A method and system for high voltage cable heterogeneous signal synchronization

Abstract

The application relates to a high-voltage cable heterogeneous signal synchronization method and system, and relates to the field of computer technology, which comprises the following steps: collecting original timestamp information; constructing original time references of various heterogeneous signals to obtain an initial aligned multi-modal signal; performing link transmission delay compensation on the initial aligned multi-modal signal to obtain a delay-compensated multi-modal signal; performing clock frequency drift correction on the delay-compensated multi-modal signal to obtain a frequency-corrected multi-modal signal; performing intermediate network equipment residence time correction on the frequency-corrected multi-modal signal to obtain a time reference unified multi-modal signal; and performing resampling on the time reference unified multi-modal signal based on signal inter-physical correlation characteristic information to obtain a time-synchronized multi-modal signal. The application has the effects of eliminating physical phase mismatch and clock deviation, ensuring alignment accuracy in a long-distance monitoring scene, and improving the data quality of high-voltage cable operation state fault diagnosis.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a method and system for synchronizing heterogeneous signals in high-voltage cables. Background Technology

[0002] In related technologies, when monitoring the operating status of high-voltage cables, multiple sensors (e.g., vibration, temperature, and current sensors) are typically deployed to form a multimodal monitoring system to sense the cable's operating condition. To handle heterogeneous signals with different sampling rates, existing technologies typically employ fixed-frequency interpolation for synchronization, resampling the heterogeneous signals onto a time axis with consistent signals through linear interpolation. For time differences between multiple sensor nodes, network time protocols are typically used for periodic correction to maintain clock synchronization among the nodes.

[0003] Regarding the aforementioned technologies, when resampling low-frequency signals using fixed-frequency interpolation, only the mathematical alignment of the time axis is considered, ignoring the physical correlation between signals such as vibration and temperature. This leads to phase distortion in the reconstructed signal, disrupting the physical timing relationship of the original signal. Second, in long-distance (e.g., over 20 kilometers) cable monitoring scenarios, signals need to be transmitted through multiple levels of network devices. Each sensor node uses an independent crystal oscillator clock source. The periodic time synchronization of the network time protocol cannot eliminate clock frequency drift caused by crystal oscillators (aging or temperature drift), causing time deviations between nodes to accumulate over time, making it impossible to maintain alignment on the time axis for extended periods. Third, signals pass through multiple intermediate network devices during transmission. The queue scheduling and forwarding processing of these devices introduce uncertain dwell time jitter. Existing technologies cannot effectively remove the asymmetric delays caused by such device forwarding, resulting in decreased time synchronization accuracy, leaving room for improvement. Summary of the Invention

[0004] To overcome the phase distortion, clock frequency drift accumulation error, and uncertainty delay introduced by intermediate network devices in the alignment process of heterogeneous signals, this application provides a method for synchronizing heterogeneous signals in high-voltage cables.

[0005] Firstly, this application provides a method for synchronizing heterogeneous signals in high-voltage cables, employing the following technical solution: The system collects hardware timestamps of incoming and outgoing packets to obtain raw timestamp information. Based on this raw timestamp information, it constructs the original time base for each heterogeneous signal to obtain initially aligned multimode signals. It then performs link transmission delay compensation on the initially aligned multimode signals to obtain delay-compensated multimode signals. Finally, it performs clock frequency drift correction on the delay-compensated multimode signals to obtain frequency-corrected multimode signals. Next, it performs intermediate network device dwell time correction on the frequency-corrected multimode signals to obtain time-base unified multimode signals. Based on the physical correlation characteristics between signals, it resamples the time-base unified multimode signals to obtain time-synchronized multimode signals.

[0006] Secondly, this application provides a high-voltage cable heterogeneous signal synchronization system, which adopts the following technical solution: The acquisition module is used to acquire hardware timestamps; the memory is used to store the program of the above-mentioned high-voltage cable heterogeneous signal synchronization method; the processor is used to load and execute the program in the memory and implement the above-mentioned high-voltage cable heterogeneous signal synchronization method.

