge nodes achieve initial time synchronization by exchanging timestamps through a self-organizing network. Combined with the absolute time signal from the cloud, the PI control algorithm is used to fine-tune the local clock to achieve unified time across the entire network. The second step includes the following contents: Step 201: Each edge node activates a built-in local timing module, such as a high-precision real-time clock (RTC) chip, starts the timing function, and records its own local time in a preset format, such as a Unix timestamp. The initial timing moment is marked as a synchronization starting reference. Step 202: The edge node forms a self-organizing network based on a preset wireless communication protocol, such as ZigBee or LoRa, and broadcasts a data packet containing its own unique identifier (node ​​ID) and the current local time to all adjacent nodes in the network at a fixed period (e.g., 100ms / time). The data packet uses a CRC check to ensure transmission integrity. Step 203: After receiving the timestamp data packet sent by the neighboring node, each node parses the local time of the neighboring node, calculates the difference between its own local time and the local time of the neighboring node as the preliminary time offset, and obtains the data packet transmission time by recording the difference between the sending and receiving times. The data packet transmission time is deducted when calculating the preliminary time offset; Step 204: The node collects preliminary time deviation values ​​from multiple (e.g., three) neighboring nodes, removes outliers that exceed a preset range (e.g., ±100 μs), and takes the arithmetic mean of the remaining deviation values ​​as a correction for its own clock. The node then adjusts the crystal oscillator frequency of the local timing module to correct the local clock in real time. For example, for every 1 μs deviation, the crystal oscillator output frequency is adjusted by 0.1 ppm. Step 205: Repeat steps 202 to 204 to exchange timestamps, calculate offsets, and correct clocks. After each correction, record the time difference between nodes. When the time difference between all adjacent nodes is detected to be less than a preset time difference threshold (e.g., 10 μs) for multiple consecutive times (e.g., 5 times), it is determined that preliminary synchronization between edge nodes is complete. Step 206: The cloud server acts as the time master node, uses a GPS-disciplined rubidium atomic clock as the clock source, and broadcasts a synchronization signal containing the absolute time and a broadcast time stamp to all edge nodes at a fixed period (e.g., 1 second / time). The synchronization signal is encrypted to prevent tampering. Step 207: After receiving the synchronization signal broadcast from the cloud, the edge node parses the absolute time, calculates its deviation from the local current time, and uses the proportional integral (PI) control algorithm to fine-tune the phase of the local timing module. For example, for every 10ns deviation, the phase is adjusted by 0.1°. This ensures that the deviation between the local time and the absolute time on the cloud is within ±5μs, achieving time uniformity across the entire network.

[0026] When using, combine the contents of steps 201 to 207: Through the synergy of timestamp exchange in the self-organizing network and absolute time synchronization signals in the cloud, high-precision time synchronization between edge nodes is achieved. The PI control algorithm is used to fine-tune the local clock and control the time deviation of the entire network within ±5μs, significantly improving the timing consistency of pulse signal transmission. This not only enhances the system's anti-interference ability, but also provides a reliable time reference for subsequent signal calibration, ensuring the stability and synchronization accuracy of the entire cloud-edge collaborative network.

