Partial discharge multichannel signal real-time synchronous acquisition method based on edge calculation

Abstract

The invention relates to the technical field of edge calculation application and partial discharge detection, in particular to a partial discharge multichannel signal real-time synchronous acquisition method based on edge calculation. According to the method, the functions of multi-channel signal acquisition, time synchronization, feature analysis, hierarchical storage and the like are integrated at edge nodes, asynchronous acquisition, unified timestamp marking and data synchronous fusion of multiple types of partial discharge signals are realized, and extraction of multi-dimensional feature parameters such as time domain, frequency domain and energy and intelligent event discrimination are locally completed. For abnormal signals, the system realizes classified storage and remote uploading after encryption and compression processing; and for normal signals, dynamic management is carried out through circular caching. According to the method, the accuracy and efficiency of multichannel signal synchronous acquisition are remarkably improved, the data safety and the system adaptive capacity are enhanced, and the method is suitable for real-time online monitoring and intelligent diagnosis of partial discharge of power equipment.

Description

Technical Field

[0001] The present invention relates to the field of edge computing applications and partial discharge detection technology, and specifically to a method for real-time synchronous acquisition of partial discharge multi-channel signals based on edge computing. Background Art

[0002] During long-term operation, the deterioration and potential defects of the insulation system of high-voltage power equipment such as transformers and GIS can easily lead to partial discharge (PD). PD is an electrical breakdown that occurs in a localized area of ​​an insulator. It does not immediately cause overall insulation failure, but if PD persists, it can seriously threaten equipment safety and shorten its service life. Therefore, online, real-time, and high-precision monitoring of PD signals is an important means of ensuring the safe operation of power systems. Existing online PD monitoring solutions often use multiple types of sensors (such as UHF, ultrasonic, and geostationary waves) to collect different physical quantities, and perform preliminary data analysis and alarms through embedded processors or SOCs. However, their multi-channel signal acquisition typically relies on centralized data synchronization and processing, which is limited by bandwidth, latency, and synchronization accuracy, making it difficult to meet the needs of efficient and accurate monitoring in complex environments.

[0003] In recent years, with the development of edge computing technology, moving data processing capabilities to the acquisition terminal has become an important trend to improve the real-time and intelligent level of monitoring systems. Edge computing can perform synchronous correction, feature extraction and preliminary judgment of multi-channel acquisition signals locally, significantly reducing the pressure of data transmission and improving the system response speed. However, existing technologies still have certain limitations in the high-precision synchronous acquisition and real-time processing of multi-channel signals, especially in terms of multi-source heterogeneous signal fusion, synchronization accuracy and anti-interference ability. A unified and efficient solution has not yet been formed. The above-mentioned existing technologies have shortcomings. How to achieve high-precision real-time synchronous acquisition of multi-channel signals of partial discharge under the edge computing architecture, improve the synchronization, real-time performance and data processing efficiency of multi-source heterogeneous signals, so as to meet the high reliability and intelligent requirements of online monitoring of power equipment is a problem that needs to be solved at present.

[0004] To this end, a real-time synchronous acquisition method of multi-channel partial discharge signals based on edge computing is proposed. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for real-time synchronous acquisition of multi-channel partial discharge signals based on edge computing to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions: Adopting an asynchronous sampling-synchronous fusion signal acquisition architecture, each channel asynchronously acquires the partial discharge signal of the electrical equipment under test through an independent high-sampling-rate analog-to-digital converter. The edge node uses a clock and digital processing chip to add a time stamp to the asynchronous data, reconstruct the timing, synchronize the partial discharge signals of each channel, compensate for inter-channel delay and jitter, and output the partial discharge signal data. Based on the adaptive synchronization algorithm, the multi-channel partial discharge signals are locally time-aligned and time-series-fused to output a multi-channel signal group; Perform multi-parameter feature analysis on multi-channel signal groups, generate multi-dimensional feature sequences through digital filtering and sliding window segmentation, and identify the type and time of partial discharge events through correlation analysis of multi-channel signal groups; Adopting an event-driven data processing and storage mechanism, based on real-time feature identification, abnormal partial discharge signals are classified, stored, and uploaded by event type. Only key information of normal signals is recorded locally, and signal data is compressed and securely managed in a hierarchical manner. Edge nodes regularly optimize synchronization parameters and discrimination thresholds based on historical detection and operation data.

[0007] Preferably, the asynchronous sampling-synchronous fusion signal acquisition architecture includes: Each signal acquisition channel is connected to an independent high-sampling rate analog-to-digital converter to collect partial discharge signals from the electrical equipment under test; the edge node is equipped with a field-programmable gate array as a digital processing chip, and a high-precision clock is used to add a timestamp to the partial discharge signal data collected from each channel; during the synchronous timing reconstruction process, the data of each channel is compared with the reference trigger signal, and the partial discharge signal data stream is adjusted based on the digital delay line to compensate for the acquisition start time difference and synchronization error between channels; during the acquisition process, the acquisition system detects short-term clock drift and uses a phase-locked loop circuit or digital filtering algorithm to stabilize the clock; after the multi-channel partial discharge signal data is synchronously reconstructed, the cyclic redundancy check method is used to verify the time stamp integrity, and the partial discharge test data with a unified time stamp is output.

[0008] Preferably, the multi-channel signal group includes: A high-precision timestamp is added to the acquisition moment of the partial discharge signal of each channel. According to the sampling delay and clock deviation parameters between the channels, a dynamic adjustment strategy is adopted to locally time-align the acquired signals. Through the timing fusion processing method of interpolation and resampling, the partial discharge signals of different channels are synchronized to generate a multi-channel signal group with a unified time base and timing consistency.