[0007] The above-described embodiments of this disclosure have the following beneficial effects: The high-voltage cable heterogeneous signal synchronization method of some embodiments of this disclosure eliminates physical phase mismatch and clock deviation, ensures alignment accuracy in long-distance monitoring scenarios, and improves the data quality for high-voltage cable operation status fault diagnosis. Specifically, the reasons for heterogeneous signal asynchrony are: phase distortion, clock frequency drift accumulation error, and uncertain delay introduced by intermediate network devices during the alignment process. Based on this, the high-voltage cable heterogeneous signal synchronization method of some embodiments of this disclosure first collects the hardware timestamps of incoming and outgoing packets to obtain raw timestamp information. Thus, the timestamps of data packets leaving or arriving (nanosecond level) are directly obtained at the driver layer, avoiding random delays introduced by the application layer and protocol stack processing, and providing a raw time reference for subsequent multi-modal signal time synchronization. Second, based on the above raw timestamp information, the raw time references for each heterogeneous signal are constructed to obtain initially aligned multi-modal signals. Thus, the multi-source signals distributed across different sensor nodes are unified to the same time coordinate system, eliminating the initial time deviation of the clock sources of each node, and laying the foundation for subsequent step-by-step compensation. Then, link transmission delay compensation is performed on the initially aligned multimode signals to obtain delay-compensated multimode signals. This suppresses link transmission delay fluctuations caused by network load changes and queue scheduling factors, avoiding the shortcomings of fixed delay compensation methods that cannot adapt to dynamic network environments. Furthermore, clock frequency drift correction is performed on the delay-compensated multimode signals to obtain frequency-corrected multimode signals. Thus, by calculating the frequency drift factor online and dynamically adjusting the local time accumulation compensation value, long-term clock accumulation errors caused by crystal oscillator aging or environmental temperature changes are suppressed, ensuring the synchronization stability of multimode signals under long-term operation. Next, intermediate network device dwell time correction is performed on the frequency-corrected multimode signals to obtain time-base unified multimode signals. Thus, by measuring and compensating for the dwell time of messages in each switching device through a transparent clock mechanism, delay jitter introduced by device forwarding uncertainty in multi-hop network topologies is eliminated, improving time synchronization accuracy in long-distance transmission scenarios. Finally, based on the physical correlation characteristics between signals, the time-base unified multimode signals are resampled to obtain time-synchronized multimode signals. Therefore, the traditional fixed-frequency interpolation method was abandoned, and a resynchronization model was constructed using the physical causal relationship between signals such as vibration and temperature, thus avoiding phase distortion problems. This improved the data quality for fault diagnosis of high-voltage cable operation status. Attached Figure Description

[0008] Figure 1 This is a flowchart of some embodiments of the high-voltage cable heterogeneous signal synchronization method according to the present disclosure. Detailed Implementation

[0009] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0010] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0011] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0012] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0013] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0014] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0015] refer to Figure 1 The flowchart 100 illustrates some embodiments of a high-voltage cable heterogeneous signal synchronization method according to the present disclosure. This high-voltage cable heterogeneous signal synchronization method includes the following steps: Step 101: Collect the hardware timestamps of incoming and outgoing messages to obtain the raw timestamp information.

[0016] In some embodiments, the execution entity (e.g., an electronic device) of the above-described high-voltage cable heterogeneous signal synchronization method can be hardware or software. When the computing device is hardware, it can be implemented as a distributed cluster composed of multiple servers or terminal devices, or as a single server or a single terminal device. When the computing device is software, it can be installed in the hardware devices listed above. It can be implemented as multiple software programs or software modules to provide distributed services, or as a single software program or software module. No specific limitations are made here.

[0017] In some embodiments, the aforementioned execution entity can collect hardware timestamps of incoming and outgoing packets to obtain raw timestamp information. The aforementioned hardware timestamp can be the time value recorded by the network interface card hardware when a data frame leaves or arrives at a physical port.

[0018] As an example, firstly, in the Linux system of the data acquisition node, after creating a UDP socket for sending sensor data, call... The function sets the SO_TIMESTAMPING option and enables the SOF_TIMESTAMPING_TX_HARDWARE and SOF_TIMESTAMPING_RAW_HARDWARE flags. Then, after the data packet is sent, the network card hardware records a timestamp as the data frame is about to leave the MAC layer. After the data packet is sent, the application can... The function receives a control message and extracts the hardware timestamp from the auxiliary data of type SCM_TIMESTAMPING to obtain the raw timestamp information.

[0019] Step 102: Based on the original timestamp information, construct the original time reference for each heterogeneous signal to obtain the initially aligned multimodal signal.

[0020] In some embodiments, the execution entity can construct an original time reference for each heterogeneous signal based on the original timestamp information to obtain an initially aligned multimodal signal. The original time reference can be a time scale that unifies the local clocks of each sensor node to a unified reference system. The initially aligned multimodal signal can be a sequence of multiple signals (e.g., vibration, temperature, current, acoustic signature) collected by each sensor after initial alignment on the time axis.

[0021] As an example, firstly, a Grandmaster Clock can be deployed in the network equipment room. This Grandmaster Clock can receive BeiDou / GPS dual-mode satellite signals. Then, a Precision Time Protocol (IEEE 1588 PTP) slave clock protocol stack runs on each sensor data acquisition node. Next, by periodically exchanging Sync, Follow_Up, Pdelay_Req, and Pdelay_Resp messages between the Grandmaster and slave clocks, the hardware timestamp of each message in the driver layer is recorded. Furthermore, the driver layer timestamp for the Grandmaster Clock to send the Sync message can be set to... The driver layer timestamp for receiving the Sync message from the clock is... The driver layer timestamp for sending the Delay_Req message from the clock is... The driver layer timestamp for the master clock receiving the Delay_Req message is... The time difference between the master and slave clocks is... Finally, the calculated time deviation Compensation is made to the local time from the clock. When the signals collected by each sensor node are timestamped locally, the synchronized time base is used as the primary reference to obtain the initially aligned multimodal signal.

[0022] Step 103: Perform link transmission delay compensation on the initially aligned multimode signal to obtain the delay-compensated multimode signal.