[0027] Step 3: The edge node dynamically detects the actual feature points of the pulse signal through a sliding window, aligns them with the pre-stored standard feature points, adjusts the window parameters, and performs local calibration based on the local initial calibration parameters; The step three includes the following contents: Step 301: Pre-store characteristic parameters of a standard pulse waveform in a local storage module of an edge node, including the peak position, rising edge starting point, and falling edge end point. The peak position is defined as the sampling point number when the pulse amplitude reaches its maximum value, the rising edge starting point is defined as the sampling point number when the pulse amplitude rises from the baseline value to 10% of the standard peak value, and the falling edge end point is defined as the sampling point number when the pulse amplitude drops from the peak value to 10% of the standard peak value. The baseline value is the average amplitude during a period without a pulse signal, and the standard peak value is a preset peak value, such as 5. Step 302: Based on the expected duration of the standard pulse (e.g., 10 μs) and the sampling rate (e.g., 1 sampling point per microsecond), the size of the initial sliding window is set to the number of sampling points that contain the complete pulse waveform, e.g., 12 sampling points, with 2 sampling points reserved for redundancy. The initial starting position of the sliding window is set to the position where the amplitude of the pulse signal is first detected to exceed 2 standard deviations of the baseline value. Step 303: Move the sliding window with a step size of 1 sampling point, and perform feature point detection on the pulse signal within the sliding window; use a first-order difference algorithm to identify rising and falling edges, with the rising edge starting point being the moment when the difference result turns from negative to positive and exceeds a preset rising threshold, and the falling edge ending point being the moment when the difference result turns from positive to negative and exceeds a preset falling threshold. Use a local extreme value search algorithm to locate the peak position, which is the sampling point corresponding to the maximum amplitude within the window; and record the sampling point number of the actual feature point within the window. Step 304: Calculate the positional deviation between the actual feature point and the standard feature point, such as the difference between the sampling point number of the actual peak position and the sampling point number of the standard peak position; if the absolute value of the positional deviation is greater than 1 sampling point, adjust the starting sampling point number of the window and shift the sliding window to align the actual feature point with the standard feature point, i.e., align their relative positions within the window, for example, both are at the 6th sampling point position in the window; repeat the adjustment until the absolute value of the deviation is ≤ 1 sampling point; Step 305: If the difference between the length from the rising edge start point to the falling edge end point of the actual pulse and the length of the standard pulse exceeds a preset length threshold, such as 5%, the window size is automatically adjusted: when the length increases, the window size is expanded by 2 sampling points each time, and when the length decreases, the window size is reduced by 2 sampling points each time. After the adjustment, steps 303 to 304 are repeated until the window completely contains the actual pulse waveform, that is, the rising edge start point and the falling edge end point are both within the window. Step 306: Based on the actual feature points after alignment, the edge node calls the locally stored initial calibration parameters for local calibration, including the phase compensation coefficient and the timing delay correction value; for phase deviation, the sampling phase of the pulse signal is adjusted through the linear interpolation algorithm to make the actual peak phase consistent with the standard peak phase; for timing deviation, the time mark of the signal is corrected according to the timestamp difference of the feature points to ensure that the timing of the pulse signal meets the standard requirements.

[0028] When using, combine the contents of steps 301 to 306: By dynamically detecting the actual pulse feature points through a sliding window and aligning them with the pre-stored standard parameters, accurate phase and timing compensation of the pulse signal is achieved. The window position and size can be adaptively adjusted to effectively eliminate feature point deviations and ensure the accuracy and consistency of the pulse waveform. Combined with the local initial calibration parameters, the phase and timing accuracy of the signal are further optimized, laying a solid foundation for subsequent global calibration and significantly improving the overall calibration efficiency and signal quality of the system.

[0029] Step 4: The cloud sends initial calibration parameters to the edge node. The edge node performs secondary calibration and uploads local optimized parameters. The cloud performs weighted fusion to generate the global optimal parameters and iterates the optimization until the calibration error stabilizes.