[0009] Preferably, the multi-parameter feature analysis includes: For the synchronized multi-channel signal, a digital filter is used to denoise the original signal and correct the baseline drift, and then the signal is segmented through a sliding window algorithm; in the time domain feature extraction process, a zero-crossing detection circuit is used to identify the extreme points and amplitude changes of the pulse signal, and pulse counting is completed through a gated shaping circuit; in the frequency domain feature analysis, a fast Fourier transform module is used to decompose the signal spectrum to obtain the energy distribution and main frequency information of each frequency band; in the energy feature calculation part, an integration circuit is used to integrate the signal amplitude within a single pulse width to obtain the single pulse energy and cycle average energy; in the spectral feature extraction, the harmonic analysis circuit is used to extract the harmonic components, and the Hilbert transform module is used to obtain the instantaneous frequency; the correlation analysis of the multi-channel signal uses a correlation coefficient calculation module to perform point-by-point correlation comparison on the signal waveform of each channel, and combined with the timestamp information, output the judgment result and corresponding time of the local discharge event.

[0010] Preferably, the event-driven data processing and storage mechanism includes: For partial discharge signal data judged as abnormal, the acquisition system integrates an encryption circuit module inside the local processing unit to perform block encryption on the original signal data; a differential compression algorithm module is used to compare the abnormal event data with the set reference signal template, and only the signal segments with obvious differences from the template are extracted and compressed. The compressed partial discharge signal data and the encryption key are stored together in the local non-volatile storage chip, and uploaded to the remote monitoring platform through a wired or wireless communication interface according to the event type group; for partial discharge signal data that is not judged as abnormal, the system adopts a circular cache storage module, using a ring storage structure to only save the key information of the signal. Expired data is automatically overwritten when the storage space reaches the preset threshold.

[0011] Preferably, the optimization of synchronization parameters and discrimination thresholds includes: The built-in statistical analysis module periodically calls the original records of partial discharge signals and abnormal event records in the historical acquisition data storage unit, and uses hardware counters to count the frequency of various abnormal events and corresponding detection results. The synchronization parameters are corrected by weighted average of the sampling delay and clock phase offset parameters of multi-channel signals through digital circuits. The feature discrimination threshold is based on historical false alarm and missed alarm data, and the signal amplitude, pulse width and frequency threshold are dynamically adjusted using a lookup table.

[0012] Preferably, the edge node includes: Local signal acquisition module, time synchronization module, feature analysis module, data hierarchical storage module and various physical communication interface modules; processor and digital signal processing chip, used for local asynchronous acquisition, time stamping, synchronous fusion, feature parameter extraction and event discrimination processing of multi-channel partial discharge signals; data is stored hierarchically in local storage units after feature analysis and event discrimination; edge nodes are used to perform online monitoring of the insulation status of the electrical equipment under test, conduct real-time analysis and fault trend prediction of partial discharge signals, automatically identify equipment anomalies and insulation degradation, and generate diagnostic reports and alarm information based on the analysis results, which are remotely transmitted to the monitoring center via the communication interface to support intelligent detection and remote management of the operating status of electrical equipment; edge nodes perform time synchronization and data exchange with other edge nodes through hardware synchronization signal lines and high-speed communication buses, and support multi-node collaborative networking; edge nodes are equipped with physically isolated power management modules to allocate independent power to each functional module.

[0013] Compared with the prior art, the present invention has the following beneficial effects: 1. For the synchronous acquisition of multi-channel partial discharge signals, an asynchronous sampling-synchronous fusion signal acquisition architecture is employed. Each channel acquires different types of partial discharge signals through independent high-sampling-rate analog-to-digital converters. High-precision clock marking, digital delay line compensation, and adaptive synchronization algorithms are then used within edge nodes to perform local microsecond-level time alignment and timing reconstruction of the multi-channel asynchronous signals. This overcomes the bottlenecks of traditional centralized synchronization methods limited by bandwidth, latency, and wiring complexity. Even in multi-node distributed networks and complex field environments, the system can still ensure high-precision synchronization and unified timing of each channel's signals, significantly improving system scalability and monitoring reliability.

[0014] 2. In terms of signal processing and intelligent identification, a multi-parameter feature analysis module is integrated based on edge nodes to simultaneously extract time, frequency, energy, and spectral features from synchronized multi-channel signals. This solution, combined with multi-channel signal correlation analysis, event template matching, and dynamic time warping algorithms, intelligently identifies abnormal partial discharge events. This solution can effectively identify atypical partial discharge signals such as weak amplitudes and waveform distortion, improving the accuracy and robustness of abnormal event detection under complex operating conditions, significantly outperforming existing solutions that rely on single features or simple identification methods.

[0015] 3. Regarding data management and system expansion, an event-driven data processing and hierarchical storage mechanism is implemented at the edge nodes. Signal data identified as abnormal in real time is differentially compressed, encrypted, and uploaded in groups by event type. For normal signals, only critical information is stored and cached in a circular manner. This solution effectively reduces local storage requirements and remote transmission bandwidth pressure, while ensuring the integrity and security of abnormal event data, meeting the practical needs of power plants for efficient data management and information security.

