Industrial personal computer data security dynamic protection method and system based on AI edge calculation

Abstract

The invention belongs to the crossing field of industrial control and edge computing, and particularly relates to an industrial personal computer data security dynamic protection method and system based on AI edge computing. Firstly, industrial control data encryption and standardization preprocessing are achieved through a hardware-level trusted execution environment; secondly, performing real-time anomaly detection on the encrypted data stream based on an AI model, generating a threat score by combining industrial control protocol feature analysis, and forming an edge layer intelligent detection capability; the authentication mode is dynamically switched based on the threat score, and self-adaptive strategy adjustment is realized in a fog layer; then optimizing edge resource allocation according to the threat index and the equipment load, and constructing a cloud elastic scheduling mechanism; and finally, a security system covering data encryption, real-time detection, dynamic protection and resource optimization is formed through grading component cooperation, and the method is suitable for real-time threat detection and self-adaptive response in an industrial internet scene.

Description

Technical Field

[0001] The present invention belongs to the cross - field of industrial control and edge computing, and particularly relates to a method and system for dynamically protecting the data security of industrial control computers based on AI edge computing. Background Technique

[0002] Currently, industrial control systems are facing increasingly complex security threats, and traditional protection technologies have shown obvious limitations. In terms of edge - computing security authentication, although existing technologies have proposed multi - level dynamic authentication strategies, which match different security levels for authentication through instruction types, their core relies on a static rule - base for decision - making, and they cannot effectively identify and respond to new types of unknown attack patterns such as zero - day vulnerabilities and advanced persistent threats (APTs). At the same time, such solutions do not fully consider the stringent real - time requirements of industrial control scenarios, and the authentication delay often exceeds the tolerable threshold of industrial control systems. In the field of threat assessment and resource allocation, although existing methods have introduced a dynamic priority adjustment mechanism based on real - time threat indices and resource limitations, their threat assessment models lack self - learning capabilities, and key parameters rely on manual presetting, making it difficult to adapt to the complex environment of heterogeneous industrial field devices, diverse protocols, and variable working conditions. In terms of intrusion detection technology, although there are methods to detect intrusion behaviors through traditional methods such as signal energy analysis, they fail to integrate advanced AI algorithms such as deep learning, and the detection accuracy for industrial - control - specific attacks with strong concealment and rapid mutation is insufficient;

[0003] Upon in - depth analysis, the key problems existing in the existing industrial control security system are as follows: First, the defense strategy is static and rigid, the update of authentication rules and detection feature libraries lags behind, and it cannot adapt to newly emerging threats in real - time; Second, the resource allocation mechanism lacks flexibility. In high - concurrency industrial control scenarios, it may either cause critical alarm delays due to insufficient resources or result in energy waste due to over - allocation of resources; Third, the security decision - making lacks context - awareness capabilities, and it does not comprehensively judge by combining industrial - control - specific operating states such as PLC instruction cycles and sensor data - flow timing characteristics, resulting in a high false - alarm rate; Fourth, the intelligence level on the edge side is low, and most solutions still use traditional rule - engines or shallow machine - learning algorithms, making it difficult to deeply mine complex features such as industrial - control protocol payloads and device - to - device interaction behaviors. These defects are particularly prominent in the context of the deep integration of the industrial Internet;

[0004] The above patents all have the problems proposed in this background technique. To solve the problems proposed in this background technique, this application designs a method and system for dynamically protecting the data security of industrial control computers based on AI edge computing. Summary of the Invention

[0005] Aiming at the deficiencies of the existing technology, the present invention proposes a method and system for dynamically protecting the data security of industrial control computers based on AI edge computing.

[0006] To achieve the above object, the present invention provides the following technical solutions: In the first aspect, the present invention provides an industrial control computer data security dynamic protection method based on AI edge computing, which includes the following specific steps:

[0007] S1. Trusted data encryption and preprocessing: Encrypt the original industrial control instructions and sensor data through a hardware-level trusted execution environment, and preprocess the original data to form a standardized encrypted data stream;

[0008] S2. Intelligent anomaly detection and protocol analysis: Based on the AI model, real-time detect the abnormal data packets in the standardized encrypted data stream, combine with the characteristics of the industrial control protocol for context verification, and generate a threat score in real-time;

