A drive information security management system and method based on a memristive neural network

Abstract

The present invention discloses a driving information security management system and method based on a memristive neural network, which relates to the technical field of information security. The method includes the steps of: S1, monitoring the synchronization between the real-time monitoring response system and the driving system in real time, and collecting feedback signal data; S2, analyzing and designing a synchronization controller through the Lyapunov stability theory; S3, using the memristive neural network to dynamically select an information encryption method according to the synchronization error, the feedback signal, and the signal encryption level; S4, ensuring that the probability of decrypting the encrypted information within the driving feedback interval is lower than a set threshold; S5, extracting and analyzing the historical encrypted signals, and updating the information encryption method and the network security level. Based on the Lyapunov stability theory, the present invention constructs a synchronization controller, and uses the memristive neural network to dynamically adjust the signal encryption level to ensure a high level of security during the dynamic change process.

Description

Technical Field

[0001] The present invention relates to the field of information security technology, and in particular, to a driving information security management system and method based on a memristive neural network. Background Art

[0002] With the development of information technology, industrial control systems and embedded systems are increasingly widely used in various industries. In these systems, the security and synchronization of information are key issues. Traditional information security management methods often focus on static encryption mechanisms or synchronization control strategies, and it is difficult to cope with security challenges in dynamic environments. Especially in systems that require real-time feedback and high synchronization, the encryption and decryption processes of information may be affected by changes in the system state, resulting in security and stability problems.

[0003] Although current solutions provide basic information encryption and synchronization control strategies, they lack sufficient support for dynamically adjusting synchronization errors, feedback signals, and encryption requirements that change in real time. This makes it difficult to simultaneously meet information security and system stability in complex system environments. Summary of the Invention

[0004] The purpose of the present invention is to provide a driving information security management system and method based on a memristive neural network to solve the problems raised in the prior art.

[0005] To achieve the above purpose, the present invention provides the following technical solution: A driving information security management method based on a memristive neural network, the method comprising the steps of:

[0006] S1. Monitor the synchronization of the response system and the driving system in real time, and collect feedback signal data;

[0007] S2. Analyze and design a synchronization controller through Lyapunov stability theory;

[0008] S3. Use a memristive neural network to dynamically select an information encryption method according to the synchronization error, feedback signal, and signal encryption level;

[0009] S4. Ensure that the probability of decrypting the encrypted information within the driving feedback interval is lower than a set threshold;

[0010] S5. Extract and analyze historical encrypted signals, and update the information encryption method and network security level.

[0011] According to the above solution, the synchronization is evaluated by monitoring the synchronization error between the response system and the driving system, and the formula is as follows:

[0012] e(t) = ‖y response (t) - y drive (t)‖;

[0013] Among them, e(t) represents the synchronization error at time t; y response (t) represents the output signal of the response system at time t; y drive (t) represents the output signal of the drive system at time t; t represents the time variable of continuous time;

[0014] The feedback signal includes the feedback of the state variables and control instructions of the response system and the drive system;

[0015] Collect the feedback signal data and perform data preprocessing. The data preprocessing includes removing duplicate data, filling in missing values, noise suppression, and normalization processing; and perform a detection and processing mechanism for abnormal data on the feedback signal data to prevent the impact of excessive errors or abnormal feedback signals on system decisions.

[0016] According to the above scheme, the Lyapunov stability theory analyzes the stability of the system synchronization error by defining an appropriate Lyapunov function;

[0017] The Lyapunov function is used to analyze whether the system can return to a stable state under the action of perturbations or errors, and ensure that the response system and the drive system remain stable during the synchronization process;

[0018] The system state is expressed as: x(t) = [x1(t), x2(t)] T ; Among them, x1(t) represents the state of the response system at time t; x2(t) represents the state of the drive system at time t; construct a Lyapunov function V(x) to describe the state of the system, and the formula is as follows:

[0019]

[0020] Among them, P represents a positive definite matrix, which is the weight coefficient of the system; x1 represents the state of the response system; x2 represents the state of the drive system;

[0021] The Lyapunov function is used to measure the magnitude of the system state difference. By analyzing the behavior of the system state over time, it is judged whether the system will stabilize;

[0022] Apply the second Lyapunov law to analyze the stability of the system by calculating the derivative of the Lyapunov function. The formula is as follows:

[0023]

[0024] Among them, when V(x) is a positive definite function, and its derivative is a negative definite function for all states x in the entire state space, that is, V(x) > 0, and When this occurs, it is concluded that the system is asymptotically stable, that is, the error between the response system and the drive system will gradually decrease and finally reach the synchronous state;

[0025] When constructing the Lyapunov function, inequality techniques are used to bound the upper bound of the error and ensure the reduction of the synchronization error. The inequality is as follows:

[0026]

[0027] Using the said inequality, analyze the relationship between the system state and the control input.