[0016] 4. In terms of system reliability and intelligent adaptive capabilities, the edge node of the present invention adopts modular hardware design and physically isolated power management. Different functional modules are independently powered and isolated from each other, achieving highly stable operation of acquisition, processing, storage and other units. At the same time, the system integrates adaptive synchronization parameters and discrimination threshold optimization mechanisms based on historical operation data, and can dynamically adjust the synchronization algorithm parameters and abnormal discrimination thresholds according to the on-site environment and equipment status. This solution not only significantly improves the system's operational reliability and anti-interference capabilities in harsh power environments such as high voltage and strong electromagnetic interference, but also enhances the system's self-learning and self-optimization capabilities. The system can continuously optimize synchronization accuracy and abnormality detection performance, effectively reduce false alarms and missed alarms, and reduce manual maintenance and debugging workload. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a flow chart of the method for real-time synchronous acquisition of multi-channel partial discharge signals based on edge computing proposed in the present invention; Figure 2 This is a block diagram of the asynchronous sampling-synchronous fusion structure principle proposed in the embodiment of the present invention; Figure 3 This is a functional block diagram of the edge node proposed in the embodiment of the present invention. DETAILED DESCRIPTION

[0018] 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.

[0019] See also Figures 1 to 3 The present invention provides a method for real-time synchronous acquisition of multi-channel partial discharge signals based on edge computing. The technical solution is as follows: Example 1 This embodiment is applied to 110kV substation GIS equipment. During operation, insulation aging, defects, and other problems may cause partial discharge.

[0020] For multi-channel partial discharge signal acquisition, an asynchronous sampling and synchronous fusion signal acquisition architecture is employed. Various types of partial discharge sensors, including UHF, ultrasonic, and geoelectric waves, are installed at each monitoring point. The analog signals output by each sensor are asynchronously acquired via independent high-sampling-rate analog-to-digital converters. The sampling systems for different channels each have independent sampling clocks, and the raw signals are not perfectly aligned on the time axis. All collected raw data first enters the FPGA chip configured in the edge node. The FPGA integrates a high-precision clock management module, which generates a standard time base using a temperature-compensated crystal oscillator or an external high-precision timing signal. The FPGA then adds a timestamp generated by this high-precision clock to all incoming acquisition channel data in real time. This timestamp can achieve microsecond or higher accuracy, ensuring that each frame of data accurately records its acquisition moment. Then, the FPGA relies on the set reference trigger signal to compare the data streams collected by each channel, automatically detect and calculate the sampling start time difference between channels caused by factors such as hardware delay, cable length, and front-end response characteristics. With the built-in digital delay line, the timing of each channel of collected data is dynamically adjusted to align them on a unified time axis. In order to overcome the short-term drift of the clock signal that may occur during long-term operation, the system also introduces a phase-locked loop circuit to continuously lock the reference clock, and uses a digital filtering algorithm when necessary to eliminate high-frequency noise or periodic interference to ensure the high stability and consistency of the clock signal. A synchronization error self-calibration module based on multi-channel cross-correlation analysis is integrated into the edge node. This module consists of an event trigger unit, a cross-correlation calculation unit, and a parameter feedback adjustment unit. The event trigger unit monitors the partial discharge signals collected by each channel in real time and automatically captures typical event waveforms. The cross-correlation calculation unit performs cross-correlation analysis on the event waveform sequences of different channels to accurately measure the actual time delay between each channel. The parameter feedback adjustment unit dynamically adjusts the time-scale alignment parameters or digital delay compensation in the synchronization fusion algorithm based on the correlation results, automatically correcting any remaining minor synchronization errors. This enables high-precision synchronization of multi-channel acquired data at the millisecond or even microsecond level, significantly improving the channel time consistency and overall synchronization performance of the system under complex conditions such as high noise and strong interference. Before output, the synchronized multi-channel data is verified for integrity and consistency using a cyclic redundancy check (CRC) method on the timestamp fields of all data frames. If a timestamp anomaly or loss is detected, a redundancy mechanism can be used to recover or provide an alarm. Finally, the qualified multi-channel signals are output at a unified time scale, forming a highly consistent synchronized multi-channel dataset that can be used for subsequent analysis.

[0021] This method enables high-precision synchronous fusion of original asynchronously sampled signals during the acquisition of multi-type, multi-channel partial discharge signals. This significantly improves the temporal consistency and time-scale accuracy of multi-channel signals, providing reliable foundational data for subsequent applications such as spatiotemporal correlation analysis of partial discharge events, discharge type identification, and fault location. This method effectively reduces the technical barriers to multi-channel synchronization that often rely on expensive distributed clocks or complex wiring, improves the flexibility and engineering adaptability of edge-side online data fusion, ensures the stability and integrity of field monitoring data, and provides strong support for the practical application of online partial discharge monitoring technology.

[0022] Furthermore, under a signal acquisition architecture that employs asynchronous sampling and synchronous fusion, multiple acquisition channels independently sample the partial discharge signal. Each acquisition channel then timestamps each set of collected signal data based on a local high-precision clock. Physically, the sampling times are not strictly aligned between the acquisition channels, thus operating in an asynchronous sampling mode. After the raw signals collected by all channels and their timestamps are uploaded to the edge computing node, the channel data is first aggregated and analyzed using the timestamps to extract the sampling delay and clock deviation information for each signal segment. Based on these parameters, the edge node adjusts the time axis of the collected signal data: channel data with sampling delays is shifted according to the timestamp sequence to align them on the same reference time axis. For cases with large clock deviations, timebase calibration is performed by compensating the timestamps. For calibrated signal data, if sampling points for each channel are missing or misaligned within a unified time base, interpolation methods such as linear interpolation and spline interpolation are used to fill in the missing data. For channels with inconsistent sampling rates or data volume differences due to clock drift, a resampling algorithm is applied to unify the sampling rate and sampling point distribution. After the above-mentioned local time alignment and timing fusion processing, the signal data from different channels are synchronized to a unified time base and organized into multi-channel signal groups in chronological order, achieving synchronous fusion of the data structure.

