A cloud computing-based power consumption equipment monitoring data processing method and system

Abstract

The embodiment of the application provides a kind of based on cloud computing's electrical equipment monitoring data processing method and system, it is related to electrical equipment monitoring data processing technical field, method includes receiving data packet from edge computing node;According to the historical clock data of the edge computing node of data packet, calibration time and network transmission characteristic parameter, the time stamp confidence interval of data packet is determined to dynamically adjust data collection window, and event snapshot is obtained;The event correlation strength between the data packet collected in event snapshot is calculated, and according to the overlap degree of event correlation strength and time stamp confidence interval, the data packet with high correlation strength is obtained and the data packet with high correlation strength is preferentially aligned;Based on the data packet with high correlation strength after alignment, reconstructs equipment working condition snapshot, and equipment working condition snapshot is output to equipment health assessment system.The application can improve the accuracy of equipment health assessment and the timeliness of early warning.

Description

Technical Field

[0001] This application relates to the field of electrical equipment monitoring data processing technology, and in particular to a cloud computing-based method and system for processing electrical equipment monitoring data. Background Technology

[0002] In modern industrial environments, particularly large industrial parks, the demand for real-time monitoring and data processing of various electrical equipment is increasing. Traditional monitoring systems often face challenges such as high data acquisition latency, difficulties in integrating heterogeneous data, and low efficiency in processing massive amounts of data. These problems can lead to delayed equipment status prediction and untimely fault warnings. Typically, we deploy an IoT sensing layer to collect data in real time, perform preliminary processing of various parameters such as voltage, current, and temperature through edge computing nodes, and then rely on a cloud computing platform to aggregate data from across regions and extract useful information, ultimately establishing a system for assessing equipment health. However, achieving high-precision time synchronization between distributed edge nodes, especially for equipment that requires close coordination and high-frequency operation, has always been a technical challenge. The inherent physical characteristics of local clocks and the influence of environmental factors can introduce small and irregular time deviations, making it difficult for cloud analytics to align data and identify key, subtle timing anomalies. Summary of the Invention

[0003] This application aims to address at least one of the technical problems existing in the prior art. To this end, this application proposes a cloud computing-based method and system for processing monitoring data of electrical equipment, aiming to improve the accuracy of equipment health assessment and the timeliness of early warning.

[0004] In a first aspect, embodiments of this application provide a cloud computing-based method for processing monitoring data of electrical equipment, comprising: Receive data packets from edge computing nodes; The timestamp confidence interval of the data packet is determined based on the historical clock data of the edge computing node, the calibration time, and the network transmission characteristic parameters. Based on the timestamp confidence interval, the data collection window is dynamically adjusted, and the data packets with overlapping confidence intervals in the data collection window are collected into an event snapshot. Calculate the event correlation strength between the data packets collected in the event snapshot, and based on the event correlation strength and the degree of overlap of the timestamp confidence intervals, obtain the data packets with high correlation strength and prioritize aligning the data packets with high correlation strength; The device condition snapshot is reconstructed based on the aligned data packets with high correlation strength, and the device condition snapshot is output to the device health assessment system.

[0005] According to some embodiments of this application, the step of calculating the event correlation strength between the data packets collected in the event snapshot, and obtaining data packets with high correlation strength and prioritizing the alignment of the data packets with high correlation strength based on the event correlation strength and the overlap of the timestamp confidence intervals, includes: Based on the edge computing nodes monitoring the internal sensor data of the precision machining equipment, the fine-tuning compensation behavior of the local controller is inferred, the inferred compensation amount is obtained, and the expected timing delay between the precision machining equipment is corrected based on the inferred compensation amount, wherein the edge computing nodes correspond one-to-one with the precision machining equipment; The actual relative timing delay between the precision machining equipment is analyzed, the persistent deviation from the expected timing delay is identified, and the confidence level of the physical causal relationship between the precision machining equipment is updated based on the statistical characteristics of the persistent deviation, thus obtaining the updated confidence level of the physical causal relationship. The overlap of the timestamp confidence intervals is dynamically adjusted based on the updated physical causality confidence level and the inference compensation amount. Based on the degree of overlap of the dynamically adjusted timestamp confidence intervals, the event association strength between the data packets collected in the event snapshot is calculated. Based on the event association strength and the degree of overlap of the timestamp confidence intervals, data packets with high association strength are obtained and prioritized for alignment.

[0006] According to some embodiments of this application, the step of analyzing the actual relative timing delay between the precision machining equipment, identifying persistent deviations from the expected timing delay, and updating the confidence level of the physical causality relationship between the precision machining equipment based on the statistical characteristics of the persistent deviation, to obtain the updated confidence level of the physical causality relationship, includes: Acquire real-time sensor data and perform high-frequency sampling on the real-time sensor data to obtain high-frequency sampled data; transmit the high-frequency sampled data to the edge computing node to calculate the instantaneous relative timing delay between the precision machining equipment in real time; Statistical analysis is performed on the instantaneous relative timing delay within each preset time window to obtain the deviation characteristics at different time granularities; By comparing the deviation characteristics at different time granularities, deviation patterns with high transientity, sporadic occurrence and extremely short duration are identified, and the deviation patterns are compared with the expected time delay. Based on the frequency, amplitude, and degree of deviation of the deviation pattern from the preset process flow rules, the confidence level of the physical causal relationship between the precision machining equipment is updated to obtain the updated confidence level of the physical causal relationship.

[0007] According to some embodiments of this application, the step of monitoring internal sensor data of precision machining equipment, inferring fine-tuning compensation behavior of the local controller, obtaining an inferred compensation amount, and correcting the expected timing delay between the precision machining equipment based on the inferred compensation amount includes: The system continuously monitors the internal control signals emitted by the local controller and the internal sensor data of the precision machining equipment, and synchronizes them in real time through the edge computing node. Detect whether there is a continuous, non-preset sub-millisecond timing deviation between the internal control signal and the actual execution result reflected by the internal sensor data; When there is a continuous, non-preset sub-millisecond timing deviation between the actual execution results, the dynamic fine-tuning compensation amount being executed by the local controller is obtained; Based on the persistence, magnitude, and correlation with the operating parameters of the precision machining equipment of the sub-millisecond timing deviation, the fine-tuning compensation logic currently used by the local controller is inferred and the dynamic fine-tuning compensation amount is quantified to obtain the inferred compensation amount. Based on the fine-tuning compensation logic and the inferred compensation amount, the expected timing delay between the precision machining equipment is corrected.

[0008] According to some embodiments of this application, the continuous monitoring of internal control signals emitted by the local controller and internal sensor data of the precision machining equipment further includes: The hardware clock source of the hardware time synchronization module timestamps the captured internal control signals and sensor data, and synchronizes the time with the standard time server through a preset hardware time synchronization protocol.

