Industrial computer terminal data processing method and system thereof

Abstract

The application discloses an industrial computer terminal data processing method and system, relates to the technical field of data processing, and comprises the following specific steps: starting the industrial computer, collecting data, and enabling the industrial computer to perform self-checking; processing the data collected by the industrial computer; generating and outputting control instructions according to the data processed and analyzed; enabling the industrial computer to perform periodic equipment self-checking; issuing an alarm for an abnormal event obtained through detection and analysis of the industrial computer, and establishing an equipment maintenance preset scheme library to process the abnormal event according to a preset scheme in the equipment maintenance preset scheme library; and evaluating and optimizing the operation effect of the industrial computer. The industrial computer terminal data processing method is used to check the integrity of core components through digital signature, eliminate malicious intrusion from the source, avoid resource conflicts in the running stage, cover hardware, time synchronization and collection channels through multi-dimensional self-checking, locally record slight faults, block the start of serious faults and report the same, and front-load security risks.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a data processing method and system for an industrial control computer terminal. Background Technology

[0002] Industrial control computers (ICCs) are ruggedized computers specifically designed for industrial environments. They are the core computing and control carriers of industrial automation systems, undertaking key functions such as data processing, equipment control, and status monitoring in industrial scenarios. They are the basic hardware support for realizing intelligent manufacturing and the industrial internet. As the control center of automated production lines and intelligent equipment, ICCs can receive production data collected by sensors and output control commands through preset algorithms to achieve precise automated control of the production process, replacing traditional manual operations and significantly improving production efficiency and accuracy.

[0003] Currently, industrial control computers generally lack a full-link security verification mechanism during the startup phase. Core firmware and operating system images are easily tampered with maliciously. Furthermore, system resources such as CPU, memory, and I / O are not dynamically allocated according to the amount of data acquisition tasks. Data acquisition delays or startup failures are often caused by resource contention. Minor faults are easily overlooked, and serious faults cannot be prevented in advance, posing safety hazards to industrial production. Moreover, most existing industrial control computers adopt a single-cycle self-test mode, which cannot dynamically adjust the self-test frequency according to the equipment failure rate and operating parameters. Faults of core components are difficult to detect in a timely manner, and fault handling relies on human experience, which is inefficient. Equipment maintenance decisions lack quantitative basis and are based solely on subjective judgment to decide on repair or replacement, which can easily lead to resource waste or downtime risks. Summary of the Invention

[0004] The purpose of this invention is to provide a data processing method and system for an industrial control computer terminal to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a data processing method for an industrial control computer terminal, comprising the following specific steps:

[0006] Step 1: Start the industrial control computer, collect data, and complete a multi-dimensional self-test including hardware self-test, time synchronization verification, and acquisition channel connectivity test;

[0007] Step 2: Process the data collected by the industrial control computer, and perform real-time calculation and analysis on the data. The data is classified into real-time control, status monitoring, and anomaly categories for hierarchical and differentiated processing.

[0008] Step 3: Generate and output control commands based on the processed and analyzed data, and perform rule verification and multi-device collaborative conflict verification on the control commands based on process conditions and safety constraints. Priority commands are issued through a hard real-time channel, and ordinary commands are issued through a reliable transmission protocol. The execution results are tracked in real time, and the commands are dynamically adjusted when the deviation exceeds the threshold.

[0009] Step 4: Enable the industrial control computer to perform periodic equipment self-checks, compare the self-check data with the equipment baseline model to identify potential anomalies, and perform local calibration, early warning reporting, and safety interlock switching of redundant equipment operations after anomaly classification.

[0010] Step 5: Issue alarms for abnormal events detected and analyzed by the industrial control computer, establish a preset equipment maintenance plan library, match handling plans from the maintenance plan library based on fault characteristics, perform standardized handling for known faults with one click, recommend similar solutions for unknown faults through knowledge graph and trigger expert consultation, and output maintenance and replacement decisions through quantitative calculation formulas for equipment replacement and maintenance.

[0011] Step Six: Evaluate and optimize the operation of the industrial control computer, establish a database, regularly evaluate the accuracy of data acquisition, the success rate of instruction execution, and the accuracy of anomaly identification, optimize the anomaly detection model based on historical data, and dynamically adjust the acquisition frequency and control logic. Adopt a hierarchical data management system that uses local SSD storage for hot data and cloud object storage for cold data.

[0012] Preferably, the self-test of the industrial control computer in step one includes the following steps:

[0013] S1.1: After the industrial control computer is powered on, it first loads the secure and trusted execution environment to verify the digital signatures of the core firmware, operating system image, and data acquisition driver;

[0014] S1.2: Pre-allocate CPU cores, memory buffers, and disk I / O bandwidth according to the preset data acquisition task volume;

[0015] S1.3: Perform hardware self-test, time synchronization verification, and acquisition channel connectivity test in sequence. Hardware self-test includes the health of acquisition port, communication module and storage medium. If the self-test is abnormal, trigger a graded alarm. The graded alarm includes warning abnormality and fault abnormality. Warning abnormality is recorded locally. Fault abnormality prevents startup and reports to operation and maintenance.