[0009] S3. Dynamic security policy adjustment: Automatically switch the authentication mode according to the threat score level, and execute the dynamic security adjustment policy;

[0010] S4. Elastic resource allocation and optimization: Dynamically allocate edge computing resources according to the real-time threat index and device load;

[0011] S5. Realize the dynamic protection of industrial control computer data security through the four-step closed-loop of encryption-detection-adjustment-optimization;

[0012] Preferably, the method for encrypting the original industrial control instructions and sensor data through the hardware-level trusted execution environment is specifically as follows:

[0013] Build a chip-level security isolation area based on the national cryptographic algorithm, bind the device identity with the PUF physical fingerprint, dynamically generate a key through quantum random numbers, encrypt the industrial control instructions, and use check protection for the sensor data. Finally, combine blockchain evidence storage and anti-side channel for fusion design.

[0014] Preferably, the method for preprocessing the original data to form a standardized encrypted data stream is specifically as follows:

[0015] Realize data standardization processing through a multi-level protocol parsing engine and a timing feature alignment algorithm, and embed the working condition context label generated by the digital twin.

[0016] Preferably, the method for real-time detecting the abnormal data packets in the standardized encrypted data stream based on the AI model and performing context verification in combination with the characteristics of the industrial control protocol is specifically as follows:

[0017] Complete the decryption of the quantum key in the TEE security environment, retain the time stamp and the PUF hardware fingerprint, and then through the spatio-temporal parallel analysis architecture, perform multi-dimensional feature fusion in combination with the working condition context provided by the digital twin, and finally output the threat score based on the dynamic weight model.

[0018] Preferably, the automatic switching of the authentication mode according to the threat score level is specifically implemented as follows:

[0019] Based on the real-time threat scoring model, high and low threat levels are divided, and the authentication mode switching is automatically triggered by the policy engine. Dynamic token authentication is used in the case of low threats; quantum-enhanced key exchange and iris biometric multi-factor authentication are started in the case of high threats.

[0020] Preferably, the implementation of the dynamic security adjustment policy is specifically as follows:

[0021] Based on the real-time threat score, logs are stored in a hierarchical manner. When the threat is low, they are recorded in the local encrypted cache. When the threat is high, they are immediately written to the quantum-secure cloud storage.

[0022] Preferably, the dynamic allocation of edge computing resources according to the real-time threat index and device load is specifically implemented as follows:

[0023] The edge computing resources are dynamically allocated according to the formula:

[0024]

[0025] where T priority is the threat priority, C util is the current CPU utilization rate, and α, β are adaptive weight factors;

[0026] The edge server collects the device load data and threat scores in real time, and then the federated learning engine dynamically adjusts the α and β coefficients. Finally, the scheduler executes the allocation policy according to the calculation results.

[0027] In a second aspect, the present invention provides an industrial control computer data security dynamic protection system based on AI edge computing, which is implemented based on the above-mentioned industrial control computer data security dynamic protection method based on AI edge computing. Specifically, it includes: a data acquisition and processing module for obtaining the operation data of the industrial control computer through sensors;

[0028] An AI detection module for performing multi-modal fusion analysis and risk assessment on the obtained operation data of the industrial control computer through AI to construct an edge layer component;

[0029] A policy regulation module for performing policy regulation on the industrial control computer system according to the risk assessment of the operation of the industrial control computer to construct a fog layer component;

[0030] A resource optimization module for realizing the efficient allocation of the computing resources of the industrial control system through a dynamic algorithm to construct a cloud component.

[0031] In a third aspect, the present invention provides an electronic device, including: a processor and a memory, wherein a computer program that can be called by the processor is stored in the memory;

[0032] The processor executes the above-mentioned industrial control computer data security dynamic protection method based on AI edge computing by calling the computer program stored in the memory.

[0033] In a fourth aspect, the present invention provides a computer-readable storage medium storing instructions, which when run on a computer, cause the computer to execute the above-mentioned industrial control computer data security dynamic protection method based on AI edge computing.