[0009] According to some embodiments of this application, when there is a continuous, non-preset sub-millisecond timing deviation between the actual execution results, obtaining the dynamic fine-tuning compensation amount being executed by the local controller includes: When there is a continuous, non-preset sub-millisecond timing deviation between the actual execution results, analyze the dynamic change pattern of the sub-millisecond timing deviation; Based on the dynamic change pattern of the sub-millisecond timing deviation, and combined with the preset compensation behavior feature library, the feature fingerprint of the fuzzy control rules or neural network compensation behavior that the local controller may have is identified. By performing real-time clustering and pattern matching on the feature fingerprint, the type of nonlinear fine-tuning compensation logic currently used by the local controller can be inferred. Based on the inferred nonlinear fine-tuning compensation logic type, the corresponding inverse compensation algorithm is dynamically selected and the compensation amount is calculated to obtain the dynamic fine-tuning compensation amount being executed by the local controller.

[0010] According to some embodiments of this application, the analysis of the dynamic change pattern of the sub-millisecond timing deviation includes: The operating parameters of the precision machining equipment were monitored to show a step change; When the operating parameters of the precision machining equipment undergo a step change, the parameter configuration of the preset time-series deviation analysis model is dynamically adjusted. Under the adjusted preset timing deviation analysis model, the timing deviation analysis process is re-initialized and the sub-millisecond timing deviation is analyzed to obtain the dynamic change pattern of the sub-millisecond timing deviation.

[0011] According to some embodiments of this application, the step of inferring the fine-tuning compensation logic currently used by the local controller based on the persistence, amplitude, and correlation with the operating parameters of the precision machining equipment of the sub-millisecond timing deviation includes: Continuously monitor the nonlinear coupling relationship between the operating parameters of the precision machining equipment and identify the changing patterns of the nonlinear coupling relationship; Based on the identified change pattern of the nonlinear coupling relationship, the sub-millisecond timing deviation is dynamically adjusted; The influence of the sub-millisecond time series deviation is decoupled by multidimensional eigenvalue decomposition to obtain the decoupled sub-millisecond time series deviation. Based on the correlation between the decoupled sub-millisecond timing deviation and the operating parameters of the precision machining equipment, the fine-tuning compensation logic currently used by the local controller is inferred.

[0012] According to some embodiments of this application, determining the timestamp confidence interval of the data packet based on the historical clock data of the edge computing node, calibration time, and network transmission characteristic parameters of the data packet includes: Based on the historical clock data, a clock drift prediction model for the edge computing node is constructed. The estimated drift amount corresponding to the locally generated timestamp is calculated through the clock drift prediction model to obtain a first time correction amount and a first uncertainty range. The historical clock data includes the deviation value between the local clock and the standard time of the NTP server after each NTP calibration of the edge computing node and the time point of each calibration. Query the time point of the most recent calibration of the edge computing node, calculate the time interval between the locally generated timestamp and the time point of the most recent calibration, and adjust the first uncertainty range according to the time interval to obtain the second uncertainty range; The network transmission characteristic parameters of the data packets are collected, including the local sending timestamp recorded by the edge computing node and the data packet arrival timestamp recorded by the cloud platform; the one-way network transmission delay and delay jitter range are estimated based on the network transmission characteristic parameters, and the delay jitter range is used as the third uncertainty range; The locally generated timestamp is corrected by combining the first time correction amount to obtain the target timestamp; the second uncertainty range and the third uncertainty range are fused to determine the timestamp confidence interval of the data packet.

[0013] Secondly, embodiments of this application provide a cloud computing-based electrical equipment monitoring data processing system, comprising: The receiving module is used to receive data packets from edge computing nodes; The confidence interval determination module is used to determine the timestamp confidence interval of the data packet based on the historical clock data of the edge computing node, the calibration time, and the network transmission characteristic parameters. The data collection module is used to dynamically adjust the data collection window according to the timestamp confidence interval, and to collect the data packets with overlapping confidence intervals in the data collection window into an event snapshot; The data alignment module is used to calculate the event correlation strength between the data packets collected in the event snapshot, and to obtain data packets with high correlation strength based on the event correlation strength and the degree of overlap of the timestamp confidence intervals, and to prioritize align the data packets with high correlation strength. The snapshot reconstruction and output module is used to reconstruct the device condition snapshot based on the aligned data packets with high correlation strength, and output the device condition snapshot to the device health assessment system.

[0014] The technical solution according to the embodiments of this application has at least the following beneficial effects: This application discloses a cloud computing-based method for processing monitoring data of electrical equipment. By receiving data packets from edge computing nodes and determining the timestamp confidence interval of the data packets based on the historical clock data, calibration time, and network transmission characteristic parameters of the edge computing nodes, it effectively solves the problem of difficulty in achieving high-precision time synchronization between distributed edge nodes in traditional monitoring systems. Furthermore, this application dynamically adjusts the data collection window based on the timestamp confidence interval and collects data packets with overlapping confidence intervals into event snapshots, overcoming the challenges of heterogeneous data fusion difficulties and low efficiency in processing massive amounts of data. Further, by calculating the event correlation strength between the data packets collected in the event snapshots and prioritizing the alignment of data packets with high correlation strength based on the event correlation strength and the degree of overlap of the timestamp confidence intervals, it significantly improves the accuracy of data alignment and solves the problems of data alignment difficulties and misjudgments caused by timestamp ambiguity in the prior art. Ultimately, by reconstructing the equipment operating condition snapshot based on the aligned data packets with high correlation strength and outputting it to the equipment health assessment system, more accurate and timely equipment status information can be provided. This effectively avoids processing defects and economic losses caused by timing deviations, and achieves sub-millisecond timing alignment for high-precision, high-frequency cooperating equipment, thereby improving the accuracy of equipment health assessment and the timeliness of early warning.

[0015] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0016] The accompanying drawings are used to provide a further understanding of the technical solutions of this application and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of this application and do not constitute a limitation on the technical solutions of this application.

[0017] Figure 1 A flowchart illustrating a cloud-based method for processing electrical equipment monitoring data, provided in one embodiment of this application; Figure 2 This is a schematic diagram of a cloud-based electrical equipment monitoring data processing system provided in one embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical methods, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0019] It should be noted that the meaning of "multiple" (or "more than") in the description of the embodiments of this application refers to two or more, and "greater than," "less than," "exceeding," etc. are understood to exclude the number itself, while "above," "below," "within," etc. are understood to include the number itself. If "first," "second," etc. are used in the description, they are only for the purpose of distinguishing technical features and should not be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated or the order of the technical features indicated.

[0020] In this application embodiment, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent the existence of A alone, the simultaneous existence of A and B, or the existence of B alone. A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of singular or plural items. For example, at least one of a, b, and c can represent: the existence of a alone, the existence of b alone, the existence of c alone, the simultaneous existence of a and b, the simultaneous existence of a and c, the simultaneous existence of b and c, or the simultaneous existence of a, b, and c, where a, b, and c can be single or multiple.

