Industrial computer multi-protocol adaptive control system based on edge computing

Abstract

The invention discloses an industrial computer multi-protocol adaptive control system based on edge computing, and relates to the technical field of industrial automation, the industrial computer multi-protocol adaptive control system comprises a data acquisition module, a protocol identification module, a dynamic conversion strategy module, a network anomaly evaluation module, a protocol misjudgment detection module and a data storage module, according to the invention, the protocol identification module carries out analysis based on message real-time characteristics and non-message level, quickly responds to new equipment and message changes, and carries out protocol misjudgment detection through the protocol identification module based on the non-message level analysis, so that the reliability of the system is improved, and the reliability of the system is improved. A processing strategy is dynamically adjusted through a dynamic conversion strategy module according to the load and the delay, and the edge side completes adaptive real-time performance; and a misjudgment detection module is combined with communication state evaluation to perform misjudgment detection on unqualified protocols, so that misjudgment is reduced, the protocol identification accuracy is improved, and stable and efficient operation of the system is guaranteed.

Description

Technical Field

[0001] This invention relates to the field of industrial automation technology, specifically to an industrial computer multi-protocol adaptive control system based on edge computing. Background Technology

[0002] Edge computing refers to an open platform that integrates core capabilities of networking, computing, storage, and applications at the network edge, close to the source of objects or data. It provides edge intelligence services to meet the key needs of industry digitalization in terms of agile connectivity, real-time business, application intelligence, security, and privacy protection. Industrial computers are adaptable to industrial environments such as high temperature resistance, dust resistance, and vibration resistance, as well as have characteristics such as high reliability, long lifespan, and strong scalability. They can operate stably in complex industrial sites. As industrial production scenarios become increasingly complex and specialized, computer systems need to support multiple communication protocols commonly used in the industrial field and be compatible with different manufacturers and different types of industrial equipment.

[0003] Existing computers are typically based on fixed rules and cannot dynamically adjust conversion strategies according to the real-time status of the equipment. Furthermore, edge nodes have low processing resources, which can easily lead to equipment overload and crashes or data loss. In addition, traditional methods that rely on single-feature recognition, such as function code matching, cannot distinguish between protocols with similar message structures. This can easily lead to misjudgments of similar features but incorrect protocols, resulting in subsequent conversion failures. The inability to detect protocol misjudgments in a timely manner reduces the quality of multi-protocol control in industrial computers. Summary of the Invention

[0004] The purpose of this invention is to provide an industrial computer multi-protocol adaptive control system based on edge computing, which solves the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an industrial computer multi-protocol adaptive control system based on edge computing, comprising a data acquisition module, a protocol identification module, a dynamic conversion strategy module, a network anomaly assessment module, a protocol misjudgment detection module, and a data storage module;

[0006] The data acquisition module is used to collect communication status data and operating status data of industrial equipment, and store them through the data storage module;

[0007] The protocol identification module obtains historical identification information of similar devices from the data storage module to obtain candidate protocols, and evaluates the candidate protocols in conjunction with the communication status data to determine whether the candidate protocol evaluation is qualified. If it is qualified, the dynamic conversion strategy module is used; if it is unqualified, the protocol misjudgment detection module is used.

[0008] When a candidate protocol passes the evaluation, the dynamic conversion strategy module dynamically adjusts the operation strategy of the industrial equipment based on communication status data and operation status data.

[0009] The network anomaly assessment module assesses network anomalies in industrial equipment based on protocol confidence assessment results and operational status data.

[0010] When a candidate protocol fails the evaluation, the protocol misjudgment detection module combines communication status data and candidate protocol compatibility to determine whether the protocol identification module has misjudged the protocol and takes appropriate measures.

[0011] Optionally, after the industrial equipment is connected, the protocol identification module first obtains historical identification information of similar equipment from the data storage module to obtain candidate protocols, and evaluates the confidence level of the candidate protocols in conjunction with the communication status data. The specific process is as follows:

[0012]

[0013] In the above formula, CPR represents the protocol identification confidence level;

[0014] F mat The message feature matching score reflects the degree of matching between the current message features and the candidate protocol.