[0034] The beneficial effects of the technical solution of the present invention are as follows:

[0035] Through four major innovative designs, a breakthrough improvement in the security protection of industrial control systems is achieved. First, an instruction cycle model of industrial control protocols such as TCP is constructed based on a timing feature extractor to achieve microsecond-level abnormal behavior detection; second, a multi-objective optimization algorithm that combines threat index, device load, and network latency is developed to achieve dynamic resource scheduling at the edge-fog layer, significantly improving the utilization rate of computing resources; third, a dynamic authentication decision tree based on AI threat scoring is designed to greatly improve the protection level in high-threat scenarios; finally, a lightweight AI model compression technology is adopted to improve the overall energy efficiency ratio of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The drawings described herein are used to provide a further understanding of the present disclosure and constitute a part of the present disclosure. The illustrative embodiments and descriptions thereof are used to explain the present disclosure and do not constitute an improper limitation of the present disclosure. In the drawings:

[0037] Figure 1 is a schematic diagram of the overall framework of the industrial control computer data security dynamic protection method based on AI edge computing of the present invention;

[0038] Figure 2 is a schematic diagram of the overall process of the industrial control computer data security dynamic protection method based on AI edge computing of the present invention;

[0039] Figure 3 is a schematic diagram of the overall framework of the industrial control computer data security dynamic protection system based on AI edge computing of the present invention; DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] To clearly elaborate on the core content of the technical solution, the embodiments of the present application are systematically described below in conjunction with the drawings of the specification. It should be particularly noted that the embodiments shown below are only typical implementation manners of the present application, and there may be various optimization and deformation schemes for its specific technical details, but they all fall within the scope of the patent protection of the present application.

[0041] Embodiment 1

[0042] To solve the technical problems proposed in the background art, the present invention provides a preferred embodiment: asFigure 1 - Figure 2 As shown in the figure, a dynamic protection method for industrial control computer data based on AI edge computing includes the following specific steps:

[0043] S1. Trusted data encryption and preprocessing: Encrypt the original industrial control instructions and sensor data through a hardware-level trusted execution environment, and preprocess the original data to form a standardized encrypted data stream;

[0044] S2. Intelligent anomaly detection and protocol analysis: Based on the AI model, real-time detect abnormal data packets in the standardized encrypted data stream, and perform context verification in combination with industrial control protocol characteristics to generate threat scores in real time;

[0045] S3. Dynamic security policy adjustment: Automatically switch the authentication mode according to the threat score level and execute the dynamic security adjustment policy;

[0046] S4. Elastic resource allocation and optimization: Dynamically allocate edge computing resources according to the real-time threat index and device load, and further construct cloud components;

[0047] S5. Realize the dynamic protection of industrial control computer data through the four-step closed loop of encryption-detection-adjustment-optimization;

[0048] Encrypt the original industrial control instructions and sensor data through a hardware-level trusted execution environment. The specific implementation method is as follows:

[0049] The hardware-level trusted execution environment (TEE) realizes industrial control data encryption through triple technologies: First, build a chip-level security isolation area based on the national cryptographic SM2 algorithm, and use PUF physical fingerprints to bind the device identity;

[0050] It should be noted that: First, extract the unique fingerprint of the device through the PUF module of the Huawei HiSec 5.0 security chip, generate a 256-bit SM3 hash identity ID, and establish an SM2 algorithm authentication channel to complete the two-way authentication between the PLC and the edge node;

[0051] Secondly, dynamically generate keys through quantum random numbers, encrypt industrial control instructions in the SM4-CTR mode, and protect sensor data with AES-256-GCM with verification;

[0052] It should be noted that: Use the SV-QRNG200 quantum random number chip to generate keys in real time. The industrial control instruction stream is encrypted in the SM4-CTR mode, and the sensor data is protected by AES-256-GCM and the key is rotated every 10 minutes using the Kyber-768 algorithm; Finally, combine blockchain evidence storage and anti-side channel design. The specific implementation process is as follows: Write the SM3 hash value of the encrypted log into the Ant Chain, and at the same time resist side channel attacks through 7nm process electromagnetic shielding and dynamic voltage regulation technology (DVFS).

