A distributed control system attack intrusion detection device and method

By designing data processing, database and attack intrusion detection modules in a decentralized control system, combining confidence rule base and linear interpolation method, efficient detection of multiple attack types is achieved, and low intrusion detection efficiency and poor accuracy in the prior art are solved. The problem has significantly improved the reliability of attack identification and early warning.

CN113721569BActive Publication Date: 2024-11-01SHANGHAI UNIVERSITY OF ELECTRIC POWER
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202110980608.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-08-25
Publication Date
2024-11-01
Estimated Expiration
2041-08-25

AI Technical Summary

Technical Problem

The existing decentralized control system has low intrusion detection efficiency, high intrusion rates and false alarm rates, making it difficult to effectively detect and early warning of various attack intrusions, especially in environments with real-time and resource-constrained.

Method used

A decentralized control system attack intrusion detection device is designed, including a data processing module, a database module and an attack intrusion detection module. By obtaining communication link data from a network router, extracting attack sign characteristic data, and using the set attack intrusion confidence rules. The library analyzes it and combines linear interpolation to calculate the confidence of attack intrusion, so as to realize the detection of ARP, DOS, SSH password blasting, man-in-the-middle and maliciously modifying set value attacks.

Benefits of technology

It improves the accuracy and efficiency of attack intrusion detection, reduces the missed and false alarm rates, ensures accurate identification and timely warning of different attack types, and enhances the security of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113721569B_ABST
    Figure CN113721569B_ABST
Patent Text Reader

Abstract

The present invention relates to a distributed control system attack intrusion detection device and method thereof, the system comprises a data processing module connected to a distributed control system network router, the data processing module is connected to a database module, the database module is bidirectionally connected to the attack intrusion detection module, the data processing module obtains communication link data from the distributed control system network router, and extracts attack symptom feature data from the communication link data; the attack intrusion detection module analyzes the attack symptom feature data according to a set attack intrusion confidence rule base to obtain an attack detection result; the database module receives and stores the attack symptom feature data and the attack detection result. Compared with the prior art, the present invention can detect and warn attack intrusions in real time and reliably by establishing an attack intrusion confidence rule base that is mapped to different attack intrusion types, thereby ensuring detection efficiency and accuracy, thereby effectively improving the safety of the operation of the distributed control system.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of industrial control system security monitoring, and in particular to a distributed control system attack intrusion detection device and method thereof. Background Art

[0002] Distributed control system is a new generation of instrument control system based on microprocessor, adopting the design principle of decentralized control function, centralized display operation, and taking into account the design principles of division and autonomy and comprehensive coordination. Distributed control system is referred to as DCS (Distributed Control System). It adopts the basic design concept of decentralized control, centralized operation and management, and adopts a multi-level hierarchical, cooperative and autonomous structure. Its main features are its centralized management and decentralized control. At present, DCS has been widely used in various industries such as power, metallurgy, petrochemical, etc.

[0003] As distributed control systems gradually use open and universal protocols, the network security threats and risks faced by distributed control systems are also increasing. Since distributed control systems directly control the operation of on-site equipment, once attacked and invaded, it will not only cause huge economic losses, but also may endanger personal safety and the environment. Intrusion detection technology is one of the most important security precautions in distributed control systems. It can detect known and unknown attacks and improve the ability to identify attacks and warn. However, due to the real-time requirements of distributed control systems and limited equipment resources, the existing intrusion detection efficiency is low, and its missed alarm rate and false alarm rate are high. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a distributed control system attack intrusion detection device and method thereof, so as to improve the detection efficiency and accuracy of attack intrusion.

[0005] The object of the present invention can be achieved by the following technical solutions: a distributed control system attack intrusion detection device, comprising a data processing module connected to a distributed control system network router, the data processing module is connected to a database module, the database module is bidirectionally connected to the attack intrusion detection module, the data processing module obtains communication link data from the distributed control system network router, and extracts attack sign feature data from the communication link data;

[0006] The attack intrusion detection module is used to analyze the attack sign feature data according to the set attack intrusion confidence rule base to obtain the attack detection result;

[0007] The database module is used to receive and store the attack symptom feature data output by the data processing module, and to receive and store the attack detection result output by the attack intrusion detection module.