[0015] W1 is the message characteristic influence coefficient, with a value of 0.5;

[0016] T cor The number of times the same type of device was correctly identified on the edge side;

[0017] T tot This represents the total number of times the same type of device was identified on the edge side.

[0018] W2 is the historical equipment impact coefficient, with a value of 0.3;

[0019] This represents the historical accuracy rate for identifying similar devices, reflecting the historical protocol identification accuracy rate of devices of the same model.

[0020] P mat The communication process matching degree reflects the default matching status of industrial equipment ports and handshake processes with candidate protocols;

[0021] W3 is the communication process impact coefficient, with a value of 0.2;

[0022] According to industry standards and system testing, the confidence threshold is set at 0.7. When the protocol recognition confidence CPR ≥ 0.7, it is considered qualified. At this time, the candidate protocol is confirmed as the actual protocol and put into use.

[0023] When the protocol identification confidence level (CPR) is less than 0.7, it is considered unqualified and will be detected by the protocol misjudgment detection module.

[0024] Optionally, the dynamic conversion strategy module dynamically adjusts the industrial equipment operation strategy based on communication status data and operating status data. First, it extracts communication delay information from the communication status data and obtains equipment load information from the operating status data. The communication delay information and equipment load information are then weighted and fused to obtain a conversion strategy score AF, which reflects the current operating status of the industrial equipment. Based on the conversion strategy score AF result, a corresponding conversion mode is adopted. The process is as follows:

[0025] When the conversion strategy score AF ≥ 0.8, the high-efficiency conversion mode is enabled;

[0026] When the conversion strategy score AF is between 0.5 and 0.8, the balanced conversion mode is enabled.

[0027] When the conversion strategy score AF < 0.5, the conservative conversion mode is enabled;

[0028] Each conversion mode has different conversion complexity, sampling frequency, and resource allocation strategies. The dynamic conversion strategy module enables dynamic adjustment of the operation strategy of industrial equipment.

[0029] Optionally, the network anomaly assessment module first extracts packet loss rate information and communication delay information from the operating status data, then compares the packet loss rate information and communication delay information with the threshold set by the system, and introduces corresponding weight coefficients to obtain packet loss rate impact items and communication delay impact items, and combines the packet loss rate impact items and communication delay impact items to assess network anomalies.

[0030] Protocol identification confidence level (CPR) is introduced as an impact factor for protocol identification. The network anomaly assessment results are adjusted according to the CPR value to obtain the network anomaly score (CS).

[0031] Optionally, based on system testing, the first anomaly threshold for the network anomaly score CS is set to 0.6, and the second anomaly threshold is set to 0.9. Corresponding measures are taken based on the different values ​​of the network anomaly score CS, as follows:

[0032] When the network anomaly score CS ≤ 0.6, it indicates that the network is normal, and the protocol conversion process continues;

[0033] When 0.6 < Network Anomaly Score (CS) < 0.9, it indicates a minor anomaly. Continue the protocol conversion process and record the anomaly log.

[0034] When the network anomaly score CS≥0.9, the protocol conversion process is paused, an alarm is triggered to the operation and maintenance platform, and the operation and maintenance personnel are notified to conduct a manual inspection.

[0035] Optionally, when the protocol identification confidence level CPR < 0.7, it indicates that the candidate protocol is unqualified, and the protocol misjudgment detection module is used to detect whether there is a misjudgment.

[0036] The protocol misjudgment detection module includes a feature supplementation unit and a compatibility detection unit. The feature supplementation unit first supplements the message features not used by the protocol identification module to obtain the new message feature matching degree, and then compares it with the message feature matching degree F. mat Weighted fusion yields the complete message feature matching degree NF. mat .

[0037] Optionally, the compatibility detection unit detects the compatibility of candidate protocols by first simulating communication to detect the communication quality of candidate protocols and obtain communication compatibility verification items.

[0038] Next, the correctness of data conversion under the candidate protocol is checked to obtain the data format matching verification item. The communication compatibility verification item and the data format matching verification item are combined to obtain the candidate protocol compatibility verification score BS.