[0023] In some embodiments, the execution entity may perform link transmission delay compensation on the initially aligned multimodal signal to obtain a delay-compensated multimodal signal. The link transmission delay may be the total time it takes for sensor data to travel from the edge acquisition node to the central processing unit.

[0024] In some optional implementations of certain embodiments, the execution entity may perform link transmission delay compensation on the initially aligned multimode signal to obtain the delay-compensated multimode signal, which may include the following steps: The first step is to establish a first prediction equation and a second prediction equation. The first prediction equation can be an equation describing the evolution of the transmission delay state over time. The second prediction equation can be an equation describing the relationship between the observed transmission delay values ​​and the state vector. The state vector can be a mathematical model characterizing the dynamic characteristics of the link transmission delay. The state vector can include: the transmission delay state component at the current moment and the delay change rate state component. In practice, the first prediction equation can be a state equation, and the second prediction equation can be an observation equation.

[0025] As an example, the first prediction equation described above could take the form of... Among them, the above It can be the first The state vector at time t, The above. This could be an estimate of the transmission delay at the current moment. (The above...) This could be an estimate of the rate of change due to delay. (The above...) , can be a state transition matrix. The above This can be the sampling time interval. (The above...) This could be process noise. The form of the second prediction equation mentioned above could be... Among them, the above This could be the current link transmission delay observation. (The above...) This can be an observation matrix. (The above...) It could be observation noise.

[0026] The second step is to obtain the interaction data of the periodic message exchange between the master clock node and the slave clock node. This interaction data can be the hardware timestamp of each message at the driver layer.

[0027] The third step is to determine the observed link transmission delay based on the aforementioned interactive data. In practice, the one-way or two-way path delay between the slave clock and the master clock can be calculated based on the interactive data to obtain the observed link transmission delay at the current moment.

[0028] As an example, first, the downlink path delay can be determined. Upbound delay The above-mentioned link transmission delay observations can be... The above. This could be the precise timestamp at which the message leaves the driver layer when the master clock sends a Sync message. (The above...) This could be the precise timestamp of the message arriving at the driver layer when the master clock receives the Sync message. (The above...) This could be the timestamp when the master clock receives the Delay_Req message and reaches the driver layer. (The above...) It can be the timestamp when the master clock leaves the driver layer after sending the Delay_Req message.

[0029] The fourth step is to determine the transmission delay status information at the current moment based on the first prediction equation described above. This transmission delay status information may include, but is not limited to, the estimated transmission delay and the estimated rate of change of delay at the current moment.

[0030] As an example, firstly, the state estimate from the previous moment can be... Substituting into the first prediction equation, we obtain the transmission delay state information at the current moment. Simultaneously, the prediction error covariance matrix can be determined. Among them, the above It can be the process noise covariance matrix. (The above...) It can be the error covariance matrix of the previous time step.

[0031] Fifth, based on the aforementioned link transmission delay observations and the second prediction equation, the current transmission delay status information is corrected to obtain the transmission delay estimate. In practice, an iterative Kalman filter can be used to determine the transmission delay estimate.

[0032] Step 6: Based on the above transmission delay estimate, the original timestamps of the initially aligned multimode signals are compensated to obtain the delay-compensated multimode signals.

[0033] As an example, firstly, the estimated transmission delay can be added to the original timestamp to obtain the absolute timestamp. Then, based on the absolute timestamp, the delay-compensated multimode signal can be output.

[0034] In some optional implementations of certain embodiments, the execution entity may correct the current transmission delay state information based on the link transmission delay observation and the second prediction equation to obtain the transmission delay estimate, which may include the following steps: The first step is to determine the observed prediction value at the current moment based on the transmission delay state information and the second prediction equation. This observed prediction value can be a theoretical observation value obtained by mapping the transmission delay state information to the observation space.

[0035] The second step is to determine the observation residual based on the observed link transmission delay and the predicted observation. The observation residual can represent the degree of deviation between the current predicted value and the actual measured value. The observation residual can be the difference between the observed link transmission delay and the predicted observation.

[0036] The third step is to obtain the prediction error covariance information corresponding to the transmission delay state information at the current moment. This prediction error covariance information can be a matrix used to characterize the uncertainty of the transmission delay state information at the current moment. This prediction error covariance information reflects the error variance between the predicted state and the actual state.

[0037] The fourth step involves determining the corrected weighting coefficients based on the second prediction equation and the prediction error covariance information described above. These corrected weighting coefficients can be used to balance the weights of predicted and observed information during the state correction process.

[0038] As an example, the formula for calculating the above-mentioned corrected weighting coefficient can be: Among them, the above This could be information about the prediction error covariance. (The above...) This can be an observation matrix. (The above...) It can be the transpose of the observation matrix. (The above...) It can be the observation noise covariance matrix.

[0039] Fifth, based on the above-mentioned corrected weighting coefficients and the above-mentioned observation residuals, the transmission delay state information at the current moment is weighted and corrected to obtain the transmission delay estimate.