[0008] Furthermore, the data processing module includes a Python packet capture submodule and a data analysis submodule connected in sequence. The Python packet capture submodule is used to obtain communication link data from a distributed control system network router, and extract attack sign feature data from the communication link data in conjunction with the data analysis module.

[0009] Furthermore, the attack intrusion detection module includes an ARP attack detection unit, a DOS attack detection unit, an SSH password brute force attack detection unit, a man-in-the-middle attack detection unit and a malicious modification of set value attack detection unit, so as to perform corresponding attack intrusion detection according to the set attack intrusion confidence rule base and obtain corresponding attack detection results.

[0010] Furthermore, the database module is specifically a Mysql database.

[0011] Furthermore, the attack and intrusion detection module is connected to a display module, and the display module is used to display the attack detection result to the user.

[0012] Furthermore, the display module is provided with a QT5 display interface.

[0013] A distributed control system attack intrusion detection method comprises the following steps:

[0014] S1. According to the set data monitoring time period, the data processing module obtains the communication link data from the distributed control system network routing, and extracts the attack sign feature data from the acquired communication link data;

[0015] S2, the data processing module transmits the attack sign characteristic data to the database module for storage;

[0016] S3. According to the set detection start time period, the attack intrusion detection module obtains the attack symptom feature data from the database module, and analyzes the attack symptom feature data according to the set attack intrusion confidence rule base to obtain the corresponding attack detection result.

[0017] Furthermore, the attack intrusion confidence rule library is provided with detection rules, confidence levels, semantic thresholds and limits corresponding to five attack intrusion types, wherein the five attack intrusion types include ARP attacks, DOS attacks, SSH password blasting attacks, man-in-the-middle attacks and malicious modification of set value attacks.

[0018] Furthermore, the step S3 specifically includes the following steps:

[0019] S31, according to the set detection start time period, the attack intrusion detection module obtains the attack sign feature data from the database module;

[0020] S32, the attack intrusion detection module compares the attack sign feature data with the thresholds and limits of the five attack intrusion types respectively, and uses a linear interpolation method to obtain the current evidence confidence corresponding to the five attack intrusion types;

[0021] S33, combining the current evidence confidence, the set corresponding premise confidence and the set corresponding conclusion confidence, calculate the corresponding establishment confidence of the five attack intrusion types;

[0022] S34, judging whether the confidence level is within the set corresponding threshold range, if it is judged to be yes, it indicates that such attack intrusion has occurred currently, otherwise it indicates that such attack intrusion has not occurred currently.

[0023] Furthermore, the calculation formula for establishing the confidence in step S33 is:

[0024] a=[1-max{0,δ1-δ′1}]×[1-max{0,δ2-δ′2}]×…×[1-max{0,δ N -δ′ N}]×β

[0025] Among them, a is the confidence of a certain attack intrusion type, δ′1 is the current evidence confidence value of the first symptom feature involved in the corresponding attack intrusion detection rule, δ1 is the premise confidence value given by the first symptom feature involved in the corresponding attack intrusion detection rule, and so on, δ′ N is the current evidence confidence value of the Nth running data involved in the corresponding attack intrusion detection rule, δ N is the confidence value of the premise condition given by the Nth running data involved in the corresponding attack intrusion detection rule, N is the total number of symptom features involved in the corresponding attack intrusion detection rule, and β is the confidence value of the conclusion in the corresponding attack intrusion detection rule.

[0026] Compared with the prior art, the present invention has the following advantages:

[0027] 1. The present invention is directed to a distributed control system. By setting a data processing module connected to a network router of the distributed control system, connecting the data processing module to a database module, and bidirectionally connecting the database module to an attack intrusion detection module, the data processing module is used to obtain communication link data from the network router of the distributed control system, and attack symptom feature data is extracted from the communication link data; the attack intrusion detection module is used to analyze the attack symptom feature data according to a set attack intrusion confidence rule library to obtain an attack detection result; thereby, the attack intrusion can be detected in real time, reliably and efficiently.

[0028] 2. The present invention fully considers different attack intrusion types, and based on the attack intrusion characteristics, establishes a one-to-one mapping relationship between the attack intrusion type and the attack intrusion characteristics, so that the attack intrusion confidence rule library is pre-set in the attack intrusion detection module, and according to the requirements of the industrial control network, the attack intrusion is detected regularly according to the set network attack intrusion confidence rule library, and finally the type of attack intrusion is obtained, thereby ensuring the accuracy of detection and reducing the missed alarm rate and false alarm rate.