[0021] In the description of this application, unless otherwise expressly defined, terms such as "setup," "installation," and "connection" should be interpreted broadly, and those skilled in the art can reasonably determine the specific meaning of the above terms in this application in conjunction with the specific content of the technical solution.

[0022] Based on the above, this application proposes a cloud computing-based method and system for processing monitoring data of electrical equipment, aiming to improve the accuracy of equipment health assessment and the timeliness of early warning.

[0023] The cloud computing-based equipment monitoring data processing method provided in this application can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, etc.; the server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms; the software can be an application that implements the cloud computing-based equipment monitoring data processing method, but is not limited to the above forms.

[0024] This application can be applied to numerous general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics devices, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via communication networks. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices. It should be noted that in various specific embodiments of this invention, when processing is required based on data related to the characteristics of an object (e.g., user attributes or sets of attribute information), permission or consent from the corresponding object is obtained first, and the collection, use, and processing of this data comply with relevant laws and standards. Furthermore, when the embodiments of the present invention need to obtain the attribute information of an object, they will obtain the separate permission or separate consent of the corresponding object through pop-up windows or redirection to a confirmation page. After obtaining the separate permission or separate consent of the corresponding object, they will then obtain the relevant data of the object necessary for the embodiments of the present invention to operate normally.

[0025] See Figure 1 , Figure 1 This is a flowchart illustrating a cloud-based method for processing monitoring data of electrical equipment according to an embodiment of this application. The cloud-based method for processing monitoring data of electrical equipment provided in this embodiment includes, but is not limited to, steps S110 to S150, which will be described in detail below.

[0026] Step S110: Receive data packets from the edge computing node; Step S120: Determine the timestamp confidence interval of the data packet based on the historical clock data of the edge computing node, calibration time, and network transmission characteristic parameters; Step S130: Dynamically adjust the data collection window based on the timestamp confidence interval, and collect data packets with overlapping confidence intervals in the data collection window into event snapshots; Step S140: Calculate the event association strength between the data packets collected in the event snapshot, and obtain the data packets with high association strength according to the event association strength and the degree of overlap of the timestamp confidence intervals, and prioritize aligning the data packets with high association strength. Step S150: Reconstruct the device condition snapshot based on the aligned data packets with high correlation strength, and output the device condition snapshot to the device health assessment system.

[0027] It should be noted that a "data packet" refers to a data unit containing monitoring data of electrical equipment and related metadata, collected and encapsulated by edge computing nodes. These data packets carry real-time information on the equipment's operating status and form the basis for subsequent processing. A "timestamp confidence interval" refers to a time range representing the interval within which the actual generation time of a data packet might fall. Due to factors such as clock drift and network latency, a single timestamp often has uncertainty; therefore, introducing a confidence interval can more accurately describe the temporal attributes of the data packet. A "data aggregation window" refers to a dynamically adjustable interval used for collecting and processing data packets in the time dimension. By adjusting this window, it can flexibly adapt to the data processing needs of different scenarios. An "event snapshot" refers to a collection of data packets with overlapping confidence intervals within a specific data aggregation window. These data packets are considered related to the same or related events and form the basis for event correlation analysis. "Event correlation strength" refers to the degree of temporal, spatial, or logical correlation between the data packets aggregated in the event snapshot. A high correlation strength indicates that these data packets may collectively reflect an important change in equipment operating conditions or an event. A "device status snapshot" refers to a comprehensive description of the equipment's operating status at a specific moment or within a specific time period, reconstructed based on aligned, highly correlated data packets. This snapshot serves as a crucial basis for analysis and decision-making by the equipment health assessment system.

[0028] In one embodiment, data packets are first received from edge computing nodes. These data packets may contain sensor data from various electrical devices, such as voltage, current, temperature, vibration, etc. Data packet reception can be achieved through standard network protocols, such as MQTT, HTTP, or TCP / IP. For example, edge computing nodes can periodically encapsulate the collected sensor data into data packets and send them to the cloud platform via wired or wireless networks. In one implementation, the cloud platform deploys a dedicated data receiving service that continuously monitors data streams from each edge computing node and performs preliminary verification and caching of the received data packets. Next, based on the historical clock data of the edge computing node, calibration time, and network transmission characteristic parameters of the data packets, the timestamp confidence interval of the data packets is determined. This step is the core of this application's solution to the time synchronization problem. Historical clock data may include the deviation between the local clock of the edge computing node and the standard time of the NTP server after each NTP calibration, and the time points of each calibration. The calibration time refers to the time point when the edge computing node most recently synchronized with the standard time server. Network transmission characteristic parameters may include the local transmission timestamp recorded by the edge computing node and the data packet arrival timestamp recorded by the cloud platform. Using this information, a clock drift prediction model can be constructed to estimate the predicted drift of locally generated timestamps. Combined with network transmission latency and jitter, a more accurate timestamp confidence interval can be determined. For example, NTP calibration records from edge computing nodes over the past few weeks or months can be used to analyze clock drift trends and patterns, establishing a linear prediction model. When a data packet is received, this model is used to predict the potential drift range of its locally generated timestamp. Simultaneously, by measuring the round-trip time or one-way delay of the data packet from the edge computing node to the cloud platform, the time uncertainty introduced by network transmission is estimated. Based on this uncertainty, the final timestamp confidence interval for the data packet can be obtained. Then, based on the timestamp confidence interval, the data aggregation window is dynamically adjusted, and data packets with overlapping confidence intervals in the data aggregation window are aggregated into event snapshots. Traditional data aggregation windows are usually of a fixed size, which can easily lead to incorrect segmentation of relevant data or incorrect aggregation of irrelevant data when timestamp uncertainty is high. This application dynamically adjusts the aggregation window to adapt to changes in the timestamp confidence interval.