[0040] As an example, firstly, the weighting coefficients can be multiplied by the observation residuals to obtain the correction. Then, the correction is added to the transmission delay state information at the current time to obtain the transmission delay estimate.

[0041] Step 104: Perform clock frequency drift correction on the delayed-compensated multimode signal to obtain the frequency-corrected multimode signal.

[0042] In some embodiments, the aforementioned execution entity can perform clock frequency drift correction on the delay-compensated multimode signal to obtain a frequency-corrected multimode signal. The aforementioned clock frequency drift can be a deviation between the crystal oscillator frequency of the local clock and the master clock crystal oscillator frequency. This deviation can gradually accumulate with time and changes in ambient temperature, resulting in a long-term drift between the local clock time and the master clock time.

[0043] In some optional implementations of certain embodiments, the execution entity may perform clock frequency drift correction on the delay-compensated multimode signal to obtain a frequency-corrected multimode signal, which may include the following steps: The first step is to obtain the transmission timestamps of at least two consecutive synchronization messages at the master clock and the reception timestamps at the slave clock. These synchronization messages can be protocol messages used for time synchronization between the master and slave clocks. The transmission timestamp can be the moment the synchronization message leaves the master clock driving layer. The reception timestamp can be the moment the synchronization message arrives at the slave clock driving layer.

[0044] As an example, suppose the first The timestamp of a synchronization message sent on the master clock can be... The received timestamp from the clock end can be... . No. The timestamp of a synchronization message sent on the master clock can be... The received timestamp from the clock end can be... .

[0045] The second step is to determine the transmission time interval of adjacent messages in the above-mentioned at least two consecutive synchronization messages based on the above-mentioned transmission timestamp and reception timestamp, so as to obtain the master clock physical time interval and slave clock local time interval.

[0046] As an example, the aforementioned master clock physical time interval can be The aforementioned local time interval from the clock can be... .

[0047] The third step involves calculating the ratio between the aforementioned physical time interval of the master clock and the aforementioned local time interval of the slave clock to obtain the frequency drift factor, which characterizes the frequency deviation between the master and slave clocks. This frequency drift factor can be the deviation rate of the slave clock frequency relative to the master clock frequency.

[0048] As an example, the formula for calculating the frequency drift factor mentioned above can be: . This indicates that the slave clock frequency is higher than the master clock frequency. This indicates that the slave clock frequency is higher than the master clock frequency.

[0049] The fourth step is to determine the time span information based on the preset update time and the current time. The preset update time can be the time when the frequency drift factor was last calculated. The time span information can be the length of time from the preset update time to the current time.

[0050] As an example, the above-mentioned preset update time could be The aforementioned current moment can be... The time span information is as follows: .

[0051] The fifth step involves performing error accumulation processing on the aforementioned frequency drift factor and time span information to obtain the dynamic cumulative offset. This dynamic cumulative offset can be the time deviation accumulated by the frequency drift within the time span.

[0052] As an example, the formula for calculating the above dynamic cumulative offset can be: Among them, the above This could be a frequency drift factor. (The above...) It can be information spanning a time period.

[0053] Step 6: Summate the aforementioned dynamic cumulative offset and the preset initial offset to obtain the total logical clock compensation. The preset initial offset can be a static time offset set at the initial synchronization moment. The total logical clock compensation can be a compensation value used to correct the local clock time.

[0054] As an example, the preset initial offset can be set to... The formula for calculating the total compensation amount of the logic clock can be: .

[0055] Step 7: Based on the total logic clock compensation amount mentioned above, correct the original hardware clock reading of the multi-mode signal after delay compensation to obtain the logic clock output time series. The logic clock output time series can be a sequence of logic time values ​​after frequency drift correction.

[0056] As an example, the raw hardware clock reading can be set to... The formula for calculating the logic clock output time series can be expressed as follows: Among them, the above It can be the total compensation amount of the logical clock at the current moment.

[0057] Step 8: Remap the timestamps of the delay-compensated multimode signal using the logic clock output time series to obtain the frequency-corrected multimode signal. In practice, the original timestamps of each data point in the delay-compensated multimode signal can be replaced with the corresponding logic clock output time series values ​​to obtain the frequency-corrected multimode signal.

[0058] Step 105: Perform intermediate network device dwell time correction on the frequency-corrected multimode signal to obtain a multimode signal with a unified time base.

[0059] In some embodiments, the aforementioned executing entity can perform intermediate network device dwell time correction on the frequency-corrected multimode signal to obtain a multimode signal with a unified time base. The dwell time can be the time it takes for a synchronization event message to travel from the ingress port to the egress port within the intermediate network device. The dwell time may include, but is not limited to, parsing processing time, queuing processing time, and forwarding processing time.

[0060] In some optional implementations of certain embodiments, the execution entity may perform intermediate network device dwell time correction on the frequency-corrected multimode signal to obtain a time-base unified multimode signal, which may include the following steps: The first step is to determine the moment when the synchronization event message enters the physical port of the intermediate network device, thus obtaining the ingress capture trigger signal. This ingress capture trigger signal can be a hardware-generated trigger signal when the synchronization event message begins to enter the device port.