[0028] According to the above scheme, the synchronization controller, based on the Lyapunov stability theory, ensures that the response system and the drive system converge stably to the synchronous state; the synchronization controller needs to satisfy minimizing the synchronization error and ensuring the robustness of the system;

[0029] The formula of the synchronization controller is as follows:

[0030]

[0031] Among them, u(t) represents the synchronization controller; e(t) represents the synchronization error; t represents the time variable of continuous time; K p 、K i and K d respectively represent proportional, integral and differential gains; the synchronization controller optimizes the performance of the synchronization controller by comprehensively adjusting the proportional, integral and differential gains.

[0032] According to the above scheme, the memristive neural network includes an input layer, a hidden layer and an output layer;

[0033] The input layer receives input variables; the input variables include synchronization error, feedback signal and signal encryption level;

[0034] The synchronization error is used to quantify the synchronization error between the response system and the drive system;

[0035] The feedback signal includes the feedback of the state variables and control instructions of the response system and the drive system;

[0036] The signal encryption level is an indication of the current encryption strength, and the indication of the current encryption strength includes low, medium and high signal encryption levels;

[0037] The hidden layer performs weighted processing and nonlinear transformation on the input signal through neurons, so as to extract effective features;

[0038] The hidden layer contains M neurons. The neurons receive input variables, perform calculations using weighted summation, and then perform non-linear transformation through an activation function to generate output signals, and transmit the output signals to the next layer of neurons or finally to the output layer;

[0039] The weighted summation is as follows:

[0040]

[0041] where, z k (t) represents the activation value of the k-th output neuron; w ki represents the connection weight between the k-th output neuron and the i-th input variable; n represents the total number of input variables; i represents the index of the input variable; x i (t) represents the input variable; t represents the time variable of continuous time; b k represents the bias term of the neuron;

[0042] The activation function is as follows:

[0043] a k (t) = σ(z k (t));

[0044] where, a k (t) represents the output signal of the neuron; σ represents the activation function;

[0045] The output layer outputs an appropriate information encryption method according to the output signal of the hidden layer.

[0046] According to the above scheme, the memristive neural network adjusts the information encryption method according to real-time data, and dynamically adjusts the connection weights according to the synchronization error and feedback signal, as follows:

[0047]

[0048] where, w ki (t) represents the connection weight between the k-th neuron and the i-th input variable; η represents the learning rate, which controls the step size of weight update; represents the gradient of the synchronization error with respect to the weight; e(t) represents the synchronization error; h(t) represents the influence factor of the feedback signal on the weight; ΔL(t) represents the influence factor of the change in the signal encryption level on the weight; α and β respectively represent the influence weight factors that control the feedback signal and the change in the signal encryption level on the weight update;

[0049] The real-time state adjustment adopts the reinforcement learning method, and adjusts the network weights through real-time feedback to enable the network to quickly adapt to the dynamic environment.

[0050] According to the above scheme, the dynamic selection of information encryption method uses a memristor neural network to dynamically select an appropriate encryption method based on synchronization error, feedback signal and signal encryption level, ensuring that the encryption strength matches the real-time requirements of the system;

[0051] By designing an encryption mechanism, the risk of information being decrypted within the drive feedback interval is reduced, ensuring information security.

[0052] According to the above scheme, the corresponding encryption strength is calculated based on the synchronization error, the feedback signal and the signal encryption level. The encryption strength affects the probability of information being decrypted. The formula is as follows:

[0053] C(t)=f(e(t),y(t),L);

[0054] Where C(t) represents the encryption strength at time t; e(t) represents the synchronization error, which is the synchronization error between the response system and the drive system; y(t) represents the feedback signal; L represents the signal encryption level; f represents a function that combines the synchronization error, feedback signal, and signal encryption level to determine the appropriate encryption strength;

[0055] The probability of being decrypted is as follows:

[0056] P decrypt (t) = g(C(t), Δt);

[0057] Among them, P decrypt (t) represents the probability that the information is decrypted at time t; Δt represents the length of the drive feedback interval; g(C(t), Δt) is a function that represents the impact of encryption strength C(t) and drive feedback interval Δt on the decryption risk;

[0058] The probability of ensuring that the encrypted information is decrypted within the driving feedback interval is lower than the set threshold, that is, P decrypt (t)<θ, where θ represents the set threshold.