[0029] In one embodiment, when edge node 1 and edge node 2 send their collected data packets to the cloud platform, the cloud platform's data receiving service calculates a "confidence interval" for the timestamp of each data packet. The cloud platform maintains a database recording historical calibration data of edge node 1 and edge node 2 with the NTP server over the past weeks or even months. For example, it finds that the crystal oscillator of edge node 1 typically speeds up gradually at a rate of 0.5 milliseconds per hour after calibration, while the crystal oscillator of edge node 2 slows down at a rate of 0.3 milliseconds per hour, and this drift has a certain regularity that can be predicted using a quadratic polynomial function. Assuming that edge node 1's last NTP calibration was 1 hour ago, its local timestamp T_A is 10:00:00.124. Based on the historical drift model, the cloud platform predicts that the crystal oscillator of edge node 1 may have accumulated a speedup of 0.5 milliseconds in the past hour. Therefore, it corrects T_A to 10:00:00.1235 and calculates a confidence interval [10:00:00.1233, 10:00:00.1237] for T_A based on the prediction model error (e.g., ±0.2 milliseconds). Similarly, for edge node 2, whose local timestamp T_B is 10:00:00.125, the cloud platform predicts that its crystal oscillator may have accumulated a slowdown of 0.3 milliseconds, corrects T_B to 10:00:00.1253, and calculates a confidence interval [10:00:00.1251, 10:00:00.1255]. If an edge node has been away from its last NTP calibration for more than a preset threshold (e.g., 2 hours), the cloud platform assumes that its timestamp drift may be more significant and therefore expands its confidence interval. For example, if edge node 1 has not been calibrated for 3 hours, its confidence interval may expand from ±0.2 milliseconds to ±0.5 milliseconds. The cloud platform estimates network transmission latency by recording the arrival time of data packets and the transmission time carried within the data packets, combined with historical network latency statistics. For example, the average latency is 10 milliseconds, and the jitter range is ±0.1 milliseconds. This jitter will also be included in the confidence interval calculation. Finally, after comprehensive evaluation, the data packet timestamp of robot A may be determined as 10:00:00.1235, with a confidence interval of [10:00:00.1232, 10:00:00.1238]; the data packet timestamp of robot B may be determined as 10:00:00.1239, with a confidence interval of [10:00:00.1236, 10:00:00.1242].

[0030] In one embodiment, when a new data packet is received, the system evaluates the overlap between its timestamp confidence interval and the confidence intervals of existing data packets within the current collection window. If significant overlap exists, it indicates that these data packets may belong to the same event, and the collection window may be expanded or moved accordingly to include these related data packets. For example, a threshold can be set; when the confidence interval of a new data packet overlaps with the confidence interval of at least one data packet within the current window by more than this threshold, the new data packet is included in the current window, and the start and end times of the collection window are dynamically adjusted based on the confidence interval ranges of all included data packets. When processing data from robots A and B, a fixed time window is no longer used. It will find that the confidence intervals [10:00:00.1232, 10:00:00.1238] for robot A and [10:00:00.1236, 10:00:00.1242] for robot B overlap (the overlapping interval is [10:00:00.1236, 10:00:00.1238]). The system will dynamically adjust the aggregation window to aggregate these two data packets into the same logical event snapshot. It will calculate a weighted average timestamp as the representative time of the snapshot, for example, 10:00:00.1237, and determine a smaller window that includes the core possible times of these two data packets. Next, it will calculate the event correlation strength between the data packets aggregated in the event snapshot, and based on the event correlation strength and the degree of overlap of the timestamp confidence intervals, it will obtain data packets with high correlation strength and prioritize aligning them. After the event snapshot is formed, further analysis of the correlation of the data packets is required. Meanwhile, the degree of overlap of timestamp confidence intervals is also an important indicator of correlation. Higher overlap indicates that these data packets are more likely to occur simultaneously in time. Furthermore, the intersection of the timestamp confidence intervals of these data packets is calculated; a larger intersection indicates a higher probability of time synchronization. Finally, data packets with high correlation coefficients and large time intersections are selected for priority alignment. Finally, the equipment condition snapshot is reconstructed based on the aligned data packets with high correlation strength, and the equipment condition snapshot is output to the equipment health assessment system. After priority alignment, data packets considered highly correlated and with high time synchronization accuracy are used to construct the equipment condition snapshot. The reconstructed equipment condition snapshot is then sent to the equipment health assessment system, which can utilize this high-precision, high-reliability condition data for more accurate fault diagnosis, remaining life prediction, and maintenance plan optimization.

[0031] It's important to note that, firstly, by monitoring sensor data collected from the precision machining equipment by edge computing nodes, the fine-tuning compensation behavior performed by the local controller to maintain equipment performance can be inferred, and the inferred compensation amount can be quantified. This inferred compensation amount is used to correct the expected timing delay between devices, making the expected timing delay closer to the actual physical causal relationship, thus providing a more accurate benchmark for subsequent correlation analysis. Secondly, by continuously analyzing the actual relative timing delay between precision machining devices and identifying the persistent deviation between it and the corrected expected timing delay, the confidence level of the physical causal relationship between devices can be dynamically updated based on the statistical characteristics of these persistent deviations. Finally, by combining the updated confidence level of the physical causal relationship and the inferred compensation amount, the system can dynamically adjust the overlap of the data packet timestamp confidence intervals. This dynamic adjustment mechanism allows the calculation of event correlation strength to fully consider the micro-dynamics within the devices and the macro-causal relationships, resulting in a more accurate event correlation strength and prioritizing the alignment of data packets with high correlation strength, ensuring the accuracy and reliability of data alignment.

[0032] In one embodiment, assuming an automated production line consisting of multiple precision machining units, such as a multi-axis CNC machine tool system, each axis (e.g., X-axis, Y-axis, Z-axis) is controlled by an independent local controller and equipped with internal sensors such as high-precision encoders and force sensors. When performing complex machining tasks, these local controllers may perform sub-millisecond-level fine-tuning compensations to eliminate vibration, improve surface finish, or compensate for tool wear. First, the edge computing node continuously monitors the internal control signals emitted by the local controller of each axis, as well as data from internal sensors such as encoders and force sensors. By analyzing this data, for example, when the Z-axis is performing high-speed feed, its local controller may perform periodic small reverse pulse compensation to suppress vibration. The system infers the fine-tuning compensation behavior being performed by the Z-axis local controller based on the timing relationships and patterns of these signals and sensor data, and quantifies the corresponding inferred compensation amount. For example, the inferred compensation amount might be a value representing a small lag or lead of the actual Z-axis position relative to the commanded position. Next, the system analyzes the actual relative timing delays between different axes (e.g., the X-axis and Z-axis). If a persistent, for example, microsecond-level deviation is detected between the actual relative timing delay and the corrected expected timing delay (considering the inferred compensation amount for the Z-axis), the system will identify this deviation pattern. For example, an action on the X-axis may always be a fixed microsecond later than an action on the Z-axis, even after considering Z-axis compensation. Based on the statistical characteristics of this persistent deviation (such as the stability and frequency of the deviation), the system updates the confidence level of the physical causality between the X and Z axes. Then, the system dynamically adjusts the overlap of the timestamp confidence intervals of related data packets on the X and Z axes based on the updated physical causality confidence level and the previously inferred compensation amount. For example, if the confidence level of the physical causality between the X and Z axes is low, or the inferred compensation amount is large, the system may adjust the overlap requirement of their data packet timestamp confidence intervals to more accurately determine whether they belong to the same event. Finally, based on the dynamically adjusted overlap of the timestamp confidence intervals, the system recalculates the event correlation strength between the X and Z axis data packets collected in the event snapshot. In this way, data packets with strong physical causal relationships and high temporal consistency (even with fine-tuning compensation) are identified as having high correlation strength and are prioritized for alignment. It should be noted that by acquiring high-frequency sampled real-time sensor data and calculating instantaneous relative timing delays in real time, the system ensures the ability to capture microscopic dynamics between devices. Secondly, statistical analysis is performed on the instantaneous relative timing delays within each preset time window to obtain deviation characteristics at different time granularities, enabling the system to examine timing deviations from both macroscopic and microscopic perspectives.By comparing deviation characteristics at different time granularities and identifying highly transient, sporadic, and extremely short-duration deviation patterns, the system can effectively distinguish noise from genuine anomalies and focus on subtle, transient temporal anomalies that may foreshadow potential problems. Finally, the confidence level of the physical causal relationship is updated based on the frequency, amplitude, and degree of deviation from preset process flow rules of the deviation pattern. This ensures that the confidence level update process fully considers the severity and persistence of abnormal behavior, thus more accurately reflecting the true state of the physical causal relationship between equipment.