[0023] Through the above method steps, it is possible to ensure that the multi-channel asynchronous sampling data is synchronized and fused with high precision before subsequent analysis, so that the multi-channel signal group is highly consistent in time reference and signal timing, providing an accurate basis for joint analysis, feature comparison and other operations of partial discharge signals, and improving the reliability and scientificity of multi-channel data statistical analysis and abnormal signal detection.

[0024] Furthermore, for the synchronized multi-channel signal group, a digital filter is initially used to process the original signal sequence to remove noise, filter out power frequency interference and high-frequency noise, and automatically correct the signal baseline with the help of the low-frequency component of the filter to eliminate the impact of background drift on feature extraction. The denoised and corrected signal stream is input into the sliding window segmentation algorithm. The sliding window moves on the signal sequence according to the preset step size, dividing the long time series into multiple continuous signal segments, each of which serves as the basic analysis unit for subsequent feature extraction. In the time domain analysis stage, the zero-crossing detection algorithm is used to automatically determine the extreme points of each pulse signal, and the maximum amplitude and waveform change amplitude of each pulse are identified and counted. The gated shaping algorithm is combined to count the pulses of the signal segments to achieve accurate quantification of the frequency of partial discharge pulses. After entering the frequency domain analysis stage, the fast Fourier transform algorithm is used to perform spectral decomposition on the segmented signal to obtain the energy distribution spectrum of different frequency bands and the signal main frequency information. In the energy parameter extraction process, an integration algorithm is used to integrate the signal amplitude within a single pulse width to obtain the single pulse energy intensity. At the same time, the energy mean within each signal cycle is statistically analyzed to reflect the intensity changes of partial discharge events. In spectral feature analysis, the harmonic component extraction algorithm is combined to separate the amplitudes of each order of harmonics, and the Hilbert transform method is used to extract the instantaneous frequency characteristics of the signal, improving the ability to identify abnormal pulses and complex waveforms. For correlation analysis between multi-channel signals, an event-triggered similarity matching module can be integrated. That is, after completing the synchronization and denoising operations on the multi-channel signal segments, the system monitors the time window of suspected partial discharge events in the signal, automatically intercepts the corresponding signal segments of each channel, and extracts multi-dimensional feature parameters for these segments, including pulse amplitude, rising and falling edge slopes, main frequency components, energy distribution, etc. The system can then perform template matching based on preset event templates, or use a dynamic time warping algorithm to perform nonlinear alignment and similarity scoring on the signal segments of each channel to identify synchronous or asynchronous partial discharge events between channels. This method can achieve high-precision identification of the synchronization of multi-channel partial discharge events in complex scenarios where multi-channel signals contain nonlinear and non-stationary characteristics such as waveform distortion, amplitude variation, and phase drift, improving the ability to collaboratively analyze abnormal events and spatially locate them. Using a correlation coefficient calculation method, signal segments from different channels are compared point by point on the same time basis. Combined with timestamp data, this method outputs the synchronization identification results and the time of occurrence of partial discharge events, providing support for spatial positioning and multi-channel collaborative analysis.

[0025] Through the above-mentioned multi-parameter feature analysis method, comprehensive feature extraction of partial discharge signals can be performed from multiple dimensions such as time domain, frequency domain, energy, spectral characteristics and multi-channel correlation, realizing multi-level discrimination and quantitative analysis of complex partial discharge events, improving the accuracy and robustness of partial discharge event identification, and providing diverse and accurate basic data for subsequent fault diagnosis, status assessment and trend prediction.

[0026] Furthermore, for multi-channel partial discharge signal data that has been synchronized and feature analyzed, an event-driven data processing and storage mechanism is employed. First, within the edge processing unit, the collected signals are automatically classified as abnormal or normal based on a multi-parameter feature extraction and discrimination algorithm. For signal data identified as abnormal, a locally integrated encryption circuit module is dispatched to encrypt the original signal data block by block using packet or streaming encryption techniques. Each data block is assigned an independent key during the encryption process, and the key is dynamically generated and managed by an internal security module. This ensures that even in the event of data interception, the original signal cannot be restored, effectively preventing data leakage. After encryption, the data is fed into a differential compression algorithm module for secondary processing. This module compares the abnormal signal with the template point by point or segment by segment using a reference signal template as a benchmark. The algorithm dynamically identifies intervals that differ significantly from the template and only extracts and efficiently compresses data from these intervals. The compression algorithm can employ methods such as threshold-based sparse coding, local adaptive dictionary methods, or event-triggered variable-length coding to further reduce data volume. The compressed differential data and generated keys are archived and stored in a local non-volatile chip by event category. They are then uploaded to a remote monitoring platform via a wired (e.g., Ethernet) or wireless (e.g., 4G / 5G, Wi-Fi) communication interface, grouped by abnormal event type and priority. The system supports multiple data transmission protocols and failure retransmission mechanisms to ensure that abnormal event data arrives in real time and reliably. For normal signal data that has not been identified as abnormal, the system adopts a circular cache mechanism, using a circular buffer to temporarily store the signal's main statistical parameters and characteristic values ​​(e.g., pulse count, main frequency, energy mean, etc.) in chronological order. The original waveform data is not persistently stored. Once the cache capacity reaches the set threshold, the earliest data is automatically overwritten by subsequent new data, enabling dynamic updates and spatial reuse. The circular cache module also supports rapid retrieval and export of key information by time window, facilitating subsequent data tracing and trend analysis. A data buffering strategy with priority scheduling is further introduced. This system automatically and dynamically manages the local data buffer based on the type and severity of partial discharge signal events. When a major or rare abnormal event is detected, more storage space and bandwidth resources are prioritized to ensure that such data is fully preserved and quickly uploaded to the remote monitoring platform. For common or low-risk events, a low-priority approach is adopted, with only brief statistical parameters saved or the original data transmitted delayed. This mechanism enables intelligent scheduling of data transmission and storage for different event categories within limited edge computing and storage resources, improving the response efficiency of critical abnormal events while reducing overall system data redundancy and resource utilization.