[0059] According to the above scheme, the historical encrypted signal is analyzed based on the historical synchronization error, the probability of being decrypted and the encryption strength;

[0060] The synchronization error is analyzed to obtain the average synchronization error, which is expressed as follows:

[0061]

[0062] in, Expressed as the average value of synchronization error; e i (t) represents the synchronization error at the jth time point t; N represents the number of historical signals;

[0063] Analyze the decryption probability, and obtain the average value of the decryption probability. The formula is as follows:

[0064]

[0065] Wherein, represents the average value of the decryption probability; C j (t) represents the decryption probability at the j-th time point t;

[0066] Analyze the encryption strength, and obtain the average value of the encryption strength. The formula is as follows:

[0067]

[0068] Wherein, represents the average value of the encryption strength; represents the encryption strength at the j-th time point t;

[0069] Combine the historical synchronization error, the decryption probability, and the analysis result of the encryption strength, and adjust the current information encryption method;

[0070] Evaluate the security of the overall network, and update the network security level according to historical analysis to ensure the security of the system.

[0071] A drive information security management system based on a memristive neural network, the system includes: a data monitoring module, a synchronization control module, a memristive neural network module, a risk assessment module, and a security optimization module;

[0072] The data monitoring module monitors the synchronization of the response system and the drive system in real time, collects the feedback signals of the response system and the drive system, and evaluates the synchronization state;

[0073] The synchronization control module designs and adjusts the synchronization control strategy by using the Lyapunov stability theory to ensure that the response system and the drive system can stably converge to the synchronization state;

[0074] The memristive neural network module dynamically selects the information encryption method according to the synchronization error, the feedback signal, and the signal encryption level to ensure that the encryption strength matches the real-time requirements and guarantee the information security; The memristive neural network module has an adaptive mechanism and can dynamically adjust the encryption method, the control strategy, and the network security level according to the attack type, the network environment, and the technical progress;

[0075] The risk assessment module analyzes the probability that the encrypted information is decrypted within the drive feedback interval to ensure that the probability of decryption within the drive feedback interval is lower than the set threshold, and reduces the decryption risk of the information during transmission;

[0076] The described security optimization module conducts data analysis on historical synchronization errors, decryption probability, and encryption strength, calculates the average value, analyzes the stability of the encrypted signal and network security; updates the current encryption scheme according to the analysis results, and evaluates and adjusts the network security level; through historical data and real-time feedback learning, adjusts the encryption strategy or optimization algorithm according to long-term operation data to improve overall security and performance.

[0077] Compared with the prior art, the beneficial effects of the present invention are:

[0078] 1. Based on the Lyapunov stability theory, the present invention constructs a synchronization controller to evaluate and adjust the synchronization of the response system and the drive system in real time, ensuring that the synchronization error gradually decreases;

[0079] 2. The present invention uses a memristive neural network to dynamically adjust the signal encryption level, effectively reducing the probability of information being decrypted within the drive feedback interval, ensuring a high level of security during the dynamic change process;

[0080] 3. By analyzing historical synchronization errors, decryption probability, and encryption strength, the present invention can optimize the current encryption method and network security level, making information security management more intelligent and efficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 It is a flowchart of the steps of a drive information security management method based on a memristive neural network according to the present invention;

[0082] Figure 2 It is a schematic structural diagram of a drive information security management system based on a memristive neural network according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0083] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0084] Embodiment: As Figure 1 - Figure 2 shown, the present invention provides a technical solution, a drive information security management method based on a memristive neural network;

[0085] S1. Monitor the synchronization of the response system and the drive system in real time, and collect feedback signal data;

[0086] The embodiments of the present application are described by taking the power monitoring center and the remote terminal device as examples. This is only for illustration purposes and is not limiting.