[0029] 3. The present invention sets detection rules, confidence levels, semantic thresholds and limits corresponding to different attack and intrusion types in the attack and intrusion confidence rule library. In the subsequent specific detection process, the current evidence confidence level is obtained by comparing the thresholds and limits. The confidence level of the premise condition and the confidence level of the conclusion in the detection rules are then combined to calculate the corresponding confidence level of the attack and intrusion type. Finally, it is determined whether the confidence level is within the corresponding threshold range to obtain the detection result of whether the attack and intrusion have occurred. This method can further ensure the accuracy of the detection results and greatly improve the reliability of subsequent early warning guidance. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a schematic diagram of the structure of the device of the present invention;

[0031] Figure 2 It is a schematic diagram of the method flow of the present invention;

[0032] Figure 3 It is a schematic diagram of the working principle of the embodiment;

[0033] Explanation of the marks in the figure: 1. Data processing module, 2. Database module, 3. Attack and intrusion detection module, 4. Display module, 11. Python packet capture submodule, 12. Data analysis submodule, 31. ARP attack detection unit, 32. DOS attack detection unit, 33. SSH password blasting attack detection unit, 34. Man-in-the-middle attack detection unit, 35. Malicious modification of set value attack detection unit. DETAILED DESCRIPTION

[0034] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0035] Example

[0036] like Figure 1 As shown, a distributed control system attack intrusion detection device includes a data processing module 1 connected to a distributed control system network router, the data processing module 1 is connected to a database module 2, the database module 2 is bidirectionally connected to an attack intrusion detection module 3, the data processing module 1 obtains communication link data from the distributed control system network router, and extracts attack sign feature data from the communication link data;

[0037] The attack intrusion detection module 3 is used to analyze the attack symptom feature data according to the set attack intrusion confidence rule base to obtain the attack detection result;

[0038] The database module 2 is used to receive and store the attack sign characteristic data output by the data processing module 1, and receive and store the attack detection results output by the attack intrusion detection module 3;

[0039] In actual applications, the attack and intrusion detection module 3 is connected to a display module 4, and the display module 4 is used to display the attack detection result to the user.

[0040] The data processing module 1 includes a Python packet capture submodule 11 and a data analysis submodule 12 which are connected in sequence. The Python packet capture submodule 11 is used to obtain communication link data from a distributed control system network router, and extract attack sign feature data from the communication link data in conjunction with the data analysis module 12;

[0041] The attack intrusion detection module 3 includes an ARP attack detection unit 31, a DOS attack detection unit 32, an SSH password blasting attack detection unit 33, a man-in-the-middle attack detection unit 34 and a malicious modification of set value attack detection unit 35, so as to perform corresponding attack intrusion detection according to the set attack intrusion confidence rule base and obtain corresponding attack detection results.

[0042] The above device is used to implement a distributed control system attack intrusion detection method, such as Figure 2 As shown, the following steps are included:

[0043] S1. According to the set data monitoring time period, the data processing module obtains the communication link data from the distributed control system network routing, and extracts the attack sign feature data from the acquired communication link data;

[0044] S2, the data processing module transmits the attack sign characteristic data to the database module for storage;

[0045] S3. According to the set detection start time period, the attack intrusion detection module obtains the attack symptom feature data from the database module, and analyzes the attack symptom feature data according to the set attack intrusion confidence rule library to obtain the corresponding attack detection results, wherein the attack intrusion confidence rule library is provided with detection rules, confidence levels, semantic thresholds and limits corresponding to five types of attack intrusions. The five types of attack intrusions include ARP attacks, DOS attacks, SSH password blasting attacks, man-in-the-middle attacks and malicious modification of set value attacks.