[0016] S1.4: Retrieve the list of currently running data collection tasks from the configuration center, and determine the data collection objects, collection frequency, and data verification database.

[0017] Preferably, the data processing and real-time computing in step two includes data cleaning, edge-side real-time analysis, and data hierarchical distribution. Data cleaning preprocesses the original collected data, removes outliers, fills in missing values, and unifies the data format. Edge-side real-time analysis involves performing low-latency computing on high-frequency data locally and using a streaming computing framework. The streaming computing framework supports multi-point correlation analysis. Data hierarchical distribution involves real-time control data directly entering the instruction generation link, prioritizing low latency, while status monitoring data is synchronously uploaded to the cloud platform for long-term trend analysis. Abnormal data is separately marked and stored in a high-priority queue for in-depth analysis.

[0018] Preferably, the generation and output of control commands in step three includes the following steps:

[0019] S3.1: Based on preset process rules and safety constraints, perform dual verification on the generated control commands. The dual verification includes rule verification and conflict verification. Rule verification checks whether the command conforms to the safe operating range of the equipment. Conflict verification checks whether there are logical conflicts in the command in the multi-device collaborative scenario. If a logical conflict occurs, the adjustment command is regenerated.

[0020] S3.2: Priority control commands are sent via a hard real-time channel, while ordinary adjustment commands use a reliable transmission protocol, supporting command retransmission and timeout retransmission;

[0021] S3.3: After the instruction is issued, the device execution result is tracked in real time. If the execution fails or the deviation from the expected value exceeds the threshold, a secondary judgment is immediately triggered to regenerate the adjustment instruction.

[0022] Preferably, the periodic device self-test in step four is performed in a tiered manner, which is divided into minute-level self-test, hour-level self-test, and daily-level self-test according to the device model, operating parameters, and historical failure rate. The minute-level self-test is for connectivity testing of the core acquisition port and communication link, the hour-level self-test is for testing of key operating parameters of the device, and the daily-level self-test is for comprehensive hardware health testing, acquisition accuracy calibration, and bad block scanning of the storage medium.

[0023] Preferably, the comparison of periodic equipment self-test data with historical equipment data in step four involves comparing the self-test data with the equipment baseline model to identify potential anomalies. These potential anomalies are categorized into Level 1, Level 2, and Level 3 anomalies. Level 1 anomalies are automatically calibrated locally and recorded in the operation log. Level 2 anomalies are reported to the operation and maintenance platform to trigger an early warning. Level 3 anomalies immediately trigger a safety interlock, switch redundant equipment, synchronously report fault details, and retain all timestamps associated with the anomaly self-test data and the complete equipment operating context to support subsequent fault tracing.

[0024] Preferably, the alarm levels for abnormal events in step five are divided into warning level and fault level. Warning level events notify maintenance personnel via SMS and the background records pending work orders. Fault level events notify the responsible person via voice call and simultaneously trigger the linkage plan. Based on the characteristics of abnormal events, the handling plan is automatically matched from the maintenance plan library. Abnormal events are divided into known faults and unknown faults. For known faults, standardized handling steps, required spare parts, and historical handling cases are directly pushed, supporting one-click automated handling. For unknown faults, similar fault handling plans are recommended based on knowledge graphs, and the expert remote consultation process is triggered simultaneously. The diagnostic conditions are input, including whether there is a line fault, whether the equipment is damaged, whether the equipment fault is clear, and whether the repair time and cost are acceptable. The equipment fault handling method is determined based on the diagnostic conditions, and the equipment replacement and maintenance calculation formula is established based on the diagnostic conditions as constraints. The fault handling method is confirmed by the replacement and maintenance calculation formula. The fault handling method includes repair and replacement, which is used to assist maintenance personnel in confirming the equipment replacement and maintenance plan. After the abnormal event handling is completed, the handling process, time consumption, and effect are recorded, and the adaptability score of the plan library is updated to continuously optimize the accuracy of plan matching.

[0025] Preferably, the calculation formula for equipment replacement and maintenance is as follows:

[0026]

[0027] in, , , and These are the weighting coefficients. , , and The values ​​are 0.3, 0.3, 0.2, and 0.2 respectively. To determine if there is a line fault, To determine if the equipment is damaged, Is the repair time and cost acceptable? To determine whether the equipment malfunction is clear, , , and The value can be 1 or 0, where 1 represents yes and 0 represents no. If the calculated value is greater than or equal to 0.5, then the faulty equipment should be repaired. If the calculated value is less than 0.5, the faulty device should be replaced.

[0028] Preferably, the industrial control computer (ICC) operation performance evaluation in step six involves periodically analyzing ICC operation data to assess data acquisition accuracy, instruction execution success rate, and anomaly identification accuracy, identifying process bottlenecks. ICC operation optimization is based on optimizing the anomaly detection model using historical data to reduce false alarm and false negative rates. It also involves adjusting the acquisition frequency, calculation database, and control logic according to the actual operating scenario to reduce system resource consumption and improve processing efficiency. When storing ICC operation data, the database divides the data into hot data and cold data for layered processing. Hot data is stored on a local SSD for fast querying, while cold data is compressed and uploaded to cloud object storage.