[0027] Through an event-driven data processing and storage mechanism, this method not only ensures the secure local encryption and efficient, differentiated compression of abnormal partial discharge signals, effectively improving data security and storage utilization, but also enables timely remote reporting of key abnormal events through event type grouping and prioritized upload, ensuring the monitoring system's rapid response to significant electrical faults. The use of key information caching and automatic overwriting for normal signals significantly reduces data redundancy and local storage pressure, facilitating long-term continuous system operation and large-scale, multi-point deployment, providing an efficient, scalable, and low-cost technical foundation for intelligent partial discharge monitoring and condition assessment of power equipment.

[0028] In the real-time synchronous acquisition process of multi-channel partial discharge signals in edge computing, a dynamic optimization mechanism for synchronization parameters and feature discrimination thresholds has been specifically designed. In specific implementation, the local historical data management module is periodically called to automatically retrieve and retrieve previously collected partial discharge raw signal data and abnormal event identification records. The built-in statistical analysis module, in conjunction with the hardware counter, performs classified statistical analysis on the frequency of occurrence of various abnormal events and the corresponding discrimination accuracy. For example, the statistical module can separately record the number of false alarms, missed alarms, and accurate discriminations within different time windows, and establish associations with historical environmental parameters, equipment operating status, and other information. To address the synchronization issues of multi-channel signals, the system regularly analyzes the synchronization errors between timing signals. Based on the statistically derived core parameters such as sampling delay and clock phase offset, the sampling timing signals of each channel are weighted averaged and corrected with the help of digital circuits. The synchronization parameters are dynamically adjusted to ensure the timing consistency of the data collected by each channel. For the optimization of feature discrimination thresholds, we continuously track historical discrimination errors and on-site false alarms and missed alarms. We rely on lookup tables to automatically and dynamically adjust parameters such as signal amplitude thresholds, pulse width thresholds, and frequency thresholds. The lookup table sets a series of threshold intervals based on changes in the actual acquisition environment and the distribution of historical false positives. It automatically selects the optimal threshold based on the latest statistical results and updates the feature discrimination criteria in real time to adapt to the complex and changing operating environment on site.

[0029] By using the adaptive optimization method of the synchronization parameters and discrimination thresholds mentioned above, the synchronization accuracy of each channel signal can be continuously improved based on the historical data and event recognition performance of the site, the acquisition timing error and data offset between channels can be effectively reduced, and the feature discrimination threshold can be dynamically adjusted to make the identification of partial discharge events more accurate and robust. This method has given powerful self-learning and adaptive capabilities to the long-term, unmanned intelligent monitoring of partial discharge of power equipment, reduced human intervention, improved the system's adaptability in complex environments and the accuracy of abnormal event detection, and laid a solid data foundation for subsequent operation and maintenance decisions and fault warnings.

[0030] Furthermore, the synchronous acquisition of multi-channel signals for partial discharges based on edge computing clarifies the functional modules and their collaborative processes within edge nodes. Edge nodes include a local signal acquisition module, an event synchronization module, a feature analysis module, a data hierarchical storage module, multiple physical communication interface modules, and a physically isolated power management module. The signal acquisition module relies on a multi-channel, high-speed analog-to-digital converter, enabling high-precision digitization of the raw analog signals from multiple discharge detection channels. The event synchronization module integrates a high-precision clock source and hardware synchronization signal lines to ensure that all channel acquisition data has a unified and precise timestamp, facilitating the subsequent synchronous fusion and timing consistency correction of multi-channel signals. All acquired data first flows into a high-performance processor or digital signal processing chip, which is responsible for the synchronous fusion of local asynchronously acquired data, parameter extraction based on multi-dimensional features such as time-frequency domain and energy, and intelligent event identification. The feature analysis module, embedded within the processor or present as a co-processing unit, can rapidly analyze the signal's timing characteristics, frequency domain characteristics, energy distribution, and inter-channel correlations. After event identification, the data is routed to the data hierarchical storage module, which categorizes and manages abnormal and normal events. Abnormal data is processed by the encryption and compression module before being encrypted and compressed for storage, ensuring data security and saving storage space. For normal data, the circular cache module stores only the necessary feature parameters and statistical information, and a ring storage structure automatically overwrites expired data. All modules are interconnected by a high-speed bus, ensuring real-time and efficient data exchange with large data volumes. For remote data transmission, edge nodes integrate multiple wired and wireless communication interfaces, enabling two-way data exchange with remote monitoring platforms or the cloud. Synchronous signal lines and high-speed buses can be used between multiple nodes for time synchronization and data sharing, facilitating networked collaborative applications. In addition, a physically isolated power management module is configured within the node to independently power each functional unit, preventing power interference and fault propagation, and improving the electrical safety and reliability of the overall system.