[0046] Step S3 mainly includes the following processes:

[0047] S31, according to the set detection start time period, the attack intrusion detection module obtains the attack sign feature data from the database module;

[0048] S32, the attack intrusion detection module compares the attack sign feature data with the thresholds and limits of the five attack intrusion types respectively, and uses a linear interpolation method to obtain the current evidence confidence corresponding to the five attack intrusion types;

[0049] S33. Combining the current evidence confidence, the set corresponding premise confidence and the set corresponding conclusion confidence, the corresponding establishment confidence of the five attack intrusion types is calculated:

[0050] a=[1-max{0,δ1-δ′1}]×[1-max{0,δ2-δ′2}]×…×[1-max{0,δ N -δ′ N}]×β

[0051] Among them, a is the confidence of a certain attack intrusion type, δ′1 is the current evidence confidence value of the first symptom feature involved in the corresponding attack intrusion detection rule, δ1 is the premise confidence value given by the first symptom feature involved in the corresponding attack intrusion detection rule, and so on, δ′ N is the current evidence confidence value of the Nth running data involved in the corresponding attack intrusion detection rule, δ N is the confidence value of the premise condition given by the Nth running data involved in the corresponding attack intrusion detection rule, N is the total number of symptom features involved in the corresponding attack intrusion detection rule, and β is the confidence value of the conclusion in the corresponding attack intrusion detection rule;

[0052] S34, judging whether the confidence level is within the set corresponding threshold range, if it is judged to be yes, it indicates that such attack intrusion has occurred currently, otherwise it indicates that such attack intrusion has not occurred currently.

[0053] This embodiment applies the above technical solution, such as Figure 3 As shown, Windows 10 operating system is used as the platform, Python is used as the detection program development tool, and Mysql database is used as the real-time database to store data information. The detection rules (including rules, thresholds and limits) are written into the Python detection program, and the python packet capture module is used to calculate the data features on the communication link, and the data features are stored in the database in real time. The attack intrusion detection program is started periodically (1s). If the sign features appear, the corresponding detection rules are directly entered for judgment, and the detection results are stored in the database in real time. If the detection results are established, the detection results are displayed in the detection interface (QT5 display interface).

[0054] The specific process of this embodiment is as follows:

[0055] 1. Periodically (1s) start real-time monitoring of the distributed control system, monitor the data sent and received on its communication link, extract information from the data, analyze the symptom characteristics of the data on the communication link, and periodically (1s) start the attack intrusion detection rule base to execute step 2;

[0056] 2. Perform attack and intrusion detection based on the set attack and intrusion confidence rule library. The five attack and intrusion types are in the following order: ARP attack, DOS attack, SSH password blasting attack, malicious tampering of setting value attack, and man-in-the-middle attack.

[0057] The attack and intrusion confidence rule base contains detection rules, confidence levels, semantic thresholds, and limits corresponding to five types of attack and intrusion:

[0058] (1) ARP attack detection method (detection period is 1s):

[0059] i) ARP attack detection rule: If the number of packets detected that do not match the system ARP table match counts mismatch (counts mismatch =[address ARP ≠address detected ]) The confidence level of the increase is 0.8, and the number of ARP table changes in the system is detected to be n ARPchange The confidence level is 0.8, so the confidence level of ARP attack is 0.95;

[0060] ii) Semantic thresholds and limits:

[0061] (a) The number of packets detected that do not match the system ARP table counts mismatch Increase: Threshold: 1, High limit: 3;

[0062] (b) The number of times the ARP table in the system has been modified under illegal conditions is detected. ARPchange : Threshold: 0 times, High limit: 1 time;

[0063] (2) DOS attack detection method (detection cycle is 1s):

[0064] i)DOS attack detection rules:

[0065] If the data transmission rate V of the network card is detected netcard The increased confidence level is 0.8, and the data transmission rate V of the entire network net Increase the confidence level to 0.8 and the DPU load rate P loadThe increased confidence is 0.7, so the confidence of a DOS attack is 0.95;

[0066] ii) Semantic thresholds and limits:

[0067] (a) Detect the data transmission rate V of the network card netcard Increase: Threshold: 1000 byte / s, High limit: 10000 byte / s;

[0068] (b) Network data transmission rate V net Increase: Threshold: 10000 byte / s, High limit: 50000 byte / s;

[0069] (c) DPU load factor P load Increase: Threshold: 10%, High Limit: 90%;

[0070] (3) SSH password brute force attack detection method (detection period is 1s):

[0071] i) SSH password blasting attack detection rules:

[0072] If the number of failed login attempts through the login port is n fail The increased confidence is 0.7, and the number of login port usages n in every five seconds use The increased confidence is 0.8, so the confidence of SSH password brute force attack is 0.95;

[0073] ii) Semantic thresholds and restrictions:

[0074] (a) Number of failed login attempts through the login port n fail Increase: Threshold: 3 times, High limit: 5 times;

[0075] (b) Number of login attempts per five seconds n use Increase: Threshold: 10 times, High limit: 30 times;

[0076] (4) Man-in-the-middle attack detection (detection cycle is 1s):

[0077] i) Man-in-the-middle attack detection method:

[0078] If the number of ARP attacks detected is N ARPattack The increased confidence is 0.7, and the number of times n different ARP attack types are detected ARPtype The increased confidence is 0.8, and the number of ARP table changes detected in the system is n ARPchange The confidence level is 0.8, so the confidence level of a man-in-the-middle attack is 0.95;

[0079] ii) Semantic Thresholds and Limits

[0080] (a) Number of ARP attacks detected N ARPattack Increase: Threshold: 1 time, High limit: 3 times;

[0081] (b) The number of times n different ARP attack types are detected ARPtype Increase: Threshold: 2 times, High limit: 10 times;

[0082] (c) The number of times n ARP table changes are detected in the system ARPchange : Threshold: 0 times, High limit: 1 time;

[0083] (5) Detection method for malicious modification of setting value attacks:

[0084] i) Malicious modification of setting value attack detection rules:

[0085] If the set value is modified n times per minute SETchange The increased confidence level is 0.7, and the system operates under stable conditions. The set value is modified n times Conditionchange The increased confidence level is 0.9 and the number of times the setpoint is modified outside the safe operating setpoint interval is n danger The confidence level of the malicious modification of the setting value attack is 1.

[0086] ii) Semantic thresholds and limits:

[0087] a)Number of times the set value is modified within 1 minute n SETchange Increase: Threshold: 1 time, High limit: 5 times;

[0088] b) The number of times the set value is modified when the system is running under stable conditions n Conditionchange Increase: Threshold: 1 time; High limit: 5 times;

[0089] c) The number of times the set value is modified outside the safe operation set value range n danger : Threshold: 1 time, High limit: 2 times.

[0090] According to the above attack intrusion confidence rule base, when performing attack intrusion detection, first monitor the real-time data on the distributed control system network in real time, compare the abnormal features found in the monitoring data with the threshold and limit of the operation features involved in the corresponding attack intrusion detection rules, set the confidence corresponding to the limit to 1, and the confidence corresponding to the threshold to 0.5, and use the linear interpolation method to obtain the confidence δ′ of the attack intrusion features involved in the detection rule;

[0091] Then compare the corresponding attack intrusion rule base to obtain the confidence level of the attack intrusion:

[0092] a=[1-max{0,δ1-δ′1}]×[1-max{0,δ2-δ′2}]×…×[1-max{0,δ N -δ′ N}]×β

[0093] Among them, δ′1 is the current evidence confidence value of the first symptom feature involved in the corresponding attack intrusion detection rule obtained above, δ1 is the premise confidence value given by the first symptom feature involved in the attack intrusion detection rule, and so on, δ′ N To obtain the current evidence confidence value of the Nth running data involved in the corresponding attack intrusion detection rule, δ N is the confidence value of the premise condition given by the Nth running data involved in the attack intrusion detection rule, N is the total number of symptom features involved in the corresponding attack intrusion detection rule, and β is the confidence value of the conclusion in the corresponding attack intrusion detection rule;

[0094] Finally, it is determined whether the confidence a of the attack intrusion satisfies the confidence range corresponding to the attack intrusion. If so, the attack intrusion detection result is established.