[0029] Another objective of this invention is to provide an industrial control computer terminal data processing system, comprising:

[0030] The data acquisition module is used to collect data from the industrial control computer in real time.

[0031] The data processing module is used to process the data collected by the data acquisition module and perform data formatting and uniform conversion.

[0032] Self-test module, the self-test module is used for self-testing the operating performance of the industrial control computer;

[0033] A calculation module, which stores calculation formulas and runs the calculation formulas to calculate and process data;

[0034] A control instruction module is used to generate and verify instructions, perform real-time tracking during instruction execution, and trigger secondary judgment and dynamic adjustment of instructions based on deviation thresholds.

[0035] The equipment health management module is used for periodic self-inspection of equipment and analysis and judgment of abnormal faults;

[0036] The equipment maintenance module is used to handle abnormal faults and formulate maintenance and replacement plans based on the on-site inspection.

[0037] The communication module provides a signal transmission channel for the operation of the industrial control computer.

[0038] Data storage module, wherein the data storage module is used for hierarchical data storage;

[0039] A visualization interaction module is used to display the device's operating status in real time and supports remote consultation interaction.

[0040] The technical effects and advantages of this invention are as follows:

[0041] (1) This invention utilizes the data processing method of industrial control computer terminal, verifies the integrity of core components through digital signature, eliminates malicious intrusion from the source, dynamically allocates system resources according to the collection task, avoids resource conflicts during operation, and performs multi-dimensional self-checks covering hardware, time synchronization, and collection channels. Minor faults are recorded locally, and serious faults are blocked from starting and reported, thus putting security risks forward.

[0042] (2) This invention utilizes the method of equipment replacement and maintenance calculation. After the maintenance worker arrives at the site to check the equipment fault, he can directly input the conditions of whether there is a line fault, whether the equipment is damaged, whether the repair time cost is acceptable, and whether the equipment fault is clear into the background. Through the equipment replacement and maintenance calculation formula, it can be directly determined whether the equipment needs maintenance or replacement, thereby improving the efficiency of equipment fault handling.

[0043] (3) This invention utilizes a three-level self-inspection mechanism of minutes, hours and days to conduct high-frequency inspections of core components and in-depth inspections of all hardware, thereby improving the accuracy of anomaly identification. Alarms are linked to SMS or voice notifications at different levels. Known faults can be handled in a standardized manner with one click, and unknown faults can trigger remote expert consultation. The equipment replacement and maintenance calculation formula outputs decisions by quantifying parameters, thereby improving maintenance efficiency and reducing resource waste. Attached Figure Description

[0044] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention, but do not constitute a limitation thereof. In the drawings:

[0045] Figure 1 This is a flowchart of the industrial control computer terminal data processing method of the present invention;

[0046] Figure 2 This is a logic diagram for equipment fault maintenance and replacement according to the present invention. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] This invention provides, for example Figures 1-2 The industrial control computer terminal data processing method shown includes the following specific steps:

[0049] Step 1: Start the industrial computer, collect data, and complete multi-dimensional self-checks including hardware self-test, time synchronization verification, and acquisition channel connectivity test. This improves the safety of the industrial computer during operation and reduces the probability of equipment damage during startup.

[0050] Step 2: Process the data collected by the industrial control computer and perform real-time calculation and analysis to understand the operation of the control computer in real time. The data is classified into real-time control, status monitoring, and anomaly categories for hierarchical and differentiated processing.

[0051] Step 3: Generate and output control commands based on the processed and analyzed data. Perform rule verification and multi-device collaborative conflict verification on the control commands based on process conditions and safety constraints. In the industrial control computer terminal data processing method, process conditions refer to a series of operating rules and parameter ranges formulated based on production process, technical standards and quality requirements in industrial production. Safety constraints are mandatory restrictions set to ensure personnel safety, equipment stability and production continuity. Together, they constitute the core basis for industrial control computer data processing and command generation. Priority commands are issued through hard real-time channels and ordinary commands are issued through reliable transmission protocols. The execution results are tracked in real time and the commands are dynamically adjusted when the deviation exceeds the threshold.

[0052] Step 4: Perform periodic equipment self-checks on the industrial control computer. Compare the self-check data with the equipment baseline model to identify potential anomalies. The equipment baseline model is a benchmark reference system for equipment performance, built using statistical analysis, machine learning, and other methods based on historical data accumulated by the industrial control computer equipment under normal operating conditions. It includes core features such as the normal range, fluctuation patterns, and correlations of key operating parameters of the equipment. It serves as a benchmark for judging whether the equipment's operating status is abnormal. After anomaly classification, local calibration, early warning reporting, and safety interlock switching of redundant equipment operations are performed respectively, thereby improving the timeliness of fault detection and facilitating maintenance of any faults that occur.