[0031] Through the modular design and process implementation of the aforementioned edge nodes, high-speed, synchronous, and hierarchical processing of multi-channel partial discharge signals can be achieved locally, greatly improving signal synchronization accuracy and feature extraction efficiency. Local intelligent identification and hierarchical storage also reduce the redundant transmission and remote storage pressure of large-scale data. Multiple communication interfaces and physical synchronization mechanisms between nodes ensure multi-node collaborative networking capabilities, enabling the system to flexibly expand and adapt to complex, distributed monitoring scenarios. Physically isolated power management further ensures the long-term stable operation of each module, effectively preventing system-level risks caused by electrical faults, thereby achieving a highly reliable, scalable, and highly adaptable method for intelligent acquisition and processing of partial discharge signals, providing a solid technical foundation for status monitoring and fault warning of power equipment.

[0032] This embodiment uses a signal acquisition process combining asynchronous sampling and synchronous fusion to asynchronously acquire multi-channel partial discharge signals from the electrical equipment under test. Each signal is digitized using independent high-sampling-rate analog-to-digital conversion. At the edge node, each channel's acquired signals are individually timestamped using a high-precision clock and digital processing unit, enabling time reconstruction and synchronous fusion of the asynchronous data to compensate for inter-channel delay and jitter. After synchronization, an adaptive synchronization algorithm is used to locally align and fuse the channel signals based on the calibrated time information, constructing a complete multi-channel synchronous signal group. Subsequently, multi-parameter feature analysis is performed on the synchronous signal group, including extraction of time, frequency, energy, and spectral features. Correlations between the multi-channel signals are also analyzed to further identify the type and time of partial discharge events. An event-driven processing and storage process is employed for the collected data. Specifically, through real-time feature analysis, abnormal partial discharge signals are classified, processed, compressed, and uploaded according to event type. Normal signals, however, only key information is stored in a local buffer, achieving hierarchical compression and data security management. Finally, historical detection data is reviewed regularly, and multi-channel synchronization parameters and feature discrimination thresholds are automatically optimized based on statistical analysis to dynamically adapt to changes in the on-site operating environment.

[0033] By integrating the above steps, this embodiment enables accurate, synchronous acquisition and efficient local processing of multi-channel partial discharge signals, significantly improving the accuracy of signal synchronization and the reliability of event identification in asynchronous acquisition scenarios. The event-driven, hierarchical data processing approach effectively reduces redundant storage and transmission of normal data, improving the efficiency of overall data management and remote interaction. Relying on adaptive optimization of parameters and thresholds, it offers excellent field adaptability and continuously improved recognition accuracy, providing advanced technical means and methodologies for real-time monitoring and intelligent early warning of partial discharge in electrical equipment.

[0034] Example 2 This embodiment is applied to 110kV substation GIS equipment. During operation, insulation aging, defects, and other problems may cause partial discharge.

[0035] In this embodiment, each signal from the synchronized fused multi-channel partial discharge signal undergoes a separate digital filtering process. Taking FIR (finite impulse response) and IIR (infinite impulse response) filters as examples, FIR filters are typically used to suppress fixed-frequency power-frequency interference and wide-bandwidth high-frequency noise. Their filter coefficients can be pre-calculated and fixed in the digital processing module using window function design or least-squares methods based on the actual sampling rate and interference frequency band. IIR filters are more suitable for high-order, narrowband filtering requirements, such as suppressing specific harmonic components. Before entering feature extraction, each signal passes through the aforementioned filter chain. Filter parameters (such as cutoff frequency and stopband attenuation) can be flexibly configured based on ambient noise test results to minimize effective signal attenuation while maximizing background noise suppression. After filtering, to prevent baseline drift from affecting accurate pulse amplitude identification, the system uses a low-pass filter or moving average algorithm to extract the signal baseline. This baseline is then subtracted from the original signal to achieve automatic baseline correction, ensuring more reliable subsequent feature extraction. Signal segmentation is implemented using a sliding window algorithm. The window length is generally set to be slightly larger than the duration of a single partial discharge pulse, and the step length can range from 1 / 2 to 1 / 10 of the sampling point, depending on the actual pulse density and computing resources. During the window sliding process, each extracted segment is treated as an independent analysis unit for subsequent processing. Regarding time domain characteristics, the zero-crossing detection algorithm determines the positive and negative changes in the signal point by point, quickly locking the start and end points of the pulse. It then applies an extreme point search within each pulse interval to accurately calculate the maximum pulse amplitude and duration. Combined with a pulse counting algorithm, the system can accumulate the number of discharge pulses within each window, providing basic data for equipment status assessment. During frequency domain analysis, the method uses the FFT algorithm to transform the signal segments within each sliding window from the time domain to the frequency domain, obtaining the amplitude spectrum and energy distribution of each frequency component. By analyzing the main frequency component, the primary frequency characteristics of the partial discharge signal can be identified, while detecting changes in high-frequency or harmonic components helps identify abnormal discharge types. In terms of energy feature extraction, the integration algorithm will perform numerical integration operations on discrete points of the signal amplitude within the time period when each pulse occurs to obtain the pulse energy intensity. It will also count the mean, variance and other data of all pulse energies in the entire cycle or window to reflect the overall activity and changing trend of the discharge event. In the spectrum feature analysis stage, harmonic component separation relies on amplitude extraction of the high-order frequency components of the FFT results to achieve the identification of high-order harmonic characteristics and nonlinear interference.The Hilbert transform is used to obtain the instantaneous amplitude and frequency of the signal, which can capture short-term mutations and complex waveform characteristics, making it particularly suitable for identifying non-periodic abnormal pulses. Multi-channel correlation analysis uses a unified timestamp as a benchmark, calculating the correlation coefficient for each channel's signal segments within the same sliding window point by point. The output measures the synchronization and similarity of the signals between channels. A high correlation indicates that the discharge event occurred simultaneously at multiple monitoring points and has a spatial distribution. The correlation analysis results can be used for spatial positioning and source identification, providing data support for subsequent equipment failure analysis.