[0039] When the false positive detection score MS of a candidate protocol is greater than or equal to 0.7, it means that the protocol false positive detection is successful, the candidate protocol can meet the usage requirements, and can be put into normal use.

[0040] When the candidate protocol misjudgment detection score MS < 0.7, it indicates that the protocol misjudgment detection has failed and cannot meet the usage requirements. In this case, manual processing should be notified.

[0041] Optionally, the data storage module includes a database, which includes a standard protocol library, device protocol fingerprints, and historical protocol identification records.

[0042] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0043] I. The protocol identification module of this invention is based on real-time features such as the start character, function code, and field length of the message. It can determine the matching degree between the current message and the candidate protocol in real time, and can respond quickly to new devices or sudden changes in messages. It provides an immediate basis for protocol identification, avoids the lag caused by relying on historical data, and analyzes the matching degree of communication process based on the communication port and handshake process of the device. It can cover protocol identification codes that are not at the message level, solve the problem of misjudgment caused by similar message features in traditional systems, and enhance the ability to identify protocols without clear message features.

[0044] Second, the dynamic conversion strategy module of this invention dynamically adjusts the data sampling frequency, protocol conversion algorithm, and data compression strategy based on the real-time load of the equipment and communication latency. When the equipment is in good condition, it improves data processing efficiency and accuracy, and prioritizes equipment stability when the condition is poor. The network anomaly assessment module evaluates the network anomaly of industrial equipment in real time to avoid data transmission errors. It does not require frequent manual adjustments and can be completed at the edge to adapt to the dynamic changes of industrial equipment and the resource constraints of edge computing. It meets the real-time requirements of industrial scenarios, realizes multi-protocol adaptive control of industrial equipment, and ensures the stable and efficient operation of industrial systems.

[0045] Third, this invention uses a protocol misjudgment detection module to evaluate communication status data and candidate protocol compatibility to determine whether the protocol identification module has misjudged the protocol. It also avoids candidate protocols that match features but are actually invalid. Furthermore, it only triggers when the protocol identification module fails to identify the protocol, supplementing the feature dimensions not covered by the protocol misjudgment detection module, reducing the burden on edge nodes, improving the accuracy of protocol identification, and avoiding device loss of control, data loss, or resource waste caused by misjudgment. Attached Figure Description

[0046] Figure 1 This is a flowchart of the system workflow of the present invention;

[0047] Figure 2 This is a block diagram of the dynamic conversion strategy module of the present invention;

[0048] Figure 3 This is a block diagram of the false positive detection module of the present invention. Detailed Implementation

[0049] 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.

[0050] For examples, please refer to Figures 1 to 3 This implementation provides an industrial computer multi-protocol adaptive control system based on edge computing, including a data acquisition module, a protocol identification module, a dynamic conversion strategy module, a network anomaly assessment module, a protocol misjudgment detection module, and a data storage module;

[0051] The data acquisition module is used to collect communication status data and operating status data of industrial equipment, and store them through the data storage module. The data storage module includes a database, which includes a standard protocol library, device protocol fingerprints, and historical protocol identification records. The standard protocol library includes message formats, ports, and function codes of different protocols. The device protocol fingerprints include the protocols used by the historical devices. The historical protocol identification records include the historical device identification time and system processing data.

[0052] The protocol identification module obtains historical identification information of similar devices from the data storage module, obtains candidate protocols, and evaluates the candidate protocols in combination with communication status data to determine whether the candidate protocol evaluation is qualified. If it is qualified, the dynamic conversion strategy module is used; if it is unqualified, the protocol misjudgment detection module is used.

[0053] Furthermore, after the device is connected to the system, the protocol identification module obtains historical identification information of similar devices from the data storage module to obtain candidate protocols. These candidate protocols serve as a reference benchmark for the system to verify the actual protocol used by the currently connected device. The confidence level of the candidate protocols is evaluated in conjunction with communication status data. The specific process is as follows:

[0054]

[0055] In the above formula, CPR is the confidence level of protocol identification; the higher the value, the more reliable the identification.