[0053] Step 5: Issue alarms for abnormal events detected and analyzed by the industrial control computer, establish a preset equipment maintenance plan library, match handling plans from the maintenance plan library based on fault characteristics, perform standardized handling for known faults with one click, recommend similar solutions for unknown faults through knowledge graph and trigger expert consultation, and output maintenance and replacement decisions through quantitative calculation formulas for equipment replacement and maintenance.

[0054] Step Six: Evaluate and optimize the operation of the industrial control computer, establish a database, regularly evaluate the accuracy of data acquisition, the success rate of instruction execution, and the accuracy of anomaly identification, optimize the anomaly detection model based on historical data, and dynamically adjust the acquisition frequency and control logic. Adopt a hierarchical data management system with hot data stored locally on SSDs and cold data stored in cloud object storage to facilitate subsequent data traceability.

[0055] Furthermore, the self-test of the industrial control computer in step one includes the following steps:

[0056] S1.1: After the industrial control computer is powered on, it first loads a secure and trusted execution environment to verify the digital signatures of the core firmware, operating system image, and data acquisition driver, preventing malicious tampering and improving the security of the industrial control computer's operation;

[0057] S1.2: Based on the preset data collection task volume, pre-allocate CPU cores, memory buffers, and disk I / O bandwidth to avoid resource contention during operation;

[0058] S1.3: Perform hardware self-test, time synchronization verification, and acquisition channel connectivity test sequentially. Hardware self-test includes the health of the acquisition port, communication module, and storage medium. Time synchronization verification error with the time synchronization server should be ≤1ms. A tiered alarm will be triggered if a self-test anomaly occurs. Tiered alarms include warning anomalies and fault anomalies. Warning anomalies are recorded locally. Warning anomalies include minor issues found during hardware self-test that do not affect core functions, such as slightly high response latency of individual channels on the acquisition port, a small number of repairable bad blocks on the storage medium, slightly low signal strength of the communication module but still within the normal range, and time synchronization verification error within acceptable limits, such as when the time synchronization error is within the normal range. The server error is ≤1ms but close to the threshold. A brief fluctuation occurred during the connectivity test of the acquisition channel, but it has recovered automatically. The fault anomaly prevents startup and reports to the operation and maintenance department. The fault anomaly includes core functional failures found by hardware self-test, such as large-scale unresponsive acquisition ports, complete failure of communication modules, storage media health below the safety threshold, failure of key hardware component verification, time synchronization verification error exceeding the allowable range, error with the time synchronization server >1ms, and acquisition channel connectivity test finding that multiple channels are completely interrupted and unable to establish a data transmission link. The fault anomaly prevents startup and reports to the operation and maintenance department to avoid direct startup of the industrial control computer, which may cause equipment damage.

[0059] S1.4: Retrieve the list of currently running data acquisition tasks from the configuration center, determine the acquisition objects, acquisition frequency, and data verification database. The acquisition objects are determined to be sensors, PLCs, machine tool controllers, etc. The acquisition frequency is 10ms / time for high-frequency data and 1min / time for low-frequency data. The data verification database includes range verification and format verification. Range verification verifies whether numerical data falls within the reasonable value range defined by business rules, also known as threshold verification or range verification. Format verification verifies whether string data conforms to preset encoding rules, syntax structure, or format specifications, also known as pattern verification.

[0060] Specifically, step two, data processing and real-time computation, includes data cleaning, edge-side real-time analysis, and data hierarchical distribution. Data cleaning preprocesses the raw acquired data, removing outliers, filling in missing values, and standardizing the data format. Outlier removal includes removing values ​​exceeding the sensor's range and values ​​that jump beyond a threshold. Missing value filling is based on time series interpolation or filling with the average of similar equipment. Standardizing the data format involves converting the data into a standard industrial data model, including timestamps, equipment IDs, measurement point IDs, values, and unit fields. Edge-side real-time analysis involves performing low-latency computations on high-frequency data locally using a streaming computing framework. Low-latency computations include frequency domain analysis of vibration signals, calculation of temperature / pressure change rates, and real-time determination of equipment operating status. The streaming computing framework supports multi-measurement point correlation analysis, such as determining equipment failure risk by combining temperature exceeding a threshold with abnormal current increases. Data hierarchical distribution ensures that real-time control data directly enters the command generation chain, prioritizing low latency and status monitoring. Measurement data is synchronously uploaded to the cloud platform for long-term trend analysis. This long-term trend analysis targets condition monitoring data collected by industrial control computers, such as periodic non-real-time control data like equipment temperature, pressure, vibration, and energy consumption. After being uploaded to the cloud platform, cross-time dimension data analysis is conducted. By comparing historical baselines with current operating data, the gradual change patterns and potential risks in equipment performance are identified. A benchmark threshold and feature fingerprint library trained based on historical normal operating data serve as the criterion for performance degradation. Time series prediction, sliding window statistics, and correlation analysis are used to identify abnormal inflection points in data trends. Abnormal data is individually marked and stored in a high-priority queue for in-depth analysis. In-depth analysis is based on a streaming computing framework, simultaneously analyzing data from multiple monitoring points of the same or related equipment. It correlates historical fault cases, equipment manuals, and expert experience to recommend handling solutions for similar faults. By training models with labeled fault data, hidden fault features are mined to achieve intelligent diagnosis of unknown faults.