[0036] The multi-parameter feature extraction and correlation discrimination method described above enables in-depth mining of the multi-channel, full-dimensional features of partial discharge signals in complex noise backgrounds and multi-source signal environments. This method not only improves the ability to extract and identify partial discharge events under a single sampling channel, but also incorporates spatial distribution characteristics and signal correlation into the overall discrimination system through multi-channel collaborative analysis, significantly enhancing the ability to distinguish partial discharge types, occurrence locations, and event synchronization. At the same time, the use of automatic parameter adjustment and multi-step verification mechanisms significantly reduces the errors and lack of adaptability caused by manual threshold setting. Ultimately, this method provides high-precision, multi-level raw data and judgment support for the subsequent fine classification of partial discharge events, abnormal trend warnings, and electrical equipment fault location, significantly improving the application value and intelligence level of partial discharge monitoring methods in actual engineering.

[0037] Example 3 This embodiment is applied to 110kV substation GIS equipment. During operation, insulation aging, defects, and other problems may cause partial discharge.

[0038] In this embodiment, the various functional modules of the edge nodes collaborate to collect, process, and transmit multi-channel partial discharge signals. The signal acquisition module utilizes a multi-channel ADC to perform parallel sampling of various discharge sensor signals. The collected raw data is first time-stamped by a hardware synchronization module driven by a high-precision clock, ensuring that the acquisition moment of each signal can be accurately determined during subsequent data processing. Time synchronization relies on the local clock. Based on the needs of multi-node collaboration, the local time reference can be dynamically corrected using external timing signals to improve synchronization accuracy during cross-node networking. The collected multi-channel data is then synchronized and fused on a local high-performance processor / DSP. Specifically, algorithms such as delay compensation, interpolation, and resampling are used to map each channel's signals to a unified reference time axis, achieving synchronization at the data structure level. During the feature analysis phase, edge nodes utilize a modular algorithm library to automatically extract time-domain, frequency-domain, energy, and correlation features tailored to the characteristics of partial discharge signals. Hardware co-processing units accelerate the implementation of some features, such as main frequency, harmonics, and pulse count, to avoid processing bottlenecks caused by large data volumes. The event identification module, based on multi-parameter thresholds and an adaptive decision algorithm, analyzes the feature results in real time and automatically determines whether an event represents an abnormal partial discharge event. For data identified as abnormal, the system invokes locally integrated encryption and compression algorithms to process the raw waveform data block by block, prioritizing the storage and upload of anomalous segments that significantly deviate from the typical template, saving storage space and improving data security. For normal data, only statistical features are stored in a circular buffer, where space is dynamically managed using an automatic overwrite strategy. All data flows utilize a high-speed bus for low-latency interconnection, enhancing the node's internal real-time processing capabilities. The node also features multiple communication interfaces, supporting automatic switching between wired, wireless, and networking modes, enabling flexible data sharing and remote interaction. For power management, physical isolation, independent distribution, and failover mechanisms are employed to ensure stable power supply to each functional module and prevent the spread of single-point failures.

[0039] Through the integration and implementation of the above elements, edge nodes can achieve the synchronous acquisition of multi-channel partial discharge signals, feature parameter extraction, intelligent event identification, and hierarchical and decentralized data management in a highly automated manner. This method improves the time synchronization and data fusion accuracy of multi-channel signals. It also significantly reduces data redundancy and remote transmission pressure through localized intelligent analysis and hierarchical storage, optimizing the monitoring system's resource utilization and response speed. Multiple communication interfaces work together with physical synchronization mechanisms, allowing flexible networking between multiple nodes to meet the actual needs of distributed, multi-source monitoring. Physically isolated power management ensures long-term stable operation of nodes and reduces external electrical interference and the risk of cascading faults. Overall, this method lays a solid technical foundation for the intelligent online acquisition and status diagnosis of partial discharge signals from large-scale, distributed power equipment, improving the system's reliability, scalability, and engineering adaptability, and providing a new approach for intelligent operation and maintenance and fault warning of power systems.

[0040] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for real-time synchronous acquisition of multi-channel partial discharge signals based on edge computing, characterized in that: include: Adopting an asynchronous sampling-synchronous fusion signal acquisition architecture, each channel asynchronously acquires the partial discharge signal of the electrical equipment under test through an independent high-sampling-rate analog-to-digital converter. The edge node uses a clock and digital processing chip to add a time stamp to the asynchronous data, reconstruct the timing, synchronize the partial discharge signals of each channel, compensate for inter-channel delay and jitter, and output the partial discharge signal data. Based on the adaptive synchronization algorithm, the multi-channel partial discharge signals are locally time-aligned and time-series-fused to output a multi-channel signal group; Perform multi-parameter feature analysis on multi-channel signal groups, generate multi-dimensional feature sequences through digital filtering and sliding window segmentation, and identify the type and time of partial discharge events through correlation analysis of multi-channel signal groups; Adopting an event-driven data processing and storage mechanism, based on real-time feature identification, abnormal partial discharge signals are classified, stored, and uploaded by event type. Only key information of normal signals is recorded locally, and signal data is compressed and securely managed in a hierarchical manner. Edge nodes regularly optimize synchronization parameters and discrimination thresholds based on historical detection and operation data.