[0056] F mat The message feature matching score, derived from real-time features such as the message start character, function code, and field length, determines the degree of matching between the current message and candidate protocols. A score closer to 1 indicates a higher degree of matching, reflecting the current message status of the device. It enables rapid response to new devices or sudden changes in messages, providing immediate judgment for protocol identification and avoiding the lag caused by relying on historical data. This solves the problem of traditional systems being unable to capture dynamic changes in messages, such as when message features are fine-tuned after a device firmware upgrade. mat It can quickly detect and adjust the protocol identification confidence level (CPR).

[0057] W1 is the message feature influence coefficient, with a value of 0.5. Since message features are the core basis for protocol identification, they are set to the highest weight.

[0058] T cor The number of times the same type of device was correctly identified on the edge side;

[0059] T tot This represents the total number of times the same type of device was identified on the edge side.

[0060] W2 is the historical device influence coefficient, which has relatively low importance for protocol identification, and its value is set to 0.2; The historical accuracy rate of identifying similar devices reflects the historical protocol identification accuracy rate of devices of the same model or in the same scenario. By using historical data, the current identification results can be optimized, reducing misjudgments caused by single abnormal messages.

[0061] P mat W3 is the communication process matching degree, and W3 is the communication process influence coefficient. It is obtained based on the analysis of the device's communication port and handshake process. It can cover protocol identification codes that are not at the message level. For example, the port number is a fixed feature of many industrial protocols. It solves the problem of misjudgment caused by similar message features in traditional systems and enhances the ability to identify protocols without clear message features. For example, some private protocols rely on handshake process confirmation. Therefore, the value is set to 0.3.

[0062] The protocol identification module comprehensively evaluates the reliability of protocol identification by integrating information from different dimensions of real-time message characteristics, historical data, and communication processes, thereby reducing false judgments and improving data accuracy. Based on industry standards and system testing, the confidence threshold is set at 0.7. When the protocol identification confidence CPR ≥ 0.7, it is considered qualified. At this time, the candidate protocol is confirmed as the actual protocol, and the actual protocol is applied to the matched device and put into use.

[0063] When a candidate protocol passes the evaluation, the dynamic conversion strategy module dynamically adjusts the operation strategy of the industrial equipment based on communication status data and operation status data.

[0064] Furthermore, since edge computing nodes are typically deployed on-site, their hardware resources are far lower than those of cloud servers. Using a fixed, complex protocol conversion strategy can easily lead to edge node overload and crashes. In industrial scenarios, equipment load fluctuates with production rhythms, and fixed strategies cannot adjust according to the load, resulting in resource waste or performance bottlenecks. This application addresses the challenges of high real-time requirements, limited equipment resources, and frequent network fluctuations in industrial scenarios by using a dynamic conversion strategy module to adjust data sampling frequency, protocol conversion algorithms, and data compression strategies. This ensures the stable and efficient operation of the industrial system. The process is as follows:

[0065]

[0066] In the above formula, AF is the conversion strategy score. The larger the value, the more the system needs to meet the real-time requirements of the business. The lower the value, the more it needs to prioritize controlling resource consumption and avoid equipment overload or network congestion.

[0067] D represents the current communication latency of the device;

[0068] D th The communication latency threshold is derived using industry standards.

[0069] γ1 is the communication delay impact coefficient, with a value of 0.5. The communication delay term directly reflects the real-time status of device communication. The lower the device communication delay is than the threshold, the larger this term becomes, and the system can adopt a more frequent message conversion strategy. When the delay exceeds the threshold, this term drops to 0, and the system needs to switch to a conservative strategy to avoid communication congestion. By setting the min function, the impact of extreme delays can be limited. Even if the device delay is much higher than the threshold, it will not cause abnormal formula output, thus ensuring stable system operation.

[0070] The communication delay impact coefficient γ1 can be adjusted according to the actual application scenario to improve applicability. For example, in scenarios with extremely high real-time requirements such as robot motion control, communication speed should be prioritized. In this case, the value can be adjusted to 0.7 or higher, but the communication delay impact coefficient γ1 + equipment load impact coefficient γ2 should be guaranteed to be 1.