[0087] Specifically, synchronization must be maintained between the power monitoring center and the remote terminal device to ensure the effectiveness of instructions and feedback. Among them, the power monitoring center is the response system, and the remote terminal device is the drive system. There may be a certain delay or error between the power monitoring center and the remote terminal device; the synchronization of the response system and the drive system is evaluated by monitoring the synchronization error between the response system and the drive system. The formula is as follows:

[0088] e(t) = ‖y response (t) - y drive (t)‖;

[0089] Among them, e(t) represents the synchronization error at time t; y response (t) represents the output signal of the response system at time t; y drive (t) represents the output signal of the drive system at time t; t represents the time variable of continuous time;

[0090] For the synchronization error, for example, the output signal of the power monitoring center is 50, while the output signal of the remote terminal device is 45, and the device does not respond completely as expected. At this time, the synchronization error is: e(t) = |50 - 45| = 5;

[0091] Next, collect the feedback signal data. Among them, the feedback signal includes the state variables of the response system and the drive system and the feedback of the control instructions; for example: the power monitoring center is the response system, and the remote terminal device is the drive system, then the state variable is the state of a substation device being "running" or "closed", and the feedback of the control instruction is that the substation circuit breaker switch changes from "off" to "on". This is only for illustration purposes and is not limiting.

[0092] Subsequently, perform data preprocessing on the collected feedback signal data. The data preprocessing includes data cleaning, noise suppression, and normalization processing; for example, quickly identify and delete exactly the same records through a hash table or an index structure, and avoid receiving the same data points multiple times due to network latency or other reasons through deduplication; perform linear or spline interpolation based on the adjacent points before and after the time series to fill in the missing values and ensure the integrity of the data; use a filtering algorithm to suppress noise, which can improve the data quality and thus enhance the accuracy of the system. At the same time, a detection and processing mechanism for abnormal data is performed on the feedback signal data. For example, based on statistical methods or machine learning methods, identify and remove outliers. For the feedback signal and the synchronization error, an error threshold can be set. Once the error exceeds the set range, an early warning or recovery mechanism is automatically started to prevent the impact of excessive errors or abnormal feedback signals on system decision-making.

[0093] S2. Analyze and design a synchronization controller through Lyapunov stability theory;

[0094] Specifically, using Lyapunov stability theory, define a suitable Lyapunov function to analyze the stability of the synchronization error of the system; among them, the Lyapunov function is used to analyze whether the system can return to a stable state under the action of disturbances or errors; use Lyapunov stability theory to analyze whether the synchronization error between the power monitoring center and the remote terminal device will gradually decrease. If the change of the synchronization error can be effectively controlled, it indicates that the system is stable.

[0095] The system state is expressed as: x(t) = [x1(t), x2(t)] T ; where, x1(t) represents the state of the response system at time t; x2(t) represents the state of the drive system at time t; construct a Lyapunov function V(x) to describe the state of the system, and the formula is as follows:

[0096]

[0097] Among them, P represents a positive definite matrix, which is the weight coefficient of the system; x1 represents the state of the response system; x2 represents the state of the drive system;

[0098] The Lyapunov function is used to measure the magnitude of the system state difference. By analyzing the behavior of the system state changing with time, it is judged whether the system will stabilize;

[0099] Apply Lyapunov's second law. By calculating the derivative of the Lyapunov function, analyze the stability of the system, and the formula is as follows:

[0100]

[0101] Among them, when V(x) is a positive definite function, and its derivative is a negative definite function for all states x in the entire state space, that is, V(x)>0, and ; then it is concluded that the system is asymptotically stable, that is, the error between the response system and the drive system will gradually decrease and finally reach a synchronous state; for example, the derivative of the obtained Lyapunov function is -2, which means that the system will gradually tend to be stable, that is, the synchronization error between the power monitoring center and the remote terminal device will decrease.

[0102] At the same time, use inequality techniques to define the upper bound of the error and ensure the reduction of the synchronization error. The inequality is as follows:

[0103]

[0104] Analyze the relationship between the system state and the control input using the inequality.

[0105] Next, based on the Lyapunov stability theory, construct a synchronization controller to ensure that the response system and the drive system converge stably to the synchronization state. The formula is as follows:

[0106]

[0107] where u(t) represents the synchronization controller; e(t) represents the synchronization error; t represents the time variable of continuous time; K p 、K i and K d represent the proportional, integral, and differential gains respectively. The synchronization controller optimizes the performance of the synchronization controller by comprehensively adjusting the proportional, integral, and differential gains, thereby effectively reducing errors and maintaining the synchronization of the system, which helps to improve communication efficiency and ensure the accuracy and timeliness of information transmission. For example, problems such as sensor noise and network delay may be encountered, which will have a negative impact on synchronization. Using the Lyapunov stability theory to construct a synchronization controller can enhance the robustness of the system and enable it to maintain good performance in the face of interference.