[0053] Preprocess the original data, and its specific implementation method is as follows:

[0054] Implement data standardization processing through a multi-level protocol parsing engine and a timing feature alignment algorithm. Among them, protocol parsing is implemented by a hardware-accelerated FPGA parsing module to achieve microsecond-level disassembling of TCP protocol fields. At the same time, nanosecond-level timing synchronization is achieved through the gPTP protocol of the time-sensitive network, and the working condition context tags generated by digital twins are embedded;

[0055] It should be noted that: First, based on the hardware parsing engine of the Xilinx Versal AI Edge FPGA chip, use parallel pipeline technology to perform microsecond-level field disassembling on industrial control protocols such as Modbus TCP and OPC UA to accurately extract key elements such as function codes and register addresses; Second, through the IEEE 802.1AS-2025 time-sensitive network protocol, combined with the hardware timestamps of the Huawei HI1812 chipset and the Beidou-3 timing system, complete nanosecond-level timing synchronization across devices; Finally, relying on the Unity Industrial digital twin platform, fuse real-time working condition parameters such as PLC operating status and equipment vibration spectra, and intelligently generate three-dimensional semantic tags including protocol original fields, production scenario context, and security threat scores.

[0056] Based on the AI model, detect abnormal data packets in real time and perform context verification in combination with industrial control protocol features. The specific implementation method is as follows:

[0057] First, complete quantum key decryption in the TEE secure environment, and retain the μs-level timestamp and PUF hardware fingerprint; Subsequently, through a spatio-temporal parallel analysis architecture, perform multi-dimensional feature fusion in combination with the working condition context provided by the digital twin; Finally, output a threat score of 0-1 based on the dynamic weight model;

[0058] It should be noted that: First, within the TEE secure environment, complete data decryption based on the quantum key distribution system of the national cryptography SM9 algorithm, and synchronously retain the μs-level timestamp generated by the Beidou-3 timing system and the PUF physical fingerprint; Subsequently, through the spatio-temporal parallel analysis architecture built by the Xilinx Versal AI Edge chip, fuse the decrypted data stream with the working condition parameters provided by the digital twin system in real time, and use the 3D-CNN model to extract spatial features and the Temporal-Transformer model to capture timing patterns; Finally, input the dynamic weight evaluation model and output a continuous score of 0-1 in combination with the threat intelligence library.

[0059] Automatically switch the authentication mode according to the threat level. The specific implementation method is as follows:

[0060] Based on a real-time threat 0-1 continuous value scoring model, high and low threat levels are divided, and the authentication mode is automatically switched by the policy engine. When the threat is low, dynamic token authentication based on the time synchronization type OTP-2025 algorithm is used, and when the threat is high, quantum-enhanced key exchange + iris biometric multi-factor authentication is initiated;

[0061] It should be noted that: First, the industrial control data stream is analyzed in real time based on the CNN-LSTM hybrid model, and threat scores are generated in combination with the digital twin working condition context; when detecting low-threat devices with a score < 0.7 points, OTP-2025 dynamic token authentication is enabled, and when detecting high-threat devices with a score ≥ 0.7 points, quantum-enhanced multi-factor authentication is initiated, that is, first the Huawei QKD-3000 completes anti-quantum key exchange, and then biometric verification is performed through the iris recognition module. Finally, the operation authority needs to be manually confirmed. All authentication records are uploaded to the chain in real time, and the scoring threshold is optimized daily through federated learning.

[0062] Execute the dynamic security adjustment policy, and its specific implementation method is:

[0063] Based on the real-time threat 0-1 scoring, logs are automatically classified and stored - logs of low-threat devices with a score < 0.7 points are recorded in the local encrypted cache, and logs of high-threat devices with a score ≥ 0.7 points are immediately written to the quantum security cloud storage. Each log embeds the Beidou-3 μs-level timestamp and the device PUF fingerprint, and multi-node time synchronization is achieved through the TSN network;

[0064] It should be noted that: Through the real-time threat scoring model, the industrial control data stream is dynamically evaluated, and a two-level storage policy is implemented according to the threat level. For low-threat devices with a score < 0.7 points, the logs are encrypted by SM4 and stored in the local NVMe cache. For high-threat devices with a score ≥ 0.7 points, the logs trigger the quantum security protocol, are encrypted through the "Mozi" quantum channel and written to the Alibaba Cloud quantum storage, and an Ant Chain certificate is generated synchronously. All logs embed four-dimensional metadata of timestamp, PUF fingerprint, working condition label, and SM3 feature code, and 30ns-level time synchronization is achieved through the Huawei HI1812 chipset.