[0071] L represents the current device load rate, which is obtained by weighted averaging of CPU and memory load rates, with CPU load rate accounting for 0.7% and memory load rate accounting for 0.3%.

[0072] L th The equipment load rate threshold;

[0073] γ2 is the equipment load influence coefficient, with a value of 0.5. The device load item takes into full account the limitations of the device's computing power. When the load is below the threshold, this item has a large value, indicating that the device has more idle resources, allowing the system to use complex conversion algorithms, such as multi-protocol parallel conversion. When the device load exceeds the threshold, this item drops to 0, switching to a lightweight strategy, such as single-protocol serial conversion, to avoid device overload and crash. This is suitable for industrial equipment with different performance levels. When the performance of older equipment is weak, the device load impact coefficient γ2 is increased to prioritize system stability. When the performance of newer equipment is strong, the device load impact coefficient γ2 is decreased to improve system operating efficiency. By setting the device load item, the performance of different devices can be fully utilized while ensuring stable system operation. The dynamic conversion strategy module combines communication latency and device load, and sets weights that can be adjusted according to the actual application scenario to achieve adaptive adjustment of the dynamic conversion strategy, ensuring system real-time performance and improving operating efficiency.

[0074] Based on the conversion strategy score AF, the complexity, sampling frequency, and resource allocation strategy of the protocol conversion are adjusted in real time to adapt to the dynamic state of industrial equipment and the resource constraints of edge computing. According to system testing, a threshold range for the conversion strategy score AF is set, and corresponding conversion modes are adopted according to different threshold ranges of the conversion strategy score AF. The process is as follows:

[0075] When the conversion strategy score AF≥0.8, the high-efficiency conversion mode is enabled, and a parallel protocol conversion algorithm is adopted; all fields of the device message are retained, the sampling frequency is increased, and multi-dimensional metadata such as timestamps and device health are added; 20% to 30% of the CPU cores of the edge nodes are allocated to the conversion task.

[0076] When the conversion strategy score AF is between 0.5 and 0.8, the balanced conversion mode is enabled. The serial conversion algorithm is used to process protocol messages according to priority. The sampling frequency remains unchanged. Only core business fields, such as sensor values ​​and control commands, are retained to simplify metadata. Resources are dynamically allocated: 10% to 20% of CPU cores are allocated to avoid overloading edge resources. This mode is the system default mode.

[0077] When the conversion strategy score AF < 0.5, the conservative conversion mode is enabled, and a lightweight conversion algorithm is adopted, such as fast matching based on pre-compiled rules. The sampling frequency is reduced, only core numerical fields such as temperature and pressure are extracted, unnecessary metadata is omitted, and ≤10% of CPU cores are allocated to prioritize the basic operation of edge nodes.

[0078] Because equipment status and network conditions are constantly fluctuating in industrial environments, fixed conversion strategies cannot adapt to dynamic changes. The dynamic conversion strategy module can dynamically adjust according to the operating status of different devices, improving data processing efficiency and accuracy when the device is in good condition, and prioritizing device stability when the condition is poor. This eliminates the need for frequent manual adjustments and can be completed at the edge, speeding up response time and meeting the real-time requirements of industrial scenarios, thus enabling adaptive control of industrial equipment using multiple protocols.

[0079] As the front end of industrial data processing, edge nodes dynamically change the operating strategies of equipment to avoid node downtime caused by resource overload, ensuring stable operation of edge computing and preventing equipment crashes or data loss caused by fixed strategies, thus ensuring continuous and stable operation of the system in a dynamic environment.

[0080] The network anomaly assessment module evaluates network anomalies in industrial equipment based on protocol confidence assessment results and operational status data.

[0081] Furthermore, the network anomaly assessment module evaluates network anomalies in industrial equipment based on protocol confidence assessment results and operational status data, as follows:

[0082]

[0083] In the above formula, CS is the network anomaly score, and the higher the value, the more severe the anomaly.

[0084] CPR is the confidence level for protocol identification.

[0085] β1 is the protocol reliability impact coefficient, with a value of 0.3. (1-CPR)×β1 is the protocol identification impact item. When the protocol identification confidence CPR is low, the value of this item increases, which will improve the network anomaly score CS and reduce data transmission errors caused by protocol identification errors.