[0061] Furthermore, step three, the generation and output of control instructions, includes the following steps:

[0062] S3.1: Based on preset process rules and safety constraints, the generated control commands are double-checked. The double check includes rule check and conflict check. Rule check is to check whether the command conforms to the safe operating range of the equipment. For example, the valve opening adjustment range does not exceed 20% / time. Conflict check is to check whether there is a logical conflict in the command in the multi-device collaborative scenario. When a logical conflict occurs, the adjustment command is regenerated to improve the accuracy of the command.

[0063] S3.2: Priority control commands, such as safety interlocks, are sent via a hard real-time channel with an end-to-end delay of ≤10ms. Ordinary adjustment commands use a reliable transmission protocol, supporting command retransmission and timeout retransmission to ensure a 100% command delivery rate.

[0064] S3.3: After the instruction is issued, the device execution result is tracked in real time. If the execution fails or the deviation from the expected value exceeds the threshold, a secondary judgment is immediately triggered to regenerate the adjustment instruction.

[0065] Furthermore, in step four, the periodic equipment self-test is performed in a tiered manner. Based on the equipment model, operating parameters, and historical failure rate, it is divided into minute-level, hourly-level, and daily-level self-tests. The minute-level self-test checks the connectivity of the core acquisition port and communication link. The hourly-level self-test checks key operating parameters of the equipment, such as CPU utilization, memory usage, disk space remaining, and sensor power supply voltage. The daily-level self-test performs a comprehensive hardware health check, acquisition accuracy calibration, and bad block scanning of the storage medium. For equipment with a high failure rate, the self-test level can be manually entered, and the self-test frequency can be set as needed. The frequency of minute-level, hourly, and daily self-tests decreases progressively. Using three different cycles of minute-level, hourly, and daily self-tests for tiered testing can reduce the workload of self-testing and ensure the effectiveness of equipment self-testing by providing targeted testing.

[0066] Furthermore, the comparison of periodic equipment self-inspection data with historical equipment data in step four involves comparing the self-inspection data with the equipment baseline model. The equipment baseline model includes benchmark thresholds and feature fingerprint databases trained based on historical normal operation data. Potential anomalies are identified and classified into Level 1, Level 2, and Level 3 anomalies. Level 1 anomalies include, for example, acquisition accuracy deviation ≤2%. Level 1 anomalies involve local automatic calibration and recording operation logs. Level 2 anomalies include, for example, sensor data drift and component temperature rise. Level 2 anomalies are reported to the operation and maintenance platform, triggering an alert. Level 3 anomalies include, for example, communication module failure and storage failure. Level 3 anomalies immediately trigger safety interlocks, switch redundant equipment, synchronously report fault details, and retain all timestamps associated with the anomaly self-inspection data and the complete equipment operating context to support subsequent fault tracing.

[0067] In particular, the alarm levels for abnormal events in step five are divided into warning level and fault level. The warning level notifies the maintenance personnel via SMS and records the pending work order in the background. The fault level notifies the responsible person via voice call and triggers the linkage plan simultaneously, such as starting backup equipment and cutting off unnecessary loads. Based on the characteristics of the abnormal event, such as fault type, equipment model and operating scenario, the handling plan is automatically matched from the maintenance plan library. The automatic matching of the handling plan includes the following steps:

[0068] Based on the abnormal data detected by the industrial control computer, multi-dimensional feature tags are automatically extracted to provide accurate retrieval basis for contingency plan matching. The multi-dimensional feature tags include fault type code, equipment model ID, fault occurrence timestamp, abnormality level, associated measurement point parameters, and equipment operating context.

[0069] Based on the type of abnormal features, different matching algorithms are used to filter candidate solutions from the contingency plan library, including matching known faults and unknown faults. The matching degree between the features of known faults and existing fault cases in the contingency plan library is ≥90%. The standardized handling contingency plan is directly associated, including standardized operating steps, a list of required spare parts, historical handling cases, and automated handling interfaces. When there are no known fault cases that completely match the known cases, semantic reasoning is performed based on the knowledge graph to extract the fault feature vector, calculate the cosine similarity with historical faults in the contingency plan library, recommend the top 3 similar solutions, add an expert consultation trigger entry, push fault details to the expert terminal simultaneously, and automatically determine the handling direction based on the equipment replacement and maintenance calculation formula.

[0070] When multiple candidate solutions are matched, they are prioritized based on historical success rate, processing time, spare parts availability, and adaptability. The weight ratios of historical success rate, processing time, spare parts availability, and adaptability are 40%, 25%, 20%, and 15%, respectively, and the method with the highest priority is pushed to the system.