[0030] The fourth step includes the following contents: Step 401: The cloud server sends initial calibration parameters to each edge node via an encrypted communication channel. The initial calibration parameters are pre-stored in the cloud database and include phase calibration coefficients, timing delay correction values, and amplitude compensation factors. The parameter format uses a JSON structure and contains a mapping between node IDs and corresponding parameters. Step 402: After receiving the initial calibration parameters, the edge node calls the local calibration module to perform secondary calibration on the locally calibrated pulse signal, wherein the phase calibration is: , is the pulse signal phase after secondary calibration, To calibrate the pulse signal phase locally, t is a time point; timing calibration: , is the time point after the second calibration, Calibrate the time point of the pulse signal locally; Amplitude calibration: , is the pulse signal amplitude after secondary calibration, is the local calibration pulse signal amplitude, is the amplitude compensation factor, is the phase calibration coefficient, is the timing delay correction value; The root mean square error (RMSE) is used to calculate the deviation between the calibrated signal and the standard pulse template. The formula is: ,in, M is the number of sampling points of the pulse signal, The first signal after calibration i Sample values, is the corresponding sampling value of the standard template; Step 403: The edge node adjusts the calibration parameters locally based on the root mean square error (RMSE) using the particle swarm optimization (PSO) algorithm. The particle swarm size is set to 20, the number of iterations is set to 50, and the inertia weight is linearly decreased from 0.9 to 0.4. The error is recalculated after each iteration, and the parameter combination that minimizes the RMSE is selected as the local optimization parameter. The local optimization parameter and the corresponding error value are packaged and uploaded to the cloud. The original pulse data is not uploaded to reduce bandwidth usage. Step 404: After receiving the local optimization parameters and error values ​​uploaded by each edge node, the cloud assigns weights according to the following rules: the node data volume weight is calculated based on the number of valid parameter groups uploaded by the node in the past hour. For example, if the total number of groups is Z and the number of groups for a node is z, then the weight is z / Z); the node reliability weight is determined based on the ratio of the node's historical error mean to the average error of the entire network. The weight values ​​are normalized to sum to 1. Step 405: The cloud uses weighted least squares to calculate the global optimal calibration parameters: for each parameter dimension, such as the phase calibration coefficient and the timing delay correction value, the local optimization parameters of each node are multiplied by the corresponding weight and then summed. The formula is: ,in, K is the total number of edge nodes, For the k The weight of the node, is the local optimization parameter of the node; Step 406: The cloud sends the global optimal calibration parameters to all edge nodes via a broadcast mechanism. The sent data packet includes the parameter version number, effective time, and check code. Step 407: After the edge node receives the global optimal calibration parameters and verifies them with the check code, it replaces the locally stored initial calibration parameters; repeats the calibration, error calculation and parameter optimization process of steps 402 to 406 until the change in the global calibration error is less than a preset change threshold, such as 0.001, for three consecutive times. It is determined that the global calibration error is stable at a low level and the iteration stops.

[0031] When used, combine the contents of steps 401 to 407: By sending initial calibration parameters from the cloud and guiding edge nodes to perform secondary calibration, local parameter optimization is achieved by combining the particle swarm optimization algorithm, which significantly improves the calibration accuracy. The cloud-based weighted fusion of the local optimization results of each node generates the global optimal parameters, and through iterative optimization, the calibration error quickly converges to a stable state. The advantages of cloud-edge collaboration are fully utilized to achieve high-precision, adaptive distributed calibration, while reducing bandwidth usage and ensuring the efficiency and scalability of the system.

[0032] An electronic device includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, a pulse signal wireless transmission and calibration method based on cloud-edge collaboration is implemented.

[0033] A computer-readable storage medium includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, a pulse signal wireless transmission and calibration method based on cloud-edge collaboration is implemented.

[0034] In the application, the several formulas involved are all calculated by taking their numerical values ​​after removing the dimensions, and the formula is a formula obtained by collecting a large amount of data and performing software simulation to obtain the latest real situation. The coefficients in the formula are set by technical personnel in this field according to actual conditions.

[0035] The above embodiments can be implemented in whole or in part through software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art will appreciate that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented using a combination of electronic hardware, computer software, and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution.

[0036] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, and may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment as needed.

[0037] The above is only a specific implementation method of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed in this application, which should be covered by the scope of protection of the present application.

Claims

1. A pulse signal wireless transmission and calibration method based on cloud-edge collaboration, characterized by: include: The edge node collects reference noise in the transmission environment, generates estimated noise through an adaptive filter, denoises the received noisy pulse signal, and performs amplitude distortion detection and preliminary correction. Edge nodes achieve preliminary time synchronization by exchanging timestamps through a self-organizing network. Combined with the absolute time signal from the cloud, the PI control algorithm is used to fine-tune the local clock to achieve unified time across the entire network. The edge node dynamically detects the actual feature points of the pulse signal through a sliding window, aligns them with the pre-stored standard feature points, adjusts the window parameters, and performs local calibration based on the local initial calibration parameters; The cloud sends initial calibration parameters to the edge node, the edge node performs secondary calibration and uploads local optimization parameters, and the cloud performs weighted fusion to generate the global optimal parameters, and iterates the optimization until the calibration error stabilizes.