[0061] The second step is to trigger the entry capture action based on the aforementioned entry capture trigger signal. This entry capture action can be a hardware operation that latches and records the current moment.

[0062] The third step is to determine the ingress hardware timestamp generated by the aforementioned ingress capture action. This ingress hardware timestamp can be the moment when the synchronization event message enters the physical port of the intermediate network device.

[0063] The fourth step is to detect the moment when the aforementioned synchronization event message leaves the physical port of the intermediate network device to obtain the egress capture trigger signal. This egress capture trigger signal can be a hardware-generated trigger signal when the synchronization event message leaves the device port.

[0064] As an example, when a synchronization event message is sent from the egress port after being processed and queued inside the device, the physical layer chip at the egress port detects the moment when the tail of the message is about to be sent and generates an egress capture trigger signal.

[0065] Fifth step: Trigger the exit capture action based on the exit capture trigger signal mentioned above.

[0066] As an example, when the exit capture trigger signal is valid, the hardware timestamp unit of the switching device reads the value of the current high-precision clock counter to latch that value.

[0067] Step 6: Determine the egress hardware timestamp generated by the aforementioned egress capture action. This egress hardware timestamp can be the moment the synchronization event message leaves the physical port of the intermediate network device.

[0068] Step 7: Subtract the aforementioned egress hardware timestamp and the aforementioned ingress hardware timestamp to obtain the original dwell time of the aforementioned synchronization event message. The original dwell time can be the time elapsed from the entry port to the exit port of the intermediate network device.

[0069] As an example, the formula for calculating the original dwell time mentioned above can be: The above. This could be the export hardware timestamp. (The above...) It can be the entry hardware timestamp.

[0070] Step 8: Based on a preset clock synchronization protocol, the original dwell time is converted into a correction value to obtain correction increment information. This correction increment information can be a correction value converted from the original dwell time to a format conforming to the protocol (e.g., PTP).

[0071] Step 9: Write the above-mentioned incremental correction information into the correction field of the above-mentioned synchronization event message to obtain the first synchronization message, wherein the above-mentioned first synchronization message is a synchronization message carrying local forwarding delay information.

[0072] As an example, firstly, the switching device can modify the header of the synchronization event message, adding the correction amount to the original value of the correction field. Then, it forwards the accumulated synchronization event message to obtain the first synchronization message.

[0073] Step 10: Based on the frequency-corrected multimode signal, perform path concatenation processing on the first synchronization message to obtain the path deviation parameters to be compensated. The path concatenation processing can involve correlating the sum of dwell times accumulated after passing through multiple intermediate network devices with the corresponding multimode signal sampling points.

[0074] Step 11: Based on the path deviation parameters to be compensated mentioned above, the delayed observations are decoupled to obtain a multimodal signal with a unified time base. The delayed observations can be the original total path delay value directly obtained from the synchronization event message.

[0075] In some optional implementations of certain embodiments, the execution entity may perform path concatenation processing on the first synchronization message based on the frequency-corrected multimode signal to obtain the path deviation parameter to be compensated, which may include the following steps: The first step, in response to the first synchronization message being cascaded and forwarded through multiple intermediate network devices, is to determine the correction field information of the first synchronization message. This correction field information may be a correction value accumulated through multiple levels of transparent clocks. The correction field information may also include the sum of the dwell times of all intermediate devices along the route.

[0076] The second step is to determine the total dwell time of the above-mentioned corrected field information.

[0077] The third step involves performing end-to-end path mapping on the sum of the aforementioned dwell times based on the network topology, to obtain device deviation information along the transmission path. Here, the aforementioned network topology can be the connection relationships between various intermediate network devices. The aforementioned device deviation information can be the total deterministic delay information caused by forwarding by each intermediate device.

[0078] The fourth step is to determine the sampling point sequence corresponding to the timestamps of the frequency-corrected multimodal signal and the first synchronization message. Specifically, there is a one-to-one correspondence between the timestamps and the sampling points in the frequency-corrected multimodal signal.

[0079] In practice, the corresponding sampling points can be retrieved from the frequency-corrected multimodal signal based on the timestamp of the first synchronization message to obtain the sampling point sequence.

[0080] The fifth step involves temporally correlating the aforementioned device deviation information with the aforementioned sampling point sequence to obtain the path deviation parameters to be compensated. These path deviation parameters can be parameters derived by subtracting the device forwarding jitter component from the total delay observations.

[0081] In some optional implementations of certain embodiments, the execution entity may decouple the delayed observations based on the path deviation parameters to be compensated to obtain a multimodal signal with a unified time base, which may include the following steps: The first step is to perform a subtraction elimination operation on the aforementioned path deviation parameters to be compensated, obtaining the subtraction elimination result. This subtraction elimination result can be the result of subtracting the equipment deviation component from the original total path delay.

[0082] The second step involves removing the asymmetric jitter information generated by intermediate device forwarding based on the above subtraction elimination results, thereby obtaining the link propagation delay information. This asymmetric jitter information can be due to uncertain delays caused by fluctuations in device queue scheduling and processing time.