[0108] S3. Use a memristive neural network to dynamically select an information encryption method according to the synchronization error, feedback signal, and signal encryption level;

[0109] In different working environments, system loads, and communication states, the synchronization error, feedback signal, and signal encryption level will change. By adjusting the encryption method in real time, maintain the best balance between resources and security to meet actual needs, which can not only improve information security but also avoid waste of computing resources caused by over-encryption. If only one of the conditions is considered alone, for example, only selecting the encryption method based on the synchronization error, the synchronization error only reflects the time difference between systems and cannot fully reflect the security requirements of data. And adjusting the encryption method according to the synchronization error may lead to over-encryption or under-encryption in some cases. For example, when the synchronization error is small but the system load is high, if high-strength encryption continues to be used, it will consume too much computing resources and affect system performance.

[0110] While ensuring that the synchronization error gradually decreases, it is necessary to ensure the security of the information transmitted between the power monitoring center and the remote terminal device. To prevent data from being maliciously stolen or tampered with, use a memristive neural network to dynamically select an encryption strategy;

[0111] Specifically, a memristive neural network is constructed based on the collected feedback signal data. The memristive neural network includes an input layer, a hidden layer, and an output layer. Among them, the input layer receives input variables, which include synchronization error, feedback signal, and signal encryption level. The synchronization error is used to quantify the synchronization error between the response system and the drive system. The feedback signal includes the state variables of the response system and the drive system and the feedback of the control instructions. The signal encryption level is an indication of the current encryption strength, and the indication of the current encryption strength includes low, medium, and high signal encryption levels. After the input variables enter the hidden layer through the input layer, the hidden layer performs weighted processing and non-linear transformation on the input signals through neurons, thereby extracting effective features. Among them, the hidden layer contains M neurons. The neurons receive the input variables, calculate using weighted summation, and then perform non-linear transformation through the activation function to generate output signals, and transmit the output signals to the next layer of neurons or finally to the output layer.

[0112] The weighted summation is as follows:

[0113]

[0114] Among them, z k (t) represents the activation value of the k-th output neuron; w ki represents the connection weight between the k-th output neuron and the i-th input variable; n represents the total number of input variables; i represents the index of the input variable; x i (t) represents the input variable; t represents the time variable of continuous time; b k represents the bias term of the neuron;

[0115] The activation function is as follows:

[0116] a k (t) = σ(z k (t));

[0117] Among them, a k (t) represents the output signal of the neuron; σ represents the activation function;

[0118] Next, the output layer receives the output signal of the hidden layer and selects an information encryption method according to the output signal of the hidden layer to ensure that the encryption strength matches the real-time requirements of the system. The information encryption methods include, but are not limited to, encryption methods such as symmetric encryption, asymmetric encryption, and quantum encryption. When selecting an encryption method, the consumption of computing resources will be comprehensively considered, and an appropriate encryption algorithm will be dynamically selected according to the resource limitations of the system to balance the encryption strength and the consumption of computing resources.

[0119] For example, the synchronization error e(t1) = 0.1 indicates a slight asynchrony between the response system and the drive system; the feedback signal y(t1) shows a normal working state; the current signal encryption level L = medium; based on the synchronization error, feedback signal, and signal encryption level, the memristive neural network obtains the encryption intensity C(t1) and outputs the AES-256-bit symmetric encryption algorithm with medium strength for information encryption. The AES-256-bit symmetric encryption algorithm has a relatively low computational complexity, but its key distribution problem is relatively complex; this is only for illustrative purposes and not for limitation.

[0120] And by designing an encryption mechanism, the risk of information being decrypted within the drive-feedback interval is reduced to ensure information security.

[0121] Meanwhile, the memristive neural network adjusts the information encryption method according to real-time data, and dynamically adjusts the connection weights based on the synchronization error and feedback signal. The formula is as follows:

[0122]

[0123] where w ki (t) represents the connection weight between the k-th neuron and the i-th input variable; η represents the learning rate, which controls the step size of weight update; represents the gradient of the synchronization error with respect to the weight; e(t) represents the synchronization error; h(t) represents the influence factor of the feedback signal on the weight; ΔL(t) represents the influence factor of the change in the signal encryption level on the weight; α and β respectively represent the influence weight factors that control the influence of the feedback signal and the change in the signal encryption level on the weight update.