[0086] P loss This represents the current network packet loss rate of the device, reflecting the stability of network data transmission.

[0087] P th The packet loss rate threshold is derived from industry standards.

[0088] β2 is the impact coefficient of packet loss rate, with a value of 0.3. This is a factor affecting packet loss rate. The larger the value, the more severe the packet loss of the device. By comparing the network packet loss rate of each device with the standard threshold, the packet loss rate of different devices can be transformed into a unified standardized indicator, eliminating the differences between devices and facilitating cross-device anomaly comparison.

[0089] D represents the current communication latency of the device;

[0090] D th The communication latency threshold is derived using industry standards.

[0091] β3 is the communication delay impact coefficient, with a value of 0.4. It quantifies the degree of abnormality in communication delay by comparing the device's communication delay with its threshold.

[0092] β1 is the protocol reliability impact coefficient β1 + packet loss rate impact coefficient β2 + communication delay impact coefficient β3 = 1;

[0093] The network anomaly assessment module evaluates network anomalies by incorporating Protocol Recognition Confidence (CPR) into network packet loss rate and communication latency. This allows it to detect situations where protocol recognition errors occur despite normal network metrics, preventing device control failures caused by incorrect conversions and improving the accuracy of network anomaly detection. Based on the system's testing of device performance at different network anomaly scores (CS), two anomaly thresholds for the CS are set: 0.6 for the first threshold and 0.9 for the second. Corresponding measures are taken based on the different CS scores, as follows:

[0094] When the network anomaly score CS ≤ 0.6, it indicates that the network is normal, and the protocol conversion process continues;

[0095] When 0.6 < Network Anomaly Score CS < 0.9, it indicates a minor anomaly. Continue the protocol conversion process, record the anomaly log, and trigger warnings, such as local prompts on edge nodes.

[0096] When the network anomaly score CS≥0.9, it indicates a serious anomaly, the protocol conversion process is suspended, and an alarm is pushed to the operation and maintenance platform.

[0097] Furthermore, when the protocol identification confidence level (CPR) is less than 0.7, it is considered unqualified. An evaluation is then conducted using a protocol misjudgment detection module, combining communication status data and candidate protocol compatibility, to determine if the protocol identification module has misjudged the protocol and to take corresponding measures. First, the missing feature dimensions in the original identification are supplemented. The process is as follows:

[0098]

[0099] In the above formula, NF mat To improve the feature matching degree of the completed message;

[0100] F mat The message feature matching degree is derived based on real-time features such as the message start character, function code, and field length.

[0101] F new The new message feature matching degree is derived based on the checksum matching degree and the address segment rule matching degree.

[0102] θ represents the historical feature weight, with a value of 0.6, used to balance the matching degree F between the new and original features and the completed message features. mat To mitigate the impact of this, the matching degree of the original features is adjusted by adding new message features, thereby reducing errors.

[0103] Meanwhile, since edge computing nodes are usually deployed on-site, their hardware resources are far lower than those of cloud servers. If all possible message features are calculated from the beginning, it will significantly increase the real-time computing burden. Therefore, a protocol misjudgment detection module is set up, and the protocol misjudgment detection process is only performed when the protocol identification confidence CPR < 0.7, which can save resource consumption.

[0104] Next, compatibility verification is performed on the candidate protocols. The compatibility verification process involves simulating communication and matching formats to eliminate invalid candidate protocols that match features but cannot communicate, thereby reducing false positives. The process is as follows:

[0105]

[0106] In the above formula, BS is the candidate protocol compatibility verification score. The higher the score, the stronger the compatibility of the candidate protocol.

[0107] N suc To simulate the number of successful communications, N tot To simulate the total number of communications;

[0108] W com The communication compatibility impact coefficient is set at 0.6, as simulated communication is the basis for verifying the compatibility of candidate protocols.