[0083] The third step is to perform time-axis translation processing on the frequency-corrected multimode signal based on the above-mentioned link propagation delay information to obtain the translated multimode signal.

[0084] In practice, the link propagation delay information can be added to the original timestamp of the frequency-corrected multimode signal to determine whether the signal is shifted forward or backward on the time axis, thus obtaining the shifted multimode signal.

[0085] The fourth step involves performing phase alignment processing on the translated multimodal signals to obtain multimodal signals with a unified time base. This phase alignment processing can be used to check whether the multimodal signals maintain the correct phase relationship after translation. In practice, cross-correlation analysis can be performed on the translated multi-channel signals, and fine-tuning can be performed if slight deviations exist to obtain multimodal signals with a unified time base.

[0086] Step 106: Based on the physical correlation characteristics between signals, the above-mentioned time-base unified multimodal signals are resampled to obtain time-synchronized multimodal signals.

[0087] In some embodiments, the aforementioned executing entity can resample the aforementioned time-referenced unified multimodal signal based on the physical correlation characteristics between signals to obtain a time-synchronized multimodal signal. The aforementioned physical correlation characteristics can be a physical causal relationship between different modal signals (e.g., vibration and temperature). For example, mechanical vibration can cause a local temperature rise; when the vibration energy continuously increases in a specific frequency band (e.g., 100-500Hz), it may indicate increased friction, and the temperature will show an upward trend. The aforementioned resampling can be based on reconstructing the temperature signal from the spectrum of the vibration signal.

[0088] In some optional implementations of certain embodiments, the execution entity may resample the time-base unified multimodal signal based on the physical correlation characteristics between signals to obtain a time-synchronized multimodal signal, which may include the following steps: The first step involves performing spectral feature analysis on the first signal in the aforementioned multimodal signal with a unified time base to obtain spectral feature distribution information. This first signal can be a high-frequency sampled signal (e.g., a vibration signal). The spectral feature distribution information can be the distribution of signal energy along the frequency axis, the dominant frequency component, or the frequency band energy.

[0089] The second step is to determine the cable's operating status information based on the aforementioned spectral characteristic distribution information. This operating status information may include, but is not limited to, a stable state, a state with slight disturbance, and a state of severe vibration. Each cable operating status corresponds to different vibration energy characteristics.

[0090] For example, if the vibration signal energy increases in the 100-500Hz frequency band, it is determined that the cable is in a state of severe vibration.

[0091] The third step is to determine the physical evolution model corresponding to the above-mentioned cable operating status information based on the physical correlation characteristics between the signals. This physical evolution model can be a mathematical model describing temperature changes over time.

[0092] As an example, if the cable's operating status is stable, a low-order polynomial model is used to describe the temperature change. For severe vibration conditions, a simplified model based on the heat conduction equation is used. The parameters of this heat conduction equation are related to the vibration integral value.

[0093] The fourth step involves sampling the second signal in the aforementioned multimodal signal with a unified time base to obtain a sampling point sequence. This second signal can be a low-frequency sampling signal (e.g., a temperature signal). The sampling point sequence can be the value of the temperature signal at the original sampling time.

[0094] Fifth, based on the above physical evolution model and the above sampling point sequence, determine the signal sampling model. The above signal sampling model can be a function used to describe the continuous change of the second signal on the time axis.

[0095] As an example, a temperature signal sequence could be Sampling points can be The current temperature sampling point can be used. and the next moment temperature sampling point As boundary conditions, continuous functions can be fitted. .

[0096] Step 6: Based on the above signal sampling model, resample the first signal to obtain a resampled sequence of the first signal. As an example, this can be done at the time points of the vibration signal. right Resampling was performed to obtain a temperature sequence aligned with the vibration signal. .

[0097] Step 7: Integrate the resampled sequence and the first signal to obtain a time-synchronized multimodal signal. For example, the resampled temperature sequence can be used... With vibration signal sequence By merging them in the time dimension, a time-synchronized multimodal signal is obtained.