[0033] In one embodiment, real-time sensor data refers to raw data acquired from various sensors (e.g., position sensors, speed sensors, force sensors, temperature sensors, etc.) within the precision machining equipment. This data is sampled at high frequencies, such as microseconds or sub-milliseconds, to capture extremely subtle and rapidly changing dynamic information during equipment operation. The high-frequency sampled data is then transmitted to edge computing nodes corresponding to each precision machining device for real-time processing near the data source, thereby calculating the instantaneous relative time delay between different precision machining devices at a specific moment. The aim is to ensure accurate perception of the real-time nature and precision of interactions between devices. Preset time windows can be set according to specific process requirements and equipment characteristics, and can be in the millisecond, second, or even longer time intervals. Statistical analysis is performed on the instantaneous relative time delay within each time window, such as calculating the mean, variance, standard deviation, maximum, and minimum values, to reveal the distribution and trends of time delay at different time scales. In this way, the deviation characteristics exhibited at different time granularities, such as short-term fluctuations and medium- to long-term drifts, can be obtained. The aim is to comprehensively understand the dynamic behavior of timing delays. By comparing deviation characteristics at different time granularities, transient deviations caused by system noise, occasional interference, or internal equipment fine-tuning can be distinguished from persistent deviations caused by equipment wear, environmental changes, etc. Specifically, it identifies deviation patterns that are highly transient, occasional, and extremely short in duration, for example, by setting thresholds, pattern matching algorithms, or machine learning models to detect abnormal spikes or brief offsets. Subsequently, these identified deviation patterns are compared with the expected timing delays between precision machining equipment to determine whether these deviations exceed the normal range or indicate specific abnormal operating conditions. The goal is to accurately identify subtle timing anomalies that have a potential impact on equipment performance and product quality. After identifying deviation patterns, quantitative analysis is required. The frequency of the deviation pattern refers to the number of times the pattern occurs within a certain period of time; the amplitude refers to the size or intensity of the deviation; and the degree of deviation from the preset process flow rules refers to the difference between the deviation pattern and the allowable timing range or behavior pattern in the normal process flow. Based on these quantitative indicators, the confidence level of the physical causal relationship between precision machining equipment can be dynamically updated.

[0034] In one embodiment, continuously monitoring the internal control signals issued by the local controller and the internal sensor data of the precision machining equipment, and synchronizing them in real time through edge computing nodes, refers to the uninterrupted acquisition and synchronization of control commands generated by the local controller and data collected by various sensors (position sensors, speed sensors, force sensors, etc.) inside the precision machining equipment through edge computing nodes. This process aims to obtain raw, high-precision timing information of the equipment's operating status and control commands, providing basic data for subsequent deviation analysis. Real-time synchronization ensures the consistency of control signals and sensor data in the time dimension, avoiding timing misalignment caused by data acquisition or transmission delays. Furthermore, detecting whether there is a continuous, non-preset sub-millisecond timing deviation between the internal control signals and the actual execution results reflected by the internal sensor data refers to comparing and analyzing the synchronized control signals and sensor data. When a continuous, non-preset sub-millisecond timing deviation exists between the actual execution results, the dynamic fine-tuning compensation amount being executed by the local controller can be obtained. This means that once the above-mentioned timing deviation is detected, the system can identify that this is a real-time, dynamic adjustment made by the local controller to optimize performance or respond to environmental changes. This compensation amount reflects the controller's autonomous correction behavior at the micro level. Furthermore, based on the persistence and magnitude of the sub-millisecond timing deviation and its correlation with the operating parameters of the precision machining equipment, the fine-tuning compensation logic currently employed by the local controller is inferred, and the dynamic fine-tuning compensation amount is quantified to obtain the inferred compensation amount.

[0035] It's important to note that a hardware time synchronization module refers to a dedicated hardware circuit integrated inside or outside an edge computing node, with a hardware clock source at its core. This hardware clock source typically employs a high-precision crystal oscillator or atomic clock to provide a stable and accurate time reference. Timestamping the captured internal control signals and sensor data involves the hardware time synchronization module using its hardware clock source to attach a precise timestamp to the data the instant it is acquired by the edge computing node. This process is performed at the hardware level, minimizing latency and jitter caused by software processing and ensuring timestamp accuracy. A pre-defined hardware time synchronization protocol can be understood as a communication standard implemented at the hardware level for synchronizing with other time sources, such as a hardware-accelerated implementation of the IEEE 1588 Precision Time Protocol (PTP) or Network Time Protocol (NTP). A standard time server refers to an external server that provides high-precision, authoritative time information, such as a time server calibrated via GPS or an atomic clock.

[0036] In one embodiment, assume an edge computing node is connected to a high-speed precision machining equipment. To accurately monitor the equipment's internal control signals and sensor data, the edge computing node integrates a hardware time synchronization module. This module contains a highly stable crystal oscillator as its hardware clock source. When the edge computing node captures servo motor control commands (internal control signals) and encoder feedback data (internal sensor data), the hardware time synchronization module immediately timestamps this data using its hardware clock source. Simultaneously, the hardware time synchronization module periodically synchronizes with a PTP master clock server within the factory via a hardware-implemented PTP protocol. This PTP master clock server may be calibrated using GPS signals to provide a high-precision standard time. In this way, all data captured by the edge computing node carries a sub-microsecond precision timestamp synchronized with the global standard time, providing an extremely reliable time reference for subsequent detection of sub-millisecond timing deviations between control commands and actual execution results.