[0109] This is a communication compatibility verification item used to verify the actual communication capability between the candidate protocol and the device. In industrial scenarios, device communication reliability is a core requirement. If the candidate protocol only matches the features but cannot actually transmit data, the system will be completely unable to control the device or collect data. By simulating communication, such as sending test commands and reading device status, the availability of the candidate protocol can be directly verified. The stronger the compatibility, the higher the availability, and the higher the probability of unqualified candidate protocols being put into use.

[0110] M mat M represents the number of items that successfully matched the data format. tot The total number of items for data format validation;

[0111] W for Data format influence coefficient;

[0112] The data format matching verification item verifies the correctness of data conversion under the candidate protocol, ensuring that the transmitted data can be correctly parsed. Even if communication is successful, if the data format does not conform to the protocol specification, such as field length exceeding the range or check code error, subsequent data analysis or control commands will fail. By checking the format rules, the validity of the data at the protocol level is ensured, reducing errors in subsequent data processing. Finally, by adding the two items together, the impact of communication reliability and format correctness on the availability of the candidate protocol is balanced, making it suitable for most industrial scenarios.

[0113] The complete message feature matching degree NF is obtained. mat After obtaining the candidate protocol compatibility verification score BS, multiply it by the candidate protocol misclassification detection score MS = F. mat ×BS, the higher the score, the greater the probability of the candidate protocol being put into normal use; and based on the system test, the success threshold of the false detection score MS is set to 0.7. The system test can simulate the actual use effect of the candidate protocol and the corresponding industrial equipment under different false detection scores MS, so as to set the specific success threshold.

[0114] When the false positive detection score MS of a candidate protocol is greater than or equal to 0.7, it means that the protocol false positive detection is successful, the candidate protocol can meet the usage requirements, and can be put into normal use.

[0115] When the candidate protocol misjudgment detection score MS < 0.7, it indicates that the protocol misjudgment detection has failed and cannot meet the usage requirements. In this case, the operation and maintenance platform should be notified for manual processing.

[0116] 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. An industrial computer multi-protocol adaptive control system based on edge computing, characterized in that: It includes a data acquisition module, a protocol identification module, a dynamic conversion strategy module, a network anomaly assessment module, a protocol misjudgment detection module, and a data storage module; The data acquisition module is used to collect communication status data and operating status data of industrial equipment, and store them through the data storage module; The protocol identification module obtains historical identification information of similar devices from the data storage module to obtain candidate protocols, and evaluates the candidate protocols in conjunction with the communication status data to determine whether the candidate protocol evaluation is qualified. If it is qualified, the dynamic conversion strategy module is used; if it is unqualified, the protocol misjudgment detection module is used. When a candidate protocol passes the evaluation, the dynamic conversion strategy module dynamically adjusts the operation strategy of the industrial equipment based on communication status data and operation status data. The network anomaly assessment module assesses network anomalies in industrial equipment based on protocol confidence assessment results and operational status data. When a candidate protocol fails the evaluation, the protocol misjudgment detection module combines communication status data and candidate protocol compatibility to determine whether the protocol identification module has misjudged the protocol and takes appropriate measures.

2. The industrial computer multi-protocol adaptive control system based on edge computing according to claim 1, characterized in that: After the industrial equipment is connected, the protocol identification module first obtains historical identification information of similar equipment from the data storage module to obtain candidate protocols, and then evaluates the confidence level of the candidate protocols in conjunction with the communication status data. The specific process is as follows: ; In the above formula, CPR represents the protocol identification confidence level; F mat The message feature matching score reflects the degree of matching between the current message features and the candidate protocol. W1 is the message characteristic influence coefficient, with a value of 0.5; T cor The number of times the same type of device was correctly identified on the edge side; T tot This represents the total number of times the same type of device was identified on the edge side. W2 is the historical equipment impact coefficient, with a value of 0.3; This represents the historical accuracy rate for identifying similar devices, reflecting the historical protocol identification accuracy rate of devices of the same model. P mat The communication process matching degree reflects the default matching status of industrial equipment ports and handshake processes with candidate protocols; W3 is the communication process impact coefficient, with a value of 0.2; According to industry standards and system testing, the confidence threshold is set at 0.