[0098] The above-described embodiments of this disclosure have the following beneficial effects: The high-voltage cable heterogeneous signal synchronization method of some embodiments of this disclosure eliminates physical phase mismatch and clock deviation, ensures alignment accuracy in long-distance monitoring scenarios, and improves the data quality for high-voltage cable operation status fault diagnosis. Specifically, the reasons for heterogeneous signal asynchrony are: phase distortion, clock frequency drift accumulation error, and uncertain delay introduced by intermediate network devices during the alignment process. Based on this, the high-voltage cable heterogeneous signal synchronization method of some embodiments of this disclosure first collects the hardware timestamps of incoming and outgoing packets to obtain raw timestamp information. Thus, the timestamps of data packets leaving or arriving (nanosecond level) are directly obtained at the driver layer, avoiding random delays introduced by the application layer and protocol stack processing, and providing a raw time reference for subsequent multi-modal signal time synchronization. Second, based on the above raw timestamp information, the raw time references for each heterogeneous signal are constructed to obtain initially aligned multi-modal signals. Thus, the multi-source signals distributed across different sensor nodes are unified to the same time coordinate system, eliminating the initial time deviation of the clock sources of each node, and laying the foundation for subsequent step-by-step compensation. Then, link transmission delay compensation is performed on the initially aligned multimode signals to obtain delay-compensated multimode signals. This suppresses link transmission delay fluctuations caused by network load changes and queue scheduling factors, avoiding the shortcomings of fixed delay compensation methods that cannot adapt to dynamic network environments. Furthermore, clock frequency drift correction is performed on the delay-compensated multimode signals to obtain frequency-corrected multimode signals. Thus, by calculating the frequency drift factor online and dynamically adjusting the local time accumulation compensation value, long-term clock accumulation errors caused by crystal oscillator aging or environmental temperature changes are suppressed, ensuring the synchronization stability of multimode signals under long-term operation. Next, intermediate network device dwell time correction is performed on the frequency-corrected multimode signals to obtain time-base unified multimode signals. Thus, by measuring and compensating for the dwell time of messages in each switching device through a transparent clock mechanism, delay jitter introduced by device forwarding uncertainty in multi-hop network topologies is eliminated, improving time synchronization accuracy in long-distance transmission scenarios. Finally, based on the physical correlation characteristics between signals, the time-base unified multimode signals are resampled to obtain time-synchronized multimode signals. Therefore, the traditional fixed-frequency interpolation method was abandoned, and a resynchronization model was constructed using the physical causal relationship between signals such as vibration and temperature, thus avoiding phase distortion problems. This improved the data quality for fault diagnosis of high-voltage cable operation status.

[0099] Based on the same inventive concept, embodiments of this application provide a high-voltage cable heterogeneous signal synchronization system, comprising: The acquisition module is used to acquire hardware timestamps; A memory for storing a program for a method of synchronizing heterogeneous signals in high-voltage cables; The processor can load and execute programs in memory to implement a method for synchronizing heterogeneous signals in high-voltage cables.

[0100] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device, and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0101] This application provides a computer-readable storage medium storing a computer program that can be loaded by a processor and executed as a method for synchronizing heterogeneous signals in high-voltage cables.

[0102] Computer storage media include, for example, USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, optical disks, and other media that can store program code.

[0103] Based on the same inventive concept, this application provides a smart terminal, including a memory and a processor, wherein the memory stores a computer program that can be loaded and executed by the processor to provide a method for synchronizing heterogeneous signals in high-voltage cables.

[0104] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device, and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0105] The above are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Any feature disclosed in this specification (including the abstract and drawings) may be replaced by other equivalent or similar features unless specifically stated otherwise. That is, unless specifically stated otherwise, each feature is only one example of a series of equivalent or similar features.

Claims

1. A method for synchronizing heterogeneous signals in high-voltage cables, characterized in that, include: Collect the hardware timestamps of incoming and outgoing messages to obtain the raw timestamp information; Based on the original timestamp information, the original time references for each heterogeneous signal are constructed to obtain the initially aligned multimodal signal; The initially aligned multimode signal is subjected to link transmission delay compensation to obtain a delay-compensated multimode signal; The delay-compensated multimode signal is subjected to clock frequency drift correction to obtain a frequency-corrected multimode signal; The intermediate network device dwell time is corrected by performing intermediate network device dwell time correction on the frequency-corrected multimode signal to obtain a multimode signal with a unified time base; Based on the physical correlation characteristics between signals, the multimodal signals with unified time references are resampled to obtain time-synchronized multimodal signals.

2. The method for synchronizing heterogeneous signals in high-voltage cables according to claim 1, characterized in that, The step of performing link transmission delay compensation on the initially aligned multimode signal to obtain the delay-compensated multimode signal includes: Establish the first and second prediction equations; Acquire the interactive data of periodic message exchanges between the master clock node and the slave clock node; Based on the interaction data, the observed link transmission delay is determined; Based on the first prediction equation, determine the transmission delay status information at the current moment; Based on the link transmission delay observation and the second prediction equation, the transmission delay status information at the current moment is corrected to obtain the transmission delay estimate. Based on the transmission delay estimate, the original timestamps of the initially aligned multimodal signals are compensated to obtain the delay-compensated multimodal signals.

3. The method for synchronizing heterogeneous signals in high-voltage cables according to claim 2, characterized in that, The step of correcting the transmission delay state information at the current moment based on the link transmission delay observation and the second prediction equation to obtain the transmission delay estimate includes: Based on the transmission delay state information at the current moment and the second prediction equation, the observed prediction value at the current moment is determined; Based on the observed link transmission delay and the predicted observation, the observation residual is determined; Obtain the prediction error covariance information corresponding to the transmission delay status information at the current moment; Based on the second prediction equation and the prediction error covariance information, the corrected weighting coefficients are determined; Based on the corrected weighting coefficients and the observation residuals, the transmission delay state information at the current moment is weighted and corrected to obtain the transmission delay estimate.