[0124] S4. Ensure that the probability of the encrypted information being decrypted within the drive-feedback interval is lower than the set threshold;

[0125] Specifically, the corresponding encryption intensity is calculated based on the synchronization error, feedback signal, and signal encryption level. The encryption intensity affects the probability of information being decrypted. The formula is as follows:

[0126] C(t) = f(e(t), y(t), L);

[0127] where C(t) represents the encryption intensity at time t; e(t) represents the synchronization error, which is the synchronization error between the response system and the drive system; y(t) represents the feedback signal; L represents the signal encryption level; f represents a function that combines the synchronization error, feedback signal, and signal encryption level to determine the appropriate encryption intensity;

[0128] The probability of being decrypted is given by the formula:

[0129] P decrypt (t) = g(C(t), Δt);

[0130] Among them, P decrypt (t) represents the probability that the information is decrypted at time t; Δt represents the time length of the drive feedback interval; g(C(t), Δt) is a function representing the influence of the encryption strength C(t) and the drive feedback interval Δt on the decryption risk;

[0131] Ensure that the probability of decrypting the encrypted information within the drive feedback interval is lower than the set threshold, that is, P decrypt (t) < θ, where θ represents the set threshold; for example: the set threshold is 0.05, since the probability of being decrypted is P decrypt (t) = 0.01 < θ = 0.05, then the current encryption policy is secure and does not need to be adjusted.

[0132] S5. Extract and analyze historical encryption signals, and update the information encryption method and network security level.

[0133] Specifically, extract all encryption communication records within a period of time from the system log or database. These records should include but are not limited to encryption methods, key usage, synchronization errors, and feedback signals, etc.; and remove duplicate data and fill in missing values for these data to ensure data integrity and consistency; analyze and screen out the features closely related to the analysis target, including synchronization errors, decryption probabilities, and encryption strengths; among them, analyze the synchronization errors to obtain the average value of synchronization errors, and the formula is as follows:

[0134]

[0135] Among them, represents the average value of the synchronization error; e i (t) represents the synchronization error at the j-th time point t; N represents the number of historical signals;

[0136] While analyzing the synchronization errors, through time series data analysis methods, such as the moving average method, identify the long-term trends and periodic changes in the time series of synchronization errors to evaluate the stability and response speed of the system;

[0137] Analyze the decryption probability to obtain the average value of the decryption probability, and the formula is as follows:

[0138]

[0139] Among them, represents the average value of the decryption probability; C j (t) represents the decryption probability at the j-th time point t;

[0140] Analyze the encryption strength to obtain the average encryption strength. The formula is as follows:

[0141]

[0142] Among them, represents the average encryption strength; represents the encryption strength at the j-th time point t;

[0143] Combine the historical synchronization error, the probability of being decrypted, and the analysis result of the encryption strength to adjust the current information encryption method. For example, construct a scoring model, assign weights to each dimension and calculate the total score, and decide whether to update the encryption method currently according to the result of the scoring model;

[0144] For example: when the synchronization error increases and the feedback signal is abnormal, the analysis shows that the current encryption strength is not sufficient to cope with potential security risks, and it may be necessary to increase the encryption strength, upgrade the encryption strategy, and adjust the current information encryption method. For example: upgrade from the AES-256-bit symmetric encryption algorithm to combine RSA asymmetric encryption for key exchange, while maintaining the AES-256-bit symmetric encryption algorithm for data encryption to enhance security; this is only for illustrative purposes and is not limited.

[0145] Evaluate the security of the overall network and update the network security level according to historical analysis to ensure the security of the data transmitted in each communication and prevent sensitive information from being decrypted by unauthorized third parties.

[0146] The present invention provides another technical solution, a drive information security management system based on a memristive neural network;

[0147] The system includes: a data monitoring module, a synchronization control module, a memristive neural network module, a risk assessment module, and a security optimization module;

[0148] The data monitoring module monitors the synchronization of the response system and the drive system in real time, collects the feedback signals of the response system and the drive system, and evaluates the synchronization state;

[0149] The synchronization control module designs and adjusts the synchronization control strategy using the Lyapunov stability theory to ensure that the response system and the drive system can stably converge to the synchronization state;

[0150] The memristive neural network module dynamically selects the information encryption method according to the synchronization error, the feedback signal, and the signal encryption level to ensure that the encryption strength matches the real-time requirements and guarantee the information security; the memristive neural network module has an adaptive mechanism and can dynamically adjust the encryption method, the control strategy, and the network security level according to the attack type, the network environment, and the technological progress;

[0151] The risk assessment module analyzes the probability that the encrypted information is decrypted within the drive feedback interval, ensures that the probability of being decrypted within the drive feedback interval is lower than the set threshold, and reduces the decryption risk of the information during transmission;

[0152] The security optimization module performs data analysis on the historical synchronization error, the probability of being decrypted, and the encryption strength, calculates the average value, and analyzes the stability of the encrypted signal and the network security; updates the current encryption scheme according to the analysis results, and evaluates and adjusts the network security level; through historical data and real-time feedback learning, adjusts the encryption strategy or optimization algorithm according to the long-term operation data to improve the overall security and performance.