[0065] According to the real-time load and threat index, edge computing resources are dynamically allocated, and its specific implementation method is:

[0066] Edge computing resources are dynamically allocated according to the formula:

[0067] Where T priority is the threat priority, C util is the current CPU utilization rate, and α, β are adaptive weight factors;

[0068] The edge server collects device load data and threat scores in real time. Subsequently, the federated learning engine dynamically adjusts the α and β coefficients every 15 minutes. Finally, the KubeEdge 5.3 scheduler executes a hierarchical policy based on the calculation results, allocating 80% of the CPU to the AI detection task for devices with a formula output value > 0.7, and releasing resources to the industrial control application process for devices with a value < 0.7. Further feedback adjustment is performed to automatically update the α and β coefficients according to the network latency index;

[0069] It should be noted that: First, the Huawei Atlas 500 edge server collects device load data in real time and the CNN-LSTM model generates a threat score of 0-1. It is synchronized to the KubeEdge 5.3 control plane through the TSN network and attached with a Beidou timestamp. The federated learning engine aggregates data every 15 minutes and dynamically adjusts the α coefficient and β coefficient. KubeEdge is based on Execute hierarchical scheduling: When the formula output value ≥ 0.7, allocate 80% of the CPU to the AI detection accelerated by FPGA. When the formula output value < 0.7, release 50% of the resources to the industrial control process. Finally, the PID controller optimizes the parameters every 30 minutes and uploads the evidence to the blockchain.

[0070] Embodiment 2

[0071] As Figure 3 shown, the industrial control computer data security dynamic protection system based on AI edge computing is implemented based on the above-mentioned industrial control computer data security dynamic protection method for AI edge computing, and specifically includes:

[0072] The data acquisition and processing module is used to obtain the operation data of the industrial control computer through sensors;

[0073] The AI detection module is used to perform multimodal fusion analysis and risk assessment on the obtained operation data of the industrial control computer through AI, and construct edge layer components:

[0074] Build a hardware core by embedding an AI security chip in the industrial control computer, collect device data in real time through a multi-protocol data acquisition interface, and combine a lightweight convolutional neural network (CNN) and a long short-term memory network (LSTM) fusion model to analyze the local data stream in real time; Deploy a lightweight anomaly detection model and a protocol whitelist rule library in the software layer, and combine the national cryptography SM4 / SM9 dual authentication controller and the local threat knowledge base to form a "collection - parsing - detection - response" closed loop;

[0075] It should be noted that: the hardware layer uses Ascend 910B AI chip and multi-protocol interface to realize μs-level data collection and analysis; the analysis layer deploys a lightweight CNN-LSTM fusion model, extracts spatial features through 3D-CNN, and analyzes timing rules through Temporal-LSTM, to achieve 200,000 points / second real-time reasoning; the security layer runs the protocol whitelist, national secret SM4 / SM9 dual authentication and local threat knowledge base in parallel; the response layer triggers graded disposal based on the dynamic 0-1 score interval score, and high-risk threat devices with a score ≥0.7 are immediately isolated and FPGA deep forensics is started; the optimization layer updates the model every 6 hours through federated learning, combined with digital twin adversarial samples for continuous evolution.

[0076] The policy control module is used to control the policy of the IPC system according to the risk assessment of IPC operation and build fog layer components:

[0077] By deploying edge servers to build fog computing nodes and integrating multiple edge gateways, the federated learning framework is used to aggregate the threat scores of each industrial computer in real time. The computing resources are dynamically allocated through the intelligent scheduling engine. Quantum key rotation and traffic mirror analysis are initiated for high-threat devices with a score ≥ 0.7, and lightweight rule verification is performed for low-threat devices. Through the multi-objective optimization algorithm module, collaborative threat assessment and resource scheduling are implemented.

[0078] It should be noted that a federated learning network is built through Huawei Atlas 500Pro fog computing nodes deployed at the edge layer. Each gateway uses lightweight real-time to generate a 0-1 device-level threat score, and aggregates encrypted threat features every 5 minutes through the FATE framework to form a global threat map. Computing resources are dynamically scheduled based on the NSGA-III multi-objective optimization algorithm - Huawei QKD-3000 quantum key rotation and FPGA-accelerated traffic mirror analysis are immediately started for high-threat devices, and lightweight rule verification is performed on low-threat devices. All operation records are stored through Ant Chain BaaS. The federated learning system optimizes 12 performance indicators every hour to achieve millisecond-level coordination of threat response and resource scheduling.