[0037] It should be noted that analyzing the dynamic change patterns of sub-millisecond time-series deviations refers to performing time series analysis on the detected sub-millisecond time-series deviations. For example, methods such as Fourier transform, wavelet analysis, or autoregressive moving average (ARMA) models can be used to reveal the periodicity, trend, randomness, and transient response characteristics of the deviations. The aim is to gain a deeper understanding of the inherent laws and generation mechanisms of the deviations. Specifically, by combining a pre-set compensation behavior feature library, the characteristic fingerprints of possible fuzzy control rules or neural network compensation behaviors of the local controller are identified. This pre-set compensation behavior feature library stores typical characteristics of time-series deviation patterns generated by various known nonlinear control logics (such as different types of fuzzy controllers and neural network controllers with different structures) under specific operating conditions. By comparing the dynamic change patterns of the sub-millisecond time-series deviations obtained from the current analysis with the patterns in the feature library, "characteristic fingerprints" that highly match a certain nonlinear control logic are identified. The purpose is to preliminarily determine the type of nonlinear compensation strategy that the local controller may employ. In practical applications, the type of nonlinear fine-tuning compensation logic currently used by the local controller is inferred through real-time clustering and pattern matching of feature fingerprints. Specifically, after identifying the initial feature fingerprints, clustering algorithms (such as K-means and DBSCAN) or pattern recognition techniques from machine learning are used to analyze the real-time collected deviation data and classify it into a preset nonlinear control logic model. The purpose is to accurately determine the specific type and parameters of the nonlinear compensation logic used by the local controller. Further, based on the inferred nonlinear fine-tuning compensation logic type, the corresponding inverse compensation algorithm is dynamically selected and the compensation amount is calculated. This means that once the nonlinear fine-tuning compensation logic type is determined, the system selects an algorithm matching that logic type from a preset inverse compensation algorithm library. If fuzzy control is inferred, the corresponding fuzzy inverse model algorithm is selected; if neural network compensation is used, the corresponding backpropagation or model inversion algorithm is selected. By inputting real-time deviation data into the selected inverse compensation algorithm, the actual dynamic fine-tuning compensation amount applied by the local controller is calculated.

[0038] In one embodiment, it is assumed that the local controller of a precision machining equipment employs a neural network-based adaptive compensation algorithm to cope with minute fluctuations in material properties and ambient temperature during machining. When the system detects a continuous, non-preset sub-millisecond timing deviation between the internal control signal and internal sensor data, it first performs time-frequency analysis on these deviation data to obtain the dynamic change pattern of the deviation. Then, it compares this dynamic change pattern with a preset compensation behavior feature library. This feature library stores the output response feature fingerprints of various neural network models under different input conditions. Through pattern matching, the system identifies that the current deviation pattern highly matches the feature fingerprint of a specific neural network compensation behavior. Next, the system performs real-time clustering and pattern matching on the identified feature fingerprints to further infer that the local controller is currently using a convolutional neural network (CNN) compensation logic with three hidden layers and using the ReLU activation function. Finally, based on the inferred CNN compensation logic type, the system dynamically selects a preset inverse compensation algorithm based on backpropagation. The real-time monitored sub-millisecond timing deviation data is input into this inverse compensation algorithm to accurately calculate the dynamic fine-tuning compensation amount currently being executed by the local controller.

[0039] In one embodiment, the operating parameters of the precision machining equipment can be understood as key indicators reflecting the current working state of the equipment, such as spindle speed, feed rate, depth of cut, machining temperature, and vibration frequency. A step change refers to a significant and sudden change in these parameters within a short period, rapidly switching from one stable state to another. This change is often accompanied by drastic shifts in equipment operating conditions, representing a crucial opportunity to observe the controller's fine-tuning compensation behavior. Furthermore, the preset timing deviation analysis model can be a timing analysis model based on statistics, machine learning, or deep learning, such as a Kalman filter, support vector machine, or recurrent neural network. Its parameter configuration may include model weights, learning rate, window size, and threshold. Dynamically adjusting these parameters aims to enable the model to better adapt to the new operating conditions following a step change in the current equipment operating parameters, thereby improving the accuracy of capturing and analyzing sub-millisecond timing deviations.

[0040] It's important to clarify that continuously monitoring the nonlinear coupling relationships between operating parameters of precision machining equipment refers to using advanced data analysis techniques, such as nonlinear regression models, neural networks, or support vector machines, to analyze the nonlinear dependencies between multiple operating parameters under different working conditions, including spindle speed, feed rate, tool wear, and ambient temperature. The changing patterns of these nonlinear coupling relationships, and abrupt changes in coupling strength within specific load or speed ranges, can reflect the complex internal adjustments made by the local controller to maintain machining accuracy. Specifically, dynamically adjusting sub-millisecond timing deviations based on identified nonlinear coupling patterns can be understood as adaptively adjusting the analysis model or threshold for sub-millisecond timing deviations based on real-time changes in the nonlinear coupling relationships between equipment operating parameters. When a certain nonlinear coupling pattern is detected indicating that the controller is performing a specific nonlinear compensation, the sensitivity or focus on sub-millisecond timing deviations can be adjusted accordingly to more accurately capture the timing characteristics related to this compensation behavior. In practical applications, the impact of sub-millisecond timing deviations is decoupled through multidimensional eigenvalue decomposition, using methods such as Principal Component Analysis (PCA), Independent Component Analysis (ICA), or Nonnegative Matrix Factorization (NMF). This allows for the isolation of timing deviation components directly related to the fine-tuning compensation behavior of the local controller, eliminating interference from other irrelevant factors and thus obtaining the decoupled sub-millisecond timing deviation. Furthermore, based on the correlation between the decoupled sub-millisecond timing deviation and the operating parameters of the precision machining equipment, the fine-tuning compensation logic currently employed by the local controller can be inferred. This means that, after eliminating other interfering factors, a more accurate correlation analysis can be performed on the timing deviations purely caused by controller fine-tuning and the equipment operating parameters (such as target position, actual position, speed commands, etc.).

[0041] It's important to note that historical clock data can be understood as a record of the deviation between the local clock and the standard time of an edge computing node when it is calibrated with a standard time server (NTP server) at different points in time. Using these historical deviation values ​​and calibration time points, a clock drift prediction model can be constructed, such as one based on linear regression, multinomial fitting, or more complex machine learning models. This model can predict the amount of clock drift that may have occurred since the last calibration based on the current locally generated timestamp, thus obtaining an initial correction to the local timestamp (the first time correction) and the potential range of uncertainty for this correction (the first uncertainty range). The purpose is to initially eliminate the systematic drift of the edge computing node's local clock caused by factors such as crystal oscillator characteristics and temperature changes. The cumulative uncertainty of clock drift typically increases with the increase of time intervals. Therefore, by calculating the time interval between the locally generated timestamp and the most recent NTP calibration time point, this growth in uncertainty can be quantified. In practical applications, network transmission characteristic parameters refer to the time information recorded at different stages during the process of data packets being sent from the edge computing node to the cloud platform. The local send timestamp is the time recorded by the local clock when a data packet leaves the edge computing node, while the data packet arrival timestamp is the time recorded by the cloud platform clock when the data packet arrives at the cloud platform. By analyzing these timestamp pairs of a large number of data packets, the average one-way delay of data packets in the network and the range of delay fluctuation (i.e., delay jitter range) can be estimated. The delay jitter range reflects the impact of factors such as network congestion and routing changes on transmission time, and is therefore used as the third uncertainty range of the timestamp. Its purpose is to quantify and incorporate the time uncertainty brought about by network transmission. The target timestamp is obtained by superimposing the locally generated timestamp with the first time correction amount, and it represents the data packet generation time after preliminary clock drift correction. Subsequently, the second uncertainty range, reflecting the cumulative uncertainty of clock drift, and the third uncertainty range, reflecting the uncertainty of network transmission, are fused. The fusion method can employ statistical methods to obtain a comprehensive final confidence interval that can cover all sources of uncertainty. This confidence interval defines a probability range within which the actual generation time of the data packet has a high degree of confidence.