4. The method for synchronizing heterogeneous signals in high-voltage cables according to claim 1, characterized in that, The step of performing clock frequency drift correction on the delay-compensated multimode signal to obtain a frequency-corrected multimode signal includes: Obtain the transmission timestamps of at least two consecutive synchronization messages at the master clock and the reception timestamps at the slave clock; Based on the sending timestamp and the receiving timestamp, the sending time interval of adjacent messages in the at least two consecutive synchronization messages is determined, and the master clock physical time interval and slave clock local time interval are obtained. The ratio of the master clock physical time interval to the slave clock local time interval is calculated to obtain the frequency drift factor, which characterizes the master-slave clock frequency deviation. Based on the preset update time and the current time, determine the time span information; The frequency drift factor and the time span information are subjected to error accumulation processing to obtain the dynamic cumulative offset; The total logic clock compensation is obtained by summing the dynamic cumulative offset and the preset initial offset. Based on the total compensation amount of the logic clock, the original hardware clock reading of the delayed-compensated multimodal signal is corrected to obtain the logic clock output time series. The timestamps of the delay-compensated multimode signal are remapped using the time series output by the logic clock to obtain the frequency-corrected multimode signal.

5. The method for synchronizing heterogeneous signals in high-voltage cables according to claim 1, characterized in that, The step of performing intermediate network device dwell time correction on the frequency-corrected multimodal signal to obtain a time-base unified multimodal signal includes: Determine the moment when the synchronization event message enters the physical port of the intermediate network device to obtain the ingress capture trigger signal; The entry capture action is triggered based on the entry capture trigger signal; Determine the entry hardware timestamp generated by the entry capture action; The moment when the synchronization event message leaves the physical port of the intermediate network device is detected to obtain the egress capture trigger signal; The exit capture action is triggered based on the exit capture trigger signal; Determine the exit hardware timestamp generated by the exit capture action; The original dwell time of the synchronization event message is obtained by subtracting the exit hardware timestamp and the entry hardware timestamp. Based on a preset clock synchronization protocol, the original dwell time is processed by a correction conversion to obtain correction increment information; The correction increment information is written into the correction field of the synchronization event message to obtain the first synchronization message, wherein the first synchronization message is a synchronization message carrying local forwarding delay information. Based on the frequency-corrected multimode signal, the first synchronization message is subjected to path concatenation processing to obtain the path deviation parameters to be compensated. Based on the path deviation parameters to be compensated, the delayed observations are decoupled to obtain a multimodal signal with a unified time reference.

6. The method for synchronizing heterogeneous signals in high-voltage cables according to claim 5, characterized in that, The step of performing path concatenation processing on the first synchronization message based on the frequency-corrected multimode signal to obtain the path deviation parameters to be compensated includes: In response to the first synchronization message being cascaded and forwarded through multiple intermediate network devices, the correction field information of the first synchronization message is determined; Determine the total residence time of the corrected field information; Based on the network topology, end-to-end path mapping is performed on the total dwell time to obtain device deviation information on the transmission path. Determine the sampling point sequence corresponding to the timestamp of the frequency-corrected multimodal signal and the first synchronization message; By temporally correlating the device deviation information with the sampling point sequence, the path deviation parameters to be compensated are obtained.

7. The method for synchronizing heterogeneous signals in high-voltage cables according to claim 5, characterized in that, The step of decoupling the delayed observations based on the path deviation parameters to be compensated to obtain a multimodal signal with a unified time reference includes: Based on the path deviation parameter to be compensated, a subtraction elimination operation is performed on the delayed observation value to obtain the subtraction elimination result; Based on the subtraction elimination result, the asymmetric jitter information generated by the intermediate device forwarding is eliminated to obtain the link propagation delay information; Based on the link propagation delay information, the frequency-corrected multimode signal is subjected to time-axis translation processing to obtain the translated multimode signal; The translated multimodal signal is phase aligned to obtain a multimodal signal with a unified time base.

8. The method for synchronizing heterogeneous signals in high-voltage cables according to claim 1, characterized in that, The step of resampling the time-referenced multimodal signal to obtain a time-synchronized multimodal signal based on the physical correlation characteristics between signals includes: The first signal in the multimodal signal with the unified time base is subjected to spectral feature analysis to obtain spectral feature distribution information; Based on the aforementioned spectral feature distribution information, the cable operating status information is determined; Based on the physical correlation characteristics between the signals, the physical evolution model corresponding to the cable operating status information is determined; The second signal in the multimodal signal with the unified time base is sampled to obtain a sampling point sequence; Based on the physical evolution model and the sampling point sequence, the signal sampling model is determined; Based on the signal sampling model, the first signal is resampled to obtain a resampled sequence of the first signal; The resampled sequence and the first signal are integrated to obtain a time-synchronized multimodal signal.

9. A high-voltage cable heterogeneous signal synchronization system, characterized in that, include: The acquisition module is used to acquire hardware timestamps; A memory for storing a program for a high-voltage cable heterogeneous signal synchronization method as described in any one of claims 1 to 8; The processor and the program in the memory can be loaded and executed by the processor to implement the high-voltage cable heterogeneous signal synchronization method as described in any one of claims 1 to 8.

Images