2. The method for real-time synchronous acquisition of multi-channel partial discharge signals based on edge computing according to claim 1 is characterized in that: The asynchronous sampling-synchronous fusion signal acquisition architecture includes: Each signal acquisition channel is connected to an independent high-sampling rate analog-to-digital converter to collect partial discharge signals from the electrical equipment under test; the edge node is equipped with a field-programmable gate array as a digital processing chip, and a high-precision clock is used to add a timestamp to the partial discharge signal data collected from each channel; during the synchronous timing reconstruction process, the data of each channel is compared with the reference trigger signal, and the partial discharge signal data stream is adjusted based on the digital delay line to compensate for the acquisition start time difference and synchronization error between channels; during the acquisition process, the acquisition system detects short-term clock drift and uses a phase-locked loop circuit or digital filtering algorithm to stabilize the clock; after the multi-channel partial discharge signal data is synchronously reconstructed, the cyclic redundancy check method is used to verify the time stamp integrity, and the partial discharge test data with a unified time stamp is output.

3. The method for real-time synchronous acquisition of multi-channel partial discharge signals based on edge computing according to claim 1 is characterized in that: The multi-channel signal group includes: A high-precision timestamp is added to the acquisition moment of the partial discharge signal of each channel. According to the sampling delay and clock deviation parameters between the channels, a dynamic adjustment strategy is adopted to locally time-align the acquired signals. Through the timing fusion processing method of interpolation and resampling, the partial discharge signals of different channels are synchronized to generate a multi-channel signal group with a unified time base and timing consistency.

4. The method for real-time synchronous acquisition of multi-channel partial discharge signals based on edge computing according to claim 1, characterized in that: The multi-parameter feature analysis includes: For the synchronized multi-channel signal, a digital filter is used to denoise the original signal and correct the baseline drift, and then the signal is segmented through a sliding window algorithm; in the time domain feature extraction process, a zero-crossing detection circuit is used to identify the extreme points and amplitude changes of the pulse signal, and pulse counting is completed through a gated shaping circuit; in the frequency domain feature analysis, a fast Fourier transform module is used to decompose the signal spectrum to obtain the energy distribution and main frequency information of each frequency band; in the energy feature calculation part, an integration circuit is used to integrate the signal amplitude within a single pulse width to obtain the single pulse energy and cycle average energy; in the spectral feature extraction, the harmonic analysis circuit is used to extract the harmonic components, and the Hilbert transform module is used to obtain the instantaneous frequency; the correlation analysis of the multi-channel signal uses a correlation coefficient calculation module to perform point-by-point correlation comparison on the signal waveform of each channel, and combined with the timestamp information, output the judgment result and corresponding time of the local discharge event.

5. The method for real-time synchronous acquisition of multi-channel partial discharge signals based on edge computing according to claim 1, characterized in that: The event-driven data processing and storage mechanism includes: For partial discharge signal data judged as abnormal, the acquisition system integrates an encryption circuit module inside the local processing unit to perform block encryption on the original signal data; a differential compression algorithm module is used to compare the abnormal event data with the set reference signal template, and only the signal segments with obvious differences from the template are extracted and compressed. The compressed partial discharge signal data and the encryption key are stored together in the local non-volatile storage chip, and uploaded to the remote monitoring platform through a wired or wireless communication interface according to the event type group; for partial discharge signal data that is not judged as abnormal, the system adopts a circular cache storage module, using a ring storage structure to only save the key information of the signal. Expired data is automatically overwritten when the storage space reaches the preset threshold.

6. The method for real-time synchronous acquisition of multi-channel partial discharge signals based on edge computing according to claim 1, characterized in that: The optimization synchronization parameters and discrimination thresholds include: The built-in statistical analysis module periodically calls the original records of partial discharge signals and abnormal event records in the historical acquisition data storage unit, and uses hardware counters to count the frequency of various abnormal events and corresponding detection results. The synchronization parameters are corrected by weighted average of the sampling delay and clock phase offset parameters of multi-channel signals through digital circuits. The feature discrimination threshold is based on historical false alarm and missed alarm data, and the signal amplitude, pulse width and frequency threshold are dynamically adjusted using a lookup table.

7. The method for real-time synchronous acquisition of multi-channel partial discharge signals based on edge computing according to claim 1, characterized in that: The edge node includes: Local signal acquisition module, time synchronization module, feature analysis module, data hierarchical storage module and various physical communication interface modules; processor and digital signal processing chip, used for local asynchronous acquisition, time stamping, synchronous fusion, feature parameter extraction and event discrimination processing of multi-channel partial discharge signals; data is stored hierarchically in local storage units after feature analysis and event discrimination; edge nodes are used to perform online monitoring of the insulation status of the electrical equipment under test, conduct real-time analysis and fault trend prediction of partial discharge signals, automatically identify equipment anomalies and insulation degradation, and generate diagnostic reports and alarm information based on the analysis results, which are remotely transmitted to the monitoring center via the communication interface to support intelligent detection and remote management of the operating status of electrical equipment; edge nodes perform time synchronization and data exchange with other edge nodes through hardware synchronization signal lines and high-speed communication buses, and support multi-node collaborative networking; edge nodes are equipped with physically isolated power management modules to allocate independent power to each functional module.

Images