[0079] The resource optimization module is used to achieve efficient allocation of computing resources in industrial control systems through dynamic algorithms and build cloud components:

[0080] By deploying a distributed global threat intelligence center to connect to the vulnerability feature database, linking cluster training hybrid models, and combining the national SM9 encrypted policy distribution channel, a three-in-one cloud defense system is built;

[0081] It should be noted that: First, deploy distributed threat intelligence nodes in the world's 5 major regions, synchronize the feature data of the CVE / NVD international vulnerability database in real time, and ensure the immutability of the data through blockchain evidence storage. Second, based on the federated learning framework, link global nodes to train a CNN-Transformer hybrid model, and encrypt and transmit the model parameters using differential privacy and the SM4 algorithm. Incrementally update the attack feature library daily. Then, build a policy distribution channel through the national secret SM9 algorithm to achieve second-level global synchronization of defense policies. The policy package includes triple verification of vulnerability feature fingerprints, handling rules, and weight parameters. Finally, form a trinity defense system of "intelligence - model - policy" in the cloud.

[0082] The specific method for constructing a hierarchical edge - fog - cloud collaborative architecture system based on the above - mentioned edge layer components, fog layer components, and cloud layer components is as follows:

[0083] By adopting a TSN time - sensitive network to construct a nanosecond - level synchronous communication backbone, deploying the KubeEdge edge management framework to achieve full - life - cycle management and control of containerized applications and the integrated design of federated learning and the national secret SM9 encryption channel, forming a closed - loop system of "edge real - time detection - fog domain intelligent decision - cloud global optimization";

[0084] Specifically, it should be noted that: Adopt a TSN time - sensitive network to construct a nanosecond - level synchronous communication backbone to ensure real - time data interaction between edge devices, fog nodes, and the cloud. Second, deploy the KubeEdge edge management framework to achieve full - life - cycle management and control of containerized applications, support the dynamic distribution of lightweight AI models from the cloud to the edge layer. Finally, through the integrated design of federated learning and the national secret SM9 encryption channel, complete the encrypted aggregation and collaborative analysis of multi - source threat data in the fog layer, form a closed - loop system of "edge real - time detection - fog domain intelligent decision - cloud global optimization", construct a hierarchical edge - fog - cloud collaborative architecture system, and achieve dynamic protection of the data security of industrial control computers.

[0085] Embodiment 3

[0086] This embodiment provides an electronic device, including: a processor and a memory, where the memory stores a computer program that can be called by the processor;

[0087] The processor executes the above - mentioned method for monitoring the operation of an underground reactor based on a sensor network by calling the computer program stored in the memory.

[0088] The electronic device may vary significantly due to different configurations or performances, and can include one or more processors and one or more memories. Among them, at least one computer program is stored in the memory, and the computer program is loaded and executed by the processor to implement the industrial control computer data security dynamic protection method based on AI edge computing provided by the above method embodiments. The electronic device can also include other components for implementing device functions. For example, the electronic device can also have components such as wired or wireless network interfaces and input / output interfaces for data input and output. This embodiment will not be elaborated here.

[0089] Embodiment 4

[0090] This embodiment provides a computer-readable storage medium, on which a rewritable computer program is stored;

[0091] When the computer program runs on a computer device, it enables the computer device to execute the above-mentioned industrial control computer data security dynamic protection method based on AI edge computing.

[0092] For example, the computer-readable storage medium can be a read-only memory, a random access memory, a read-only optical disc, magnetic tape, floppy disk, and optical data storage device, etc.

[0093] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions according to the embodiments of the present invention are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired network or / and a wireless network. The computer-readable storage medium can be any available medium that the computer can access or a data storage device such as a server or data center that contains one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

Claims

1. An industrial control computer data security dynamic protection method based on AI edge computing, characterized in that It includes the following specific steps: S1. Trusted data encryption and preprocessing: Encrypt the original industrial control instructions and sensor data through a hardware-level trusted execution environment, and preprocess the original data to form a standardized encrypted data stream; S2. Intelligent anomaly detection and protocol analysis: Based on the AI model, real-time detect abnormal data packets in the standardized encrypted data stream, perform context verification in combination with industrial control protocol characteristics, and generate threat scores in real time; S3. Dynamic security policy adjustment: Automatically switch the authentication mode according to the threat score level and execute the dynamic security adjustment policy; S4. Elastic resource allocation and optimization: Dynamically allocate edge computing resources according to the real-time threat index and device load; S5. Four-dimensional collaboration and closed-loop protection: Realize the dynamic protection of industrial control computer data security through a four-step closed-loop of encryption-detection-adjustment-optimization.