7. When the protocol recognition confidence CPR ≥ 0.7, it is considered qualified. At this time, the candidate protocol is confirmed as the actual protocol and put into use. When the protocol identification confidence level (CPR) is less than 0.7, it is considered unqualified and will be detected by the protocol misjudgment detection module.

3. The industrial computer multi-protocol adaptive control system based on edge computing according to claim 2, characterized in that: The dynamic conversion strategy module dynamically adjusts the industrial equipment operation strategy based on communication status data and operational status data. First, it extracts communication latency information from the communication status data and obtains equipment load information from the operational status data. The communication latency information and equipment load information are then weighted and fused to obtain a conversion strategy score AF, which reflects the current operational status of the industrial equipment. Based on the conversion strategy score AF result, a corresponding conversion mode is adopted. The process is as follows: When the conversion strategy score AF ≥ 0.8, the high-efficiency conversion mode is enabled; When the conversion strategy score AF is between 0.5 and 0.8, the balanced conversion mode is enabled. When the conversion strategy score AF < 0.5, the conservative conversion mode is enabled; Each conversion mode has different conversion complexity, sampling frequency, and resource allocation strategies. The dynamic conversion strategy module enables dynamic adjustment of the operation strategy of industrial equipment.

4. The industrial computer multi-protocol adaptive control system based on edge computing according to claim 3, characterized in that: The network anomaly assessment module first extracts packet loss rate information and communication delay information from the operating status data. Then, it compares the packet loss rate information and communication delay information with the thresholds set by the system and introduces corresponding weight coefficients to obtain packet loss rate impact items and communication delay impact items. The network anomaly situation is assessed by combining the packet loss rate impact items and communication delay impact items. Protocol identification confidence level (CPR) is introduced as an impact factor for protocol identification. The network anomaly assessment results are adjusted according to the CPR value to obtain the network anomaly score (CS).

5. The industrial computer multi-protocol adaptive control system based on edge computing according to claim 4, characterized in that: Based on system testing, the first anomaly threshold for the network anomaly score (CS) is set to 0.6, and the second anomaly threshold is set to 0.

9. Corresponding measures are taken based on the different values ​​of the CS score, as follows: When the network anomaly score CS ≤ 0.6, it indicates that the network is normal, and the protocol conversion process continues; When 0.6 < Network Anomaly Score (CS) < 0.9, it indicates a minor anomaly. Continue the protocol conversion process and record the anomaly log. When the network anomaly score CS≥0.9, the protocol conversion process is paused, an alarm is triggered to the operation and maintenance platform, and the operation and maintenance personnel are notified to conduct a manual inspection.

6. The industrial computer multi-protocol adaptive control system based on edge computing according to claim 5, characterized in that: When the protocol identification confidence level (CPR) is less than 0.7, it indicates that the candidate protocol is unqualified. The protocol misjudgment detection module is used to detect whether there is a misjudgment. The protocol misjudgment detection module includes a feature supplementation unit and a compatibility detection unit. The feature supplementation unit first supplements the message features not used by the protocol identification module to obtain the new message feature matching degree, and then compares it with the message feature matching degree F. mat Weighted fusion yields the complete message feature matching degree NF. mat .

7. The industrial computer multi-protocol adaptive control system based on edge computing according to claim 6, characterized in that: The compatibility detection unit detects the compatibility of candidate protocols. First, it detects the communication quality of candidate protocols through simulated communication to obtain communication compatibility verification items. Next, the correctness of data conversion under the candidate protocol is checked to obtain the data format matching verification item. The communication compatibility verification item and the data format matching verification item are combined to obtain the candidate protocol compatibility verification score BS. When the false positive detection score MS of a candidate protocol is greater than or equal to 0.7, it means that the protocol false positive detection is successful, the candidate protocol can meet the usage requirements, and can be put into normal use. When the candidate protocol misjudgment detection score MS < 0.7, it indicates that the protocol misjudgment detection has failed and cannot meet the usage requirements. In this case, manual processing should be notified.

8. The industrial computer multi-protocol adaptive control system based on edge computing according to claim 1, characterized in that: The data storage module includes a database, which contains a standard protocol library, device protocol fingerprints, and historical protocol identification records.

Images