[0153] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and without departing from the spirit or basic characteristics of the present invention, the present invention can be implemented in other specific forms. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be embraced within the present invention. Any reference signs in the claims should not be construed as limiting the claims involved.

Claims

1. A drive information security management method based on a memristive neural network, characterized in that: The method includes the steps of: S1. Monitor the synchronization between the response system and the drive system in real time, and collect feedback signal data; S2. Analyze and design a synchronization controller through the Lyapunov stability theory; S3. Utilize a memristive neural network to dynamically select an information encryption method according to the synchronization error, feedback signal, and signal encryption level; The memristive neural network adjusts the information encryption method according to real-time data, and dynamically adjusts the connection weights according to the synchronization error and feedback signal. The formula is as follows: ； where, w ki (t) represents the connection weight between the k-th neuron and the i-th input variable; η represents the learning rate, controlling the step size of weight update; represents the gradient of the synchronization error with respect to the weight; e(t) represents the synchronization error; h(t) represents the influence factor of the feedback signal on the weight; represents the influence factor of the change in the signal encryption level on the weight; α and β respectively represent the influence weight factors controlling the influence of the feedback signal and the change in the signal encryption level on the weight update; S4. Ensure that the probability of decrypting the encrypted information within the drive feedback interval is lower than a set threshold; S5. Extract and analyze historical encrypted signals, and update the information encryption method and network security level; Analyze historical encrypted signals according to historical synchronization errors, decryption probabilities, and encryption intensities; Analyze the synchronization error to obtain the average value of the synchronization error. The formula is as follows: ； Among them, is expressed as the average value of the synchronization error; e i (t) is expressed as the synchronization error at the j-th time point t; N is expressed as the number of historical signals; Analyze the decryption probability to obtain the average value of the decryption probability. The formula is as follows: ； Among them, represents the average value of the probability of being decrypted; C j (t) represents the probability of being decrypted at the j-th time point t; Analyze the encryption intensity to obtain the average value of the encryption intensity. The formula is as follows: ； Among them, is expressed as the average encryption strength; is expressed as the encryption strength at the j-th time point t; Combine the analysis results of historical synchronization errors, decryption probabilities, and encryption intensities to adjust the current information encryption method; Evaluate the security of the overall network, and update the network security level according to historical analysis to ensure the security of the system.

2. A drive information security management method based on a memristive neural network according to claim 1, characterized in that: The synchronization is evaluated by monitoring the synchronization error between the response system and the drive system. The formula is as follows: e(t)=‖y response (t)-y drive (t)‖; Among them, e(t) represents the synchronization error at time t; y response (t) represents the output signal of the response system at time t; y drive (t) represents the output signal of the drive system at time t; t represents the time variable of continuous time; The feedback signal includes the state variables of the response system and the drive system and the feedback of control instructions.

3. A drive information security management method based on a memristive neural network according to claim 1, characterized in that: The Lyapunov stability theory analyzes the stability of the synchronization error of the system by defining an appropriate Lyapunov function; The Lyapunov function is used to analyze whether the system can return to a stable state under the action of disturbances or errors, and ensure that the response system and the drive system remain stable during the synchronization process; The system state is represented as: x(t) = [x1(t), x2(t)] T ; where, x1(t) represents the state of the response system at time t; x2(t) represents the state of the drive system at time t; construct the Lyapunov function V(x) to describe the state of the system, and the formula is as follows: ； Among them, P represents a positive definite matrix, which is the weight coefficient of the system; x1 represents the state of the response system; x2 represents the state of the drive system; The Lyapunov function is used to measure the magnitude of the system state difference, and judge whether the system will stabilize by analyzing the behavior of the system state changing with time; Apply the second Lyapunov law to analyze the stability of the system by calculating the derivative of the Lyapunov function. The formula is as follows: ； where, when is a positive definite function, and its derivative is a negative definite function for all states x in the entire state space, that is , and ; then it is concluded that the system is asymptotically stable, that is, the error between the response system and the drive system will gradually decrease and finally reach the synchronous state; When constructing the Lyapunov function, use inequality techniques to define the upper bound of the error and ensure the reduction of the synchronization error. The inequality is as follows: ； Use the inequality to analyze the relationship between the system state and the control input.