[0095] For example, when performing intrusion detection on a DOS attack, the data transmission rate V of the network card in the communication link is monitored. netcard , the data transmission flow rate of the network V net , and the DPU load factor P load The three real-time data features are compared with the thresholds and limits given by the three data in the DOS attack intrusion detection rules to obtain δ′1, δ′2 and δ′3 respectively, where δ′1 is the current evidence confidence value of the transmission flow rate operation data of the specified network card, δ′2 is the current evidence confidence value of the transmission flow rate operation data of the network, and δ′3 is the current evidence confidence value of the DPU load rate operation data; then the confidence of the DOS attack detection is obtained. When calculating, δ1 is the confidence value of the prerequisite condition of the network card transmission data flow rate involved in the DOS attack detection rule, which is 0.8, and so on. δ2 is 0.8, δ3 is 0.7, and β is the conclusion confidence value in the DOS attack detection rule, which is 0.95. The confidence a of the DOS attack detection is obtained by the above method, and it is judged whether a meets the set confidence range corresponding to the establishment of the DOS attack detection. If it meets the confidence range, a DOS attack warning prompt is issued in the system.

Claims

1. A distributed control system attack intrusion detection method, applied to a distributed control system attack intrusion detection device, characterized in that: The device comprises a data processing module (1) connected to a distributed control system network router, the data processing module (1) being connected to a database module (2), the database module (2) being bidirectionally connected to an attack intrusion detection module (3), the data processing module (1) acquiring communication link data from the distributed control system network router, and extracting attack sign feature data from the communication link data; The attack intrusion detection module (3) is used to analyze the attack sign feature data according to the set attack intrusion confidence rule base to obtain the attack detection result; The database module (2) is used to receive and store the attack sign characteristic data output by the data processing module (1), and to receive and store the attack detection results output by the attack intrusion detection module (3); The method comprises the following steps: S1. According to the set data monitoring time period, the data processing module obtains the communication link data from the distributed control system network routing, and extracts the attack sign feature data from the acquired communication link data; S2, the data processing module transmits the attack sign characteristic data to the database module for storage; S3. According to the set detection start time period, the attack intrusion detection module obtains the attack symptom feature data from the database module, and analyzes the attack symptom feature data according to the set attack intrusion confidence rule base to obtain the corresponding attack detection result; Step S3 specifically includes the following steps: S31, according to the set detection start time period, the attack intrusion detection module obtains the attack sign feature data from the database module; S32, the attack intrusion detection module compares the attack sign feature data with the thresholds and limits of the five attack intrusion types respectively, and uses a linear interpolation method to obtain the current evidence confidence corresponding to the five attack intrusion types; S33, combining the current evidence confidence, the set corresponding premise confidence and the set corresponding conclusion confidence, calculate the corresponding establishment confidence of the five attack intrusion types; S34, judging whether the confidence level is within the set corresponding threshold range, if it is judged to be yes, it indicates that such attack intrusion has occurred currently, otherwise it indicates that such attack intrusion has not occurred currently; In step S32, attack intrusion detection is performed according to the set attack intrusion confidence rule base. The five attack intrusion types are in order: ARP attack, DOS attack, SSH password blasting attack, malicious tampering setting value attack, and man-in-the-middle attack. The attack intrusion confidence rule base contains detection rules, confidence levels, semantic thresholds, and limits corresponding to the five attack intrusion types: (1) ARP attack detection method, detection period is 1s: i) ARP attack detection rule: If the number of packets detected that do not match the system ARP table match counts mismatch (counts mismatch =[address ARP ≠address detected ]) The confidence level of the increase is 0.8, and the number of ARP table changes in the system is detected to be n ARPchange The confidence level is 0.8, so the confidence level of ARP attack is 0.95; ii) Semantic thresholds and limits: (a) The number of packets detected that do not match the system ARP table counts mismatch Increase: Threshold: 1, High limit: 3; (b) The number of times the ARP table in the system has been modified under illegal conditions is detected. ARPchange : Threshold: 0 times, High limit: 1 time; (2) DOS attack detection method, detection period is 1s: i)DOS attack detection rules: If the data transmission rate V of the network card is detected netcard The increased confidence level is 0.8, and the data transmission rate V of the entire network net Increase the confidence level to 0.8 and the DPU load rate P load The increased confidence is 0.7, so the confidence of a DOS attack is 0.95; ii) Semantic thresholds and limits: (a) Detect the data transmission rate V of the network card netcard Increase: Threshold: 1000 byte / s, High limit: 10000 byte / s; (b) Network data transmission rate V net Increase: Threshold: 10000 byte / s, High limit: 50000 byte / s; (c) DPU load factor P load Increase: Threshold: 10%, High Limit: 90%; (3) SSH password brute force attack detection method (detection period is 1s): i) SSH password blasting attack detection rules: If the number of failed login attempts through the login port is n fail The increased confidence is 0.7, and the number of login port usages n in every five seconds use The increased confidence is 0.8, so the confidence of SSH password brute force attack is 0.95; ii) Semantic thresholds and restrictions: (a) Number of failed login attempts through the login port n fail Increase: Threshold: 3 times, High limit: 5 times; (b) Number of login attempts per five seconds n use Increase: Threshold: 10 times, High limit: 30 times; (4) Man-in-the-middle attack detection, detection cycle is 1s: i) Man-in-the-middle attack detection method: If the number of ARP attacks detected is N ARPattack The increased confidence is 0.7, and the number of times n different ARP attack types are detected ARPtype The increased confidence is 0.8, and the number of ARP table changes detected in the system is n ARPchange The confidence level is 0.8, so the confidence level of a man-in-the-middle attack is 0.95; ii) Semantic Thresholds and Limits (a) Number of ARP attacks detected N ARPattack Increase: Threshold: 1 time, High limit: 3 times; (b) The number of times n different ARP attack types are detected ARPtype Increase: Threshold: 2 times, High limit: 10 times; (c) The number of times n ARP table changes are detected in the system ARPchange : Threshold: 0 times, High limit: 1 time; (5) Detection method for malicious modification of setting value attacks: i) Malicious modification of setting value attack detection rules: If the set value is modified n times per minute SETchange The increased confidence level is 0.7, and the system operates under stable conditions. The set value is modified n times Conditionchange The increased confidence level is 0.9 and the number of times the setpoint is modified outside the safe operating setpoint interval is n danger The confidence level of the malicious modification of the setting value attack is 1. ii) Semantic thresholds and limits: a)Number of times the set value is modified within 1 minute n sETchange Increase: Threshold: 1 time, High limit: 5 times; b) The number of times the set value is modified when the system is running under stable conditions n Conditionchange Increase: Threshold: 1 time; High limit: 5 times; c) The number of times the set value is modified outside the safe operation set value range n danger : Threshold: 1 time, High limit: 2 times.