[0071] Furthermore, abnormal events are categorized into known and unknown faults. For known faults, standardized handling steps, required spare parts, and historical handling cases are directly pushed to the system, supporting one-click automated handling. For unknown faults, similar fault handling solutions are recommended based on a knowledge graph, including the following steps:

[0072] Build a knowledge graph and automatically extract multi-dimensional feature labels;

[0073] By using a pre-set diagnostic database, such as overheating, abnormal current, and heat dissipation system failure, the association paths in the knowledge graph are matched, the cosine similarity between the current fault feature vector and the historical fault feature vector in the knowledge graph is calculated, and similar fault cases with a similarity of ≥70% are selected. Through the association relationship of the knowledge graph, the hidden fault causal chain is mined, such as sensor failure leading to data abnormality, which in turn leads to control command deviation.

[0074] Standardized handling steps, required spare parts, and precautions are extracted from the matched similar failure cases. Combined with the equipment replacement and maintenance calculation formula, decision suggestions for priority maintenance or replacement are automatically generated. The content pushed includes: details of similar failure cases, a list of recommended handling steps, and access to expert consultation. Fault details are also pushed to the expert terminal at the same time. The details of similar failure cases include fault phenomena, handling process, and result feedback.

[0075] The knowledge graph recommends using an attribute graph model to store entities and relationships, supporting efficient deep traversal queries. It builds query indexes for high-frequency fault types, keeping graph query response time within 100ms. It supports incremental updates, integrating new fault cases and expert experience into the knowledge graph in real time. Rule-based reasoning is implemented using the Rete algorithm to quickly match complex combinations of fault features. Graph embedding technology is introduced to transform entities and relationships into low-dimensional vectors. It supports machine learning-based similarity calculations and uncertain reasoning. For fuzzy fault features, such as abnormal equipment operating sounds, probabilistic matching is performed. The weights of cases in the knowledge graph are updated based on the fault handling results, such as increasing the weight of successful cases by 10% and decreasing the weight of failed cases by 5%. New experiences from expert consultations and updates to industry standards are regularly integrated into the knowledge graph. A fault feature extraction model is trained based on historical data to improve the accuracy and efficiency of feature extraction.

[0076] The process of remote expert consultation is triggered synchronously, and the diagnostic information is input. The diagnostic information includes whether there is a line fault, whether the equipment is damaged, whether the equipment fault is clear, and whether the repair time and cost are acceptable. Based on the diagnostic information, the method of handling the equipment fault is determined, and the equipment replacement and maintenance calculation formula is established based on the diagnostic information as a constraint. The fault handling method is confirmed by the replacement and maintenance calculation formula. The fault handling method includes repair and replacement, which helps maintenance personnel to confirm the equipment replacement and maintenance plan, thereby reducing equipment maintenance costs and ensuring equipment maintenance efficiency. After the anomaly is handled, the handling process, time consumption and effect are recorded, and the adaptability score of the contingency plan library is updated to continuously optimize the accuracy of contingency plan matching.

[0077] In particular, the formula for calculating equipment replacement and maintenance is as follows:

[0078]

[0079] in, , , and These are the weighting coefficients. , , and The values ​​are 0.3, 0.3, 0.2, and 0.2 respectively. To determine if there is a line fault, To determine if the equipment is damaged, Is the repair time and cost acceptable? To determine whether the equipment malfunction is clear, , , and The value can be 1 or 0, where 1 represents yes and 0 represents no. If the calculated value is greater than or equal to 0.5, then the faulty equipment should be repaired. If the calculated value is less than 0.5, the faulty equipment will be replaced. The acceptability of the repair time cost is related to whether the equipment fault is clear. When the equipment fault is unclear, the repair time cost is considered unacceptable. Conversely, when the repair time cost is unacceptable, the equipment fault is only considered unclear. When the equipment fault is clear, the repair time cost can be either unacceptable or acceptable.

[0080] When there is a line fault, but the equipment is not damaged, the repair time and cost are acceptable, and the equipment fault is clear. , and All values ​​are 1. If the value is 0, then If the calculated value is 1, then the line only needs maintenance.

[0081] When there is a line fault, but the equipment is not damaged, the repair time and cost are unacceptable, or the equipment fault is unclear. =1, , and All values ​​are 0. The calculated value is 0.6, indicating that the line fault is minor and only requires maintenance.

[0082] When there is no line fault, but the equipment is damaged, the repair time and cost are acceptable, and the equipment fault is clear. , , and All values ​​are 1. If the calculated value is 0.7, then the faulty equipment can be maintained.

[0083] When there is no line fault, but the equipment is damaged and the repair time and cost are unacceptable, and the equipment fault is clear. and The value of is 0. and All values ​​are 1. If the calculated value is 0.2, then the faulty equipment should be replaced.

[0084] When there is no fault in the line, but the equipment is damaged, the repair time and cost are unacceptable, or the equipment fault is unclear. , and The value of is 0. All values ​​are 1. If the calculated value is 0, the faulty equipment should be replaced.

[0085] When a line fails, equipment is damaged, repair time and costs are unacceptable, or the equipment fault is unclear... , and The value of is 0. and All values ​​are 1. If the calculated value is 0.3, the faulty equipment should be replaced.