4. A drive information security management method based on a memristive neural network according to claim 3, characterized in that: The synchronization controller, based on the Lyapunov stability theory, ensures that the response system and the drive system stably converge to the synchronization state; the synchronization controller needs to satisfy minimizing the synchronization error and ensuring the robustness of the system; The synchronization controller has the following formula: ； wherein, u(t) represents the synchronization controller; e(t) represents the synchronization error; t represents the time variable of continuous time; K p , K i and K d respectively represent the proportional, integral and derivative gains; the synchronization controller optimizes the performance of the synchronization controller by comprehensively adjusting the proportional, integral and derivative gains.

5. A drive information security management method based on a memristive neural network according to claim 1, characterized in that: The memristive neural network includes an input layer, a hidden layer, and an output layer; The input layer receives input variables; the input variables include synchronization error, feedback signal, and signal encryption level; The synchronization error is used to quantify the synchronization error between the response system and the drive system; The feedback signal includes the state variables of the response system and the drive system and the feedback of the control instruction; The signal encryption level is an indication of the current encryption strength, and the indication of the current encryption strength includes low, medium, and high signal encryption levels; The hidden layer performs weighted processing and non-linear transformation on the input signal through neurons, thereby extracting effective features; The hidden layer contains M neurons. The neurons receive input variables, calculate using weighted summation, then perform non-linear transformation through an activation function, generate an output signal, and transmit the output signal to the next layer of neurons or finally to the output layer; The weighted summation has the following formula: ； where z k (t) represents the activation value of the k-th output neuron; w ki represents the connection weight between the k-th output neuron and the i-th input variable; n represents the total number of input variables; i represents the index of the input variable; x i (t) represents the input variable; t represents the time variable of continuous time; b k represents the bias term of the neuron; The activation function has the following formula: ； Among them, a k (t) represents the output signal of the neuron; σ represents the activation function; The output layer outputs an appropriate information encryption method according to the output signal of the hidden layer.

6. A drive information security management method based on a memristive neural network according to claim 1, characterized in that: The dynamic selection of the information encryption method uses a memristive neural network to dynamically select an appropriate encryption method based on the synchronization error, feedback signal, and signal encryption level, ensuring that the encryption strength matches the real-time requirements of the system; By designing an encryption mechanism, the risk of information being decrypted during the drive feedback interval is reduced to ensure information security.

7. A drive information security management method based on a memristive neural network according to claim 1, characterized in that: Calculate the corresponding encryption strength according to the synchronization error, feedback signal, and signal encryption level. The encryption strength affects the probability of information being decrypted, and the formula is as follows: ； Where C(t) represents the encryption strength at time t; e(t) represents the synchronization error, which is the synchronization error between the response system and the drive system; y(t) represents the feedback signal; L represents the signal encryption level; f represents a function that combines the synchronization error, feedback signal, and signal encryption level to determine the appropriate encryption strength; The probability of being decrypted has the following formula: ； where, P decrypt (t) represents the probability that the information is decrypted at time t; Δt represents the time length of the drive feedback interval; g(C(t), Δt) is a function representing the influence of the encryption strength C(t) and the drive feedback interval Δt on the decryption risk; The probability that the encrypted information is decrypted within the drive feedback interval is ensured to be lower than a set threshold, i.e., P decrypt (t) < θ, where θ represents the set threshold.

8. A drive information security management system based on a memristive neural network, which is applied to a drive information security management method based on a memristive neural network according to any one of claims 1-7, and is characterized in that: The system includes: a data monitoring module, a synchronization control module, a memristive neural network module, a risk assessment module, and a security optimization module; The data monitoring module monitors the synchronization of the response system and the drive system in real time, collects the feedback signals of the response system and the drive system, and evaluates the synchronization state; The synchronization control module designs and adjusts the synchronization control strategy using the Lyapunov stability theory to ensure that the response system and the drive system can stably converge to the synchronization state; The memristive neural network module dynamically selects an information encryption method according to the synchronization error, feedback signal, and signal encryption level, ensuring that the encryption strength matches the real-time requirements and guaranteeing information security; The risk assessment module analyzes the probability of the encrypted information being decrypted within the drive feedback interval, ensures that the probability of being decrypted within the drive feedback interval is lower than the set threshold, and reduces the decryption risk of the information during transmission; The security optimization module conducts data analysis on the historical synchronization error, the probability of being decrypted, and the encryption strength, calculates the average value, and analyzes the stability of the encrypted signal and the network security; updates the current encryption scheme according to the analysis results, and evaluates and adjusts the network security level.

Images