2. The pulse signal wireless transmission and calibration method based on cloud-edge collaboration according to claim 1 is characterized in that: The edge node receives the noisy pulse signal through the wireless receiving module and simultaneously collects the reference noise from the same source as the noise in the pulse signal; the initial parameters of the adaptive filter are randomly set within a preset range, and the reference noise is filtered to generate estimated noise; the actual noise in the noisy pulse signal is extracted, the mean square error between the estimated noise and the actual noise is calculated, and the filter parameters are adjusted by the gradient descent method until the error is less than the preset error threshold.

3. The pulse signal wireless transmission and calibration method based on cloud-edge collaboration according to claim 2 is characterized by: After time alignment between the noisy pulse signal and the estimated noise, a point-by-point subtraction operation is performed to obtain the denoised pulse signal; the amplitude distortion area of ​​the denoised pulse signal is detected, and a linear correction algorithm is used to perform preliminary amplitude correction to make the signal amplitude fall within the amplitude range of the standard pulse.

4. The pulse signal wireless transmission and calibration method based on cloud-edge collaboration according to claim 1 is characterized in that: Each edge node records its own local time through a local timing module. It broadcasts a timestamp packet containing the node ID and local time to all adjacent nodes in the network through a self-organizing network at a fixed period. It receives timestamp packets from adjacent nodes, calculates the initial deviation from its own local time, and uses this as a correction after deducting the transmission time. It collects the deviation values ​​of multiple adjacent nodes, removes outliers, and takes the average value as the local clock correction value, adjusting the crystal oscillator frequency to correct the clock in real time. Repeat the timestamp exchange and clock correction until the time difference between adjacent nodes is less than the preset time difference threshold.

5. The pulse signal wireless transmission and calibration method based on cloud-edge collaboration according to claim 4 is characterized in that: The cloud server acts as the time master node and broadcasts a synchronization signal containing the absolute time and the broadcast time mark to all edge nodes at a fixed period. After receiving the synchronization signal broadcast by the cloud, the edge node parses the absolute time, calculates its deviation from the local current time, and uses the PI control algorithm to fine-tune the local clock phase.

6. The pulse signal wireless transmission and calibration method based on cloud-edge collaboration according to claim 1 is characterized in that: The characteristic parameters of the standard pulse waveform are pre-stored in the edge node, including the peak position, the starting point of the rising edge and the ending point of the falling edge; Set the initial sliding window size and starting position according to the standard pulse duration and sampling rate; Move the sliding window to detect the actual pulse feature points, calculate the position deviation from the standard feature points, and align the feature points by adjusting the window position; If the actual pulse length deviates from the standard length by more than the preset length threshold, the window size is dynamically adjusted to ensure that the waveform is completely contained. Based on the aligned feature points, the local initial calibration parameters are called for phase and timing compensation to correct the phase deviation and time mark of the signal.

7. The pulse signal wireless transmission and calibration method based on cloud-edge collaboration according to claim 1 is characterized in that: The cloud sends initial calibration parameters to each edge node through an encrypted channel. The parameters include phase calibration coefficient, timing delay correction value and amplitude compensation factor. The edge node uses the initial calibration parameters to perform secondary calibration on the locally calibrated pulse signal and calculates the root mean square error between the calibrated signal and the standard template.

8. The pulse signal wireless transmission and calibration method based on cloud-edge collaboration according to claim 7 is characterized in that: Based on the mean square error, the edge node adjusts the calibration parameters locally through the particle swarm optimization algorithm to obtain the local optimized parameters and corresponding error values, and uploads them to the cloud; the cloud assigns weights according to the node data volume and reliability, and uses the weighted least squares method to fuse the local optimized parameters to generate the global optimal parameters; the global optimal parameters are broadcast to the edge node to replace the local parameters, and the iterative optimization is repeated until the change in the global calibration error is less than the preset change threshold.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, it implements the pulse signal wireless transmission and calibration method based on cloud-edge collaboration as described in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed, implements the pulse signal wireless transmission and calibration method based on cloud-edge collaboration according to any one of claims 1 to 8.

Images