[0086] Furthermore, the evaluation of the industrial control computer's (ICC) operational performance in step six involves regularly analyzing ICC operational data, such as weekly or monthly, to assess data acquisition accuracy, instruction execution success rate, and anomaly identification accuracy, identify process bottlenecks, and optimize ICC operation. This optimization is based on historical data to improve the anomaly detection model, reduce false alarm and false negative rates, and adjust the acquisition frequency, calculation database, and control logic according to actual operating scenarios to reduce system resource consumption and improve processing efficiency. When storing ICC operational data, the database categorizes the data into hot and cold data for layered processing. Hot data consists of frequently used data, such as data from the past three months, while cold data consists of infrequently used data, such as data from more than three months ago. Hot data is stored on a local SSD for fast querying, while cold data is compressed and uploaded to cloud object storage to meet compliance audit requirements.

[0087] Another objective of this invention is to provide an industrial control computer terminal data processing system, comprising a data acquisition module, a data processing module, a self-test module, a calculation module, a control command module, an equipment health management module, an equipment maintenance module, a communication module, a data storage module, and a visualization interaction module. The data acquisition module is used to collect data from the industrial control computer in real time. The data processing module processes the data collected by the data acquisition module and performs data formatting and unified conversion. The self-test module is used for self-testing the operating performance of the industrial control computer. The calculation module stores calculation formulas and runs these formulas to calculate and process the data. The control command module generates and verifies commands, and performs real-time verification during command execution. The system includes real-time tracking, deviation threshold triggering secondary judgment and dynamic instruction adjustment, equipment health management module for periodic equipment self-inspection and analysis of abnormal faults, equipment maintenance module for handling abnormal faults and formulating maintenance and replacement plans based on on-site inspection, communication module for providing signal transmission channels for industrial control computer operations, data storage module for hierarchical data storage, and visualization interaction module for real-time display of equipment operating status and support for remote consultation interaction. The system establishes connections between the data acquisition module, data processing module, self-inspection module, calculation module, control instruction module, equipment health management module, equipment maintenance module, communication module, data storage module, and visualization interaction module.

[0088] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A data processing method for an industrial control computer terminal, characterized in that: The specific steps include the following: Step 1: Start the industrial computer, collect data, and perform a self-test. Step 2: Process the data collected by the industrial control computer and perform real-time calculation and analysis. The data is hierarchically distributed so that real-time control data directly enters the instruction generation link, status monitoring data is synchronously uploaded to the cloud platform for long-term trend analysis, and abnormal data is marked separately and stored in a high-priority queue for in-depth analysis. Step 3: Generate and output control commands based on the processed and analyzed data; Step 4: The industrial control computer performs periodic equipment self-checks and compares the results with historical equipment data. Any equipment that shows abnormalities during the self-check is reported. Periodic equipment self-checks are tiered based on equipment model, operating parameters, and historical failure rate: minute-level self-checks, hour-level self-checks, and daily-level self-checks. Minute-level self-checks test the connectivity of core acquisition ports and communication links; hourly-level self-checks test key operating parameters of the equipment; and daily-level self-checks perform comprehensive hardware health checks, acquisition accuracy calibration, and bad block scanning of storage media. Step 5: Issue alarms for abnormal events detected and analyzed by the industrial control computer, and establish a preset equipment maintenance solution library. Handle abnormal events according to the preset solutions in the library, categorizing them into known and unknown faults. For unknown faults, recommend similar fault handling solutions based on a knowledge graph, simultaneously triggering a remote expert consultation process, inputting the diagnostic results, and establishing an equipment replacement and maintenance calculation formula based on the diagnostic results as constraints. The fault handling method is confirmed through this replacement and maintenance calculation formula, which is as follows: in, and These are the weighting coefficients. and The values ​​are 0.3, 0.3, 0.2, and 0.2 respectively. To determine if there is a line fault, To determine if the equipment is damaged, Is the repair time and cost acceptable? To determine whether the equipment malfunction is clear, and The value can be 1 or 0, where 1 represents yes and 0 represents no. If the calculated value is greater than or equal to 0.5, then the faulty equipment should be repaired. If the calculated value is less than 0.5, the faulty device should be replaced. Step Six: Evaluate and optimize the operation of the industrial control computer, and establish a database to store the industrial control computer's operating data.

2. The industrial control computer terminal data processing method according to claim 1, characterized in that: The self-test of the industrial control computer in step one includes the following steps: S1.1: After the industrial control computer is powered on, it first loads the secure and trusted execution environment to verify the digital signatures of the core firmware, operating system image, and data acquisition driver; S1.2: Pre-allocate CPU cores, memory buffers, and disk I / O bandwidth according to the preset data acquisition task volume; S1.3: Perform hardware self-test, time synchronization verification, and acquisition channel connectivity test in sequence. Hardware self-test includes the health of acquisition port, communication module, and storage medium. If the self-test is abnormal, trigger a graded alarm. The graded alarm includes warning abnormality and fault abnormality. Warning abnormality is recorded locally. Warning abnormality includes problems found in hardware self-test that do not affect core functions. Fault abnormality prevents startup and reports to operation and maintenance. Fault abnormality includes core function failures found in hardware self-test, time synchronization verification error exceeding the allowable range, and acquisition channel connectivity test finding that multiple channels are completely interrupted. S1.4: Retrieve the list of currently running data collection tasks from the configuration center, and determine the data collection objects, collection frequency, and data verification database.