[0042] See Figure 2 , Figure 2 This is a schematic diagram of a cloud-based electrical equipment monitoring data processing system provided in one embodiment of this application. The cloud-based electrical equipment monitoring data processing system 200 includes: The receiving module 210 is used to receive data packets from the edge computing node; The confidence interval determination module 220 is used to determine the timestamp confidence interval of the data packet based on the historical clock data of the edge computing node, the calibration time, and the network transmission characteristic parameters. The data collection module 230 is used to dynamically adjust the data collection window based on the timestamp confidence interval, and to collect data packets with overlapping confidence intervals in the data collection window into an event snapshot; The data alignment module 240 is used to calculate the event correlation strength between the data packets collected in the event snapshot, and to obtain the data packets with high correlation strength according to the event correlation strength and the degree of overlap of the timestamp confidence intervals, and to prioritize the alignment of the data packets with high correlation strength. The snapshot reconstruction and output module 250 is used to reconstruct the device condition snapshot based on the aligned data packets with high correlation strength, and output the device condition snapshot to the device health assessment system.

[0043] It should be noted that the information interaction and execution process between the above modules are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, which will not be repeated here.

[0044] It will be understood by those skilled in the art that all or some of the steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, as is known to those skilled in the art, communication media typically include computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0045] The foregoing has provided a detailed description of the preferred embodiments of this application. However, this application is not limited to the above-described embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of this application. All such equivalent modifications or substitutions are included within the scope defined in this application.

Claims

1. A method for processing monitoring data of electrical equipment based on cloud computing, characterized in that, include: Receive data packets from edge computing nodes; The timestamp confidence interval of the data packet is determined based on the historical clock data of the edge computing node, the calibration time, and the network transmission characteristic parameters. Based on the timestamp confidence interval, the data collection window is dynamically adjusted, and the data packets with overlapping confidence intervals in the data collection window are collected into an event snapshot. Calculate the event correlation strength between the data packets collected in the event snapshot, and based on the event correlation strength and the degree of overlap of the timestamp confidence intervals, obtain the data packets with high correlation strength and prioritize aligning the data packets with high correlation strength; The device condition snapshot is reconstructed based on the aligned data packets with high correlation strength, and the device condition snapshot is output to the device health assessment system. The timestamp confidence interval represents the interval into which the actual generation time of the data packet falls; The step of calculating the event correlation strength between the data packets collected in the event snapshot, and obtaining data packets with high correlation strength based on the event correlation strength and the overlap of the timestamp confidence intervals, and prioritizing the alignment of the data packets with high correlation strength, includes: Based on the edge computing nodes monitoring the internal sensor data of the precision machining equipment, the fine-tuning compensation behavior of the local controller is inferred, the inferred compensation amount is obtained, and the expected timing delay between the precision machining equipment is corrected based on the inferred compensation amount, wherein the edge computing nodes correspond one-to-one with the precision machining equipment; The actual relative timing delay between the precision machining equipment is analyzed, the persistent deviation from the expected timing delay is identified, and the confidence level of the physical causal relationship between the precision machining equipment is updated based on the statistical characteristics of the persistent deviation, thus obtaining the updated confidence level of the physical causal relationship. The overlap of the timestamp confidence intervals is dynamically adjusted based on the updated physical causality confidence level and the inference compensation amount. Based on the degree of overlap of the dynamically adjusted timestamp confidence intervals, the event association strength between the data packets collected in the event snapshot is calculated, and based on the event association strength and the degree of overlap of the timestamp confidence intervals, data packets with high association strength are obtained and prioritized for alignment. Determining the timestamp confidence interval of the data packet based on the historical clock data, calibration time, and network transmission characteristic parameters of the edge computing node includes: Based on the historical clock data, a clock drift prediction model for the edge computing node is constructed. The estimated drift amount corresponding to the locally generated timestamp is calculated through the clock drift prediction model to obtain a first time correction amount and a first uncertainty range. The historical clock data includes the deviation value between the local clock and the standard time of the NTP server after each NTP calibration of the edge computing node and the time point of each calibration. Query the time point of the most recent calibration of the edge computing node, calculate the time interval between the locally generated timestamp and the time point of the most recent calibration, and adjust the first uncertainty range according to the time interval to obtain the second uncertainty range; The network transmission characteristic parameters of the data packets are collected, including the local sending timestamp recorded by the edge computing node and the data packet arrival timestamp recorded by the cloud platform; the one-way network transmission delay and delay jitter range are estimated based on the network transmission characteristic parameters, and the delay jitter range is used as the third uncertainty range; The locally generated timestamp is corrected by combining the first time correction amount to obtain the target timestamp; the second uncertainty range and the third uncertainty range are fused to determine the timestamp confidence interval of the data packet.

2. The method according to claim 1, characterized in that, The analysis of the actual relative timing delay between the precision machining equipment identifies persistent deviations from the expected timing delay, and updates the confidence level of the physical causality relationship between the precision machining equipment based on the statistical characteristics of the persistent deviations, resulting in an updated confidence level of the physical causality relationship, including: Acquire real-time sensor data and perform high-frequency sampling on the real-time sensor data to obtain high-frequency sampled data; transmit the high-frequency sampled data to the edge computing node to calculate the instantaneous relative timing delay between the precision machining equipment in real time; Statistical analysis is performed on the instantaneous relative timing delay within each preset time window to obtain the deviation characteristics at different time granularities; By comparing the deviation characteristics at different time granularities, deviation patterns with high transientity, sporadic occurrence and extremely short duration are identified, and the deviation patterns are compared with the expected time delay. Based on the frequency, amplitude, and degree of deviation of the deviation pattern from the preset process flow rules, the confidence level of the physical causal relationship between the precision machining equipment is updated to obtain the updated confidence level of the physical causal relationship.