2. The industrial control computer data security dynamic protection method based on AI edge computing according to claim 1, wherein In S1, the original industrial control instructions and sensor data are encrypted through a hardware-level trusted execution environment. The specific implementation method is: Build a chip-level security isolation area based on the national cryptographic algorithm, bind the device identity with the PUF physical fingerprint, dynamically generate keys through quantum random numbers, encrypt the industrial control instructions, and use checksum protection for the sensor data. Finally, design is carried out in combination with blockchain evidence storage and anti-side channel.

3. The industrial control computer data security dynamic protection method based on AI edge computing according to claim 1, wherein In S1, the original data is preprocessed to form a standardized encrypted data stream. The specific implementation method is: Realize data standardization processing through a multi-level protocol parsing engine and a timing feature alignment algorithm, and embed the working condition context tags generated by the digital twin.

4. The industrial computer data security dynamic protection method based on AI edge computing according to claim 1, wherein, In S2, based on the AI model, real-time detect abnormal data packets in the standardized encrypted data stream, and perform context verification in combination with industrial control protocol characteristics. The specific implementation method is: First, complete the quantum key decryption in the TEE security environment, retain the time stamp and the PUF hardware fingerprint, and then through the spatio-temporal parallel analysis architecture, perform multi-dimensional feature fusion in combination with the working condition context provided by the digital twin. Finally, output the threat score based on the dynamic weight model.

5. The industrial control computer data security dynamic protection method based on AI edge computing according to claim 1, wherein In S3, the authentication mode is automatically switched according to the threat score level. The specific implementation method is: Based on the real-time threat scoring model, divide the threat levels into high and low levels, and automatically trigger the authentication mode switch through the policy engine. Use dynamic token authentication at low threats; start quantum-enhanced key exchange and iris biometric multi-factor authentication at high threats.

6. The industrial control computer data security dynamic protection method based on AI edge computing according to claim 1, characterized in that In S3, execute the dynamic security adjustment policy. The specific implementation method is: Automatically classify and store logs based on the real-time threat score. Record them in the local encrypted cache at low threats, and immediately write them into the quantum security cloud storage at high threats.

7. The method for dynamically protecting the data security of an industrial control computer based on AI edge computing according to claim 1, wherein In S4, dynamically allocate edge computing resources according to the real-time threat index and device load. The specific implementation method is: Dynamically allocate edge computing resources according to the formula: where T priority is the threat priority, C util is the current CPU utilization rate, and α, β are adaptive weight factors; The edge server collects device load data and threat scores in real time, and then the federated learning engine dynamically adjusts the α and β coefficients. Finally, the scheduler executes the allocation policy according to the calculation results.

8. An industrial control computer data security dynamic protection system based on AI edge computing, which is implemented based on the method of the industrial control computer data security dynamic protection system based on AI edge computing according to any one of claims 1-7, characterized in that, It specifically includes a data acquisition and processing module, an AI detection module, a policy regulation module, and a resource optimization module; Among them, the data acquisition module is used to obtain the operating data of the industrial control computer through sensors; The AI detection module is used to perform multi-modal fusion analysis and risk assessment on the obtained operation data of the industrial control computer through AI, and construct edge layer components; The policy regulation module is used to regulate the dynamic protection policy of the industrial control computer data security according to the risk assessment of the industrial control computer operation, and construct fog layer components; The resource optimization module is used to realize the efficient allocation of the computing resources of the industrial control system through a dynamic algorithm, and construct cloud components.

9. An electronic device, comprising: A processor and a memory, wherein the memory stores computer programs that can be called by the processor; It is characterized in that the processor executes the dynamic protection method for industrial control computer data security based on AI edge computing according to any one of claims 1-7 by calling the computer programs stored in the memory.

10. A computer-readable storage medium, characterized in that, Instructions are stored, and when the instructions are run on a computer, the computer is caused to execute the dynamic protection method for industrial control computer data security based on AI edge computing according to any one of claims 1-7.

Images