2. A distributed control system attack intrusion detection method according to claim 1, characterized in that: The data processing module (1) comprises a Python packet capture submodule (11) and a data analysis submodule (12) which are connected in sequence, wherein the Python packet capture submodule (11) is used to obtain communication link data from a distributed control system network router, and in combination with the data analysis module, extract attack sign feature data from the communication link data.

3. A distributed control system attack intrusion detection method according to claim 1, characterized in that: The attack intrusion detection module (3) comprises an ARP attack detection unit (31), a DOS attack detection unit (32), an SSH password blasting attack detection unit (33), a man-in-the-middle attack detection unit (34) and a malicious setting value modification attack detection unit (35), so as to respectively perform corresponding attack intrusion detection according to a set attack intrusion confidence rule base and obtain corresponding attack detection results.

4. A distributed control system attack intrusion detection method according to any one of claims 1 to 3, characterized in that: The database module (2) is specifically a Mysql database.

5. A distributed control system attack intrusion detection method according to claim 1, characterized in that: The attack and intrusion detection module (3) is connected to a display module (4), and the display module (4) is used to display the attack detection result to the user.

6. A distributed control system attack intrusion detection method according to claim 5, characterized in that: The display module (4) is provided with a QT5 display interface.

7. A distributed control system attack intrusion detection method according to claim 1, characterized in that: The calculation formula for establishing the confidence in step S33 is: a=[1-max{0,δ1-δ′1}]×[1-max{0,δ2-δ2 ′ }]×…×[1-max{0,δ N -d′ N }]×b Among them, a is the confidence of a certain attack intrusion type, δ′1 is the current evidence confidence value of the first symptom feature involved in the corresponding attack intrusion detection rule, δ1 is the premise confidence value given by the first symptom feature involved in the corresponding attack intrusion detection rule, and so on, δ′ N is the current evidence confidence value of the Nth running data involved in the corresponding attack intrusion detection rule, δ N is the confidence value of the premise condition given by the Nth running data involved in the corresponding attack intrusion detection rule, N is the total number of symptom features involved in the corresponding attack intrusion detection rule, and β is the confidence value of the conclusion in the corresponding attack intrusion detection rule.