3. The method according to claim 1, characterized in that, The monitoring of internal sensor data of the precision machining equipment is used to infer the fine-tuning compensation behavior of the local controller, obtain the inferred compensation amount, and correct the expected timing delay between the precision machining equipment based on the inferred compensation amount, including: The system continuously monitors the internal control signals emitted by the local controller and the internal sensor data of the precision machining equipment, and synchronizes them in real time through the edge computing node. Detect whether there is a continuous, non-preset sub-millisecond timing deviation between the internal control signal and the actual execution result reflected by the internal sensor data; When there is a continuous, non-preset sub-millisecond timing deviation between the actual execution results, the dynamic fine-tuning compensation amount being executed by the local controller is obtained; Based on the persistence, magnitude, and correlation with the operating parameters of the precision machining equipment of the sub-millisecond timing deviation, the fine-tuning compensation logic currently used by the local controller is inferred and the dynamic fine-tuning compensation amount is quantified to obtain the inferred compensation amount. Based on the fine-tuning compensation logic and the inferred compensation amount, the expected timing delay between the precision machining equipment is corrected.

4. The method according to claim 3, characterized in that, The continuous monitoring of internal control signals from the local controller and internal sensor data from the precision machining equipment also includes: The hardware clock source of the hardware time synchronization module timestamps the captured internal control signals and sensor data, and synchronizes the time with the standard time server through a preset hardware time synchronization protocol.

5. The method according to claim 3, characterized in that, When there is a continuous, non-preset sub-millisecond timing deviation between the actual execution results, the dynamic fine-tuning compensation amount being executed by the local controller is obtained, including: When there is a continuous, non-preset sub-millisecond timing deviation between the actual execution results, analyze the dynamic change pattern of the sub-millisecond timing deviation; Based on the dynamic change pattern of the sub-millisecond timing deviation, and combined with the preset compensation behavior feature library, the feature fingerprint of the fuzzy control rules or neural network compensation behavior that the local controller may have is identified. By performing real-time clustering and pattern matching on the feature fingerprint, the type of nonlinear fine-tuning compensation logic currently used by the local controller can be inferred. Based on the inferred nonlinear fine-tuning compensation logic type, the corresponding inverse compensation algorithm is dynamically selected and the compensation amount is calculated to obtain the dynamic fine-tuning compensation amount being executed by the local controller.

6. The method according to claim 5, characterized in that, The analysis of the dynamic change pattern of the sub-millisecond time series deviation includes: The operating parameters of the precision machining equipment were monitored to show a step change; When the operating parameters of the precision machining equipment undergo a step change, the parameter configuration of the preset time-series deviation analysis model is dynamically adjusted. Under the adjusted preset timing deviation analysis model, the timing deviation analysis process is re-initialized and the analysis is performed. Sub-millisecond timing deviations are used to obtain the dynamic change pattern of the sub-millisecond timing deviations.

7. The method according to claim 3, characterized in that, The step of inferring the fine-tuning compensation logic currently used by the local controller based on the persistence, magnitude, and correlation with the operating parameters of the precision machining equipment of the sub-millisecond timing deviation includes: Continuously monitor the nonlinear coupling relationship between the operating parameters of the precision machining equipment and identify the changing patterns of the nonlinear coupling relationship; Based on the identified change pattern of the nonlinear coupling relationship, the sub-millisecond timing deviation is dynamically adjusted; The influence of the sub-millisecond time series deviation is decoupled by multidimensional eigenvalue decomposition to obtain the decoupled sub-millisecond time series deviation. Based on the correlation between the decoupled sub-millisecond timing deviation and the operating parameters of the precision machining equipment, the fine-tuning compensation logic currently used by the local controller is inferred.

8. A cloud computing-based data processing system for monitoring electrical equipment, characterized in that, include: The receiving module is used to receive data packets from edge computing nodes; The confidence interval determination module is used to determine the timestamp confidence interval of the data packet based on the historical clock data of the edge computing node, the calibration time, and the network transmission characteristic parameters. The data collection module is used to dynamically adjust the data collection window according to the timestamp confidence interval, and to collect the data packets with overlapping confidence intervals in the data collection window into an event snapshot; The data alignment module is used to calculate the event correlation strength between the data packets collected in the event snapshot, and to obtain data packets with high correlation strength based on the event correlation strength and the degree of overlap of the timestamp confidence intervals, and to prioritize align the data packets with high correlation strength. The snapshot reconstruction and output module is used to reconstruct a device condition snapshot based on the aligned data packets with high correlation strength, and output the device condition snapshot to the device health assessment system. The timestamp confidence interval represents the interval into which the actual generation time of the data packet falls; The step of calculating the event correlation strength between the data packets collected in the event snapshot, and obtaining data packets with high correlation strength based on the event correlation strength and the overlap of the timestamp confidence intervals, and prioritizing the alignment of the data packets with high correlation strength, includes: Based on the edge computing nodes monitoring the internal sensor data of the precision machining equipment, the fine-tuning compensation behavior of the local controller is inferred, the inferred compensation amount is obtained, and the expected timing delay between the precision machining equipment is corrected based on the inferred compensation amount, wherein the edge computing nodes correspond one-to-one with the precision machining equipment; The actual relative timing delay between the precision machining equipment is analyzed, the persistent deviation from the expected timing delay is identified, and the confidence level of the physical causal relationship between the precision machining equipment is updated based on the statistical characteristics of the persistent deviation, thus obtaining the updated confidence level of the physical causal relationship. The overlap of the timestamp confidence intervals is dynamically adjusted based on the updated physical causality confidence level and the inference compensation amount. Based on the degree of overlap of the dynamically adjusted timestamp confidence intervals, the event association strength between the data packets collected in the event snapshot is calculated, and based on the event association strength and the degree of overlap of the timestamp confidence intervals, data packets with high association strength are obtained and prioritized for alignment. Determining the timestamp confidence interval of the data packet based on the historical clock data, calibration time, and network transmission characteristic parameters of the edge computing node includes: Based on the historical clock data, a clock drift prediction model for the edge computing node is constructed. The estimated drift amount corresponding to the locally generated timestamp is calculated through the clock drift prediction model to obtain a first time correction amount and a first uncertainty range. The historical clock data includes the deviation value between the local clock and the standard time of the NTP server after each NTP calibration of the edge computing node and the time point of each calibration. Query the time point of the most recent calibration of the edge computing node, calculate the time interval between the locally generated timestamp and the time point of the most recent calibration, and adjust the first uncertainty range according to the time interval to obtain the second uncertainty range; The network transmission characteristic parameters of the data packet are collected, including the local sending timestamp recorded by the edge computing node and the data packet arrival timestamp recorded by the cloud platform; the one-way network transmission delay and delay jitter range are estimated based on the network transmission characteristic parameters, and the delay jitter range is used as the third uncertainty range; the locally generated timestamp is corrected by combining the first time correction amount to obtain the target timestamp; the second uncertainty range and the third uncertainty range are fused to determine the timestamp confidence interval of the data packet.

Images