3. The industrial control computer terminal data processing method according to claim 1, characterized in that: The data processing and real-time computing in step two includes data cleaning, edge-side real-time analysis, and data hierarchical distribution. Data cleaning is to preprocess the original collected data, remove outliers, fill in missing values, and unify the data format. Edge-side real-time analysis is to complete low-latency computing of high-frequency data locally and adopt a streaming computing framework. The streaming computing framework supports multi-point correlation analysis.

4. The industrial control computer terminal data processing method according to claim 1, characterized in that: The generation and output of control commands in step three includes the following steps: S3.1: Based on the preset process database and safety constraints, the generated control commands are double-checked. The double check includes rule check and conflict check. Rule check is to check whether the command conforms to the safe operating range of the equipment. Conflict check is to check whether the command has logical conflicts in the multi-device collaborative scenario. If a logical conflict occurs, the adjustment command is regenerated. S3.2: Priority control commands are sent via a hard real-time channel, while ordinary adjustment commands use a reliable transmission protocol, supporting command retransmission and timeout retransmission; S3.3: After the instruction is issued, the device execution result is tracked in real time. If the execution fails or the deviation from the expected value exceeds the threshold, a secondary judgment is immediately triggered to regenerate the adjustment instruction.

5. The industrial control computer terminal data processing method according to claim 4, characterized in that: The comparison of periodic equipment self-test data with historical equipment data in step four involves comparing the self-test data with the equipment baseline model to identify potential anomalies. These potential anomalies are categorized into Level 1, Level 2, and Level 3 anomalies. Level 1 anomalies are automatically calibrated locally and recorded in the operation log. Level 2 anomalies are reported to the operation and maintenance platform, triggering an early warning. Level 3 anomalies immediately trigger a safety interlock, switch redundant equipment, synchronously report fault details, and retain all timestamps associated with the anomaly self-test data and the complete equipment operating context to support subsequent fault tracing.

6. The industrial control computer terminal data processing method according to claim 1, characterized in that: In step five, the alarm levels for abnormal events are divided into warning level and fault level. Warning level events notify maintenance personnel via SMS and the background records pending work orders. Fault level events notify the responsible person via voice call and simultaneously trigger the linkage plan. Based on the characteristics of the abnormal event, the system automatically matches the handling solution from the maintenance plan library. For known faults, it directly pushes standardized handling steps, required spare parts, and historical handling cases, supporting one-click automated handling. The diagnostic situation includes whether there is a line fault, whether the equipment is damaged, whether the equipment fault is clear, and whether the repair time and cost are acceptable. Based on the diagnostic situation, the system determines the equipment fault handling method, which includes repair and replacement, to help maintenance personnel confirm the equipment replacement and maintenance plan. After the abnormal event handling is completed, the handling process, time consumption, and effect are recorded, and the adaptability score of the plan library is updated to continuously optimize the accuracy of plan matching.

7. The industrial control computer terminal data processing method according to claim 1, characterized in that: The industrial control computer (ICC) performance evaluation in step six involves periodically analyzing ICC operating data to assess data acquisition accuracy, instruction execution success rate, and anomaly identification accuracy, identifying process bottlenecks. ICC operation optimization is based on historical data to optimize the anomaly detection model, reducing false alarm and false negative rates. It also involves adjusting the acquisition frequency, calculation database, and control logic according to the actual operating scenario to reduce system resource consumption. When storing ICC operating data, the database divides the data into hot and cold data for layered processing. Hot data is stored on local SSDs for fast querying, while cold data is compressed and uploaded to cloud object storage.

8. An industrial control computer terminal data processing system, implementing the industrial control computer terminal data processing method according to any one of claims 1-7, characterized in that: include: The data acquisition module is used to collect data from the industrial control computer in real time. The data processing module is used to process the data collected by the data acquisition module and perform data formatting and uniform conversion. Self-test module, the self-test module is used for self-testing the operating performance of the industrial control computer; A calculation module, which stores calculation formulas and runs the calculation formulas to calculate and process data; A control instruction module is used to generate and verify instructions, perform real-time tracking during instruction execution, and trigger secondary judgment and dynamic adjustment of instructions based on deviation thresholds. The equipment health management module is used for periodic self-inspection of equipment and analysis and judgment of abnormal faults; The equipment maintenance module is used to handle abnormal faults and formulate maintenance and replacement plans based on the on-site inspection. The communication module provides a signal transmission channel for the operation of the industrial control computer. Data storage module, wherein the data storage module is used for hierarchical data storage; A visualization interaction module is used to display the device's operating status in real time and supports remote consultation interaction.

Images