Train control method, device and equipment for composite network attack, medium and product

Abstract

The invention discloses a composite network attack train control method, device and equipment, a medium and a product. The method comprises the following steps: acquiring a first running speed and first channel communication data of a train; based on the first channel communication data, judging whether a DoS attack aiming at the transmission of the first set point occurs in the train or not, and if so, compensating the first set point to obtain a second set point; a first traction control instruction is generated based on the second set point and the first running speed, a second running speed of the train at the next moment is obtained through prediction in combination with the discrete nonlinear model, then data reconstruction is conducted on the second running speed to obtain a third running speed, and whether the FD I attack aiming at transmission of the second running speed occurs or not is judged according to the third running speed; if yes, taking the speed as a fourth operation speed after compensation of the FD I attack until a preset iteration termination condition is met, and obtaining a target operation speed; and controlling the train based on the target running speed. The control precision of the train under the composite network attack can be improved.

Description

Technical Field

[0001] This invention relates to the field of train control, and more particularly to train control methods, apparatus, equipment, media, and products that are susceptible to complex network attacks. Background Technology

[0002] High-speed trains, as a fast and comfortable mode of transportation, have become a key component of modern railway transportation. In the high-speed train control system, the Automatic Train Operation (ATO) system is the core for achieving precise speed regulation and automatic operation. Its control accuracy directly affects traction energy consumption, on-time performance, passenger comfort, and operational safety. High-speed trains operating on fixed lines require high-precision speed control to strictly track the pre-set desired trajectory, ensuring on-time operation, energy efficiency optimization, and preventing safety accidents caused by speeding, over-driving, or other control malfunctions.

[0003] Currently, for speed tracking problems during repetitive train operations, existing technologies mainly employ traditional feedback control (such as PID control) or iterative learning control (ILC) methods. However, most existing control schemes are designed based on ideal communication environments or only consider resisting a single type of network attack, lacking a resilient control mechanism capable of uniformly addressing complex network attack scenarios where both types of attacks occur simultaneously. Under complex attacks, DoS (Denial of Service) attacks cause data loss by interfering with communication channels, while FDI (Fake Data Injection) attacks cause information distortion by tampering with data content. After the system suffers from these two different and independent attacks, the information acquired by the controller becomes neither complete nor accurate, causing the control law updates of existing methods to fail, tracking performance to deteriorate sharply, and control accuracy to drop significantly. Summary of the Invention

[0004] This invention provides a train control method, apparatus, equipment, medium, and product for complex network attacks, which can improve the control accuracy of trains under complex network attacks.

[0005] In a first aspect, an embodiment of the present invention provides a train control method against complex network attacks, comprising:

[0006] The train's first operating speed and first channel communication data are acquired in real time, wherein the first channel communication data is obtained by monitoring the first operating speed during transmission over the communication network;

[0007] Based on the communication data of the first channel, it is determined whether the train has been subjected to a DoS attack targeting the transmission of the first set point. If so, the first set point is compensated to obtain the second set point, wherein the first set point is transmitted in the communication network.

[0008] Based on the second setpoint and the first operating speed, a first traction control command is generated. Combining the discrete nonlinear model and the first traction control command, the second operating speed of the train at the next moment is predicted. The second operating speed is reconstructed to obtain a third operating speed. Based on the third operating speed, it is determined whether the train has experienced an FDI attack targeting the second operating speed. If so, the third operating speed is used as the fourth operating speed after compensation for the FDI attack. This process continues until a preset iteration termination condition is met to obtain the target operating speed.

[0009] The train is controlled based on the target operating speed.

[0010] By acquiring the train's initial operating speed and the first channel communication data monitored during its transmission through the communication network in real time, this provides accurate and reliable physical state input and raw network state information for subsequent control decisions, improving the baseline accuracy of the control system from the source. Based on the first channel communication data, it determines whether the train is experiencing a DoS attack targeting the first setpoint in the communication network transmission, and compensates for the affected first setpoint to obtain a second setpoint. This effectively resists control command interruptions caused by packet loss, providing a foundation for maintaining control accuracy. Based on the second setpoint and the first operating speed, traction control commands are generated, ensuring that the traction control commands closely match the current operating state and the compensated target, directly contributing to improved control accuracy. Finally, by combining the traction control commands and a discrete nonlinear model, the train's second operating speed at the next moment is predicted. The system reconstructs the second operating speed to obtain a third operating speed. Based on this third operating speed, it determines whether an FDI attack targeting the second operating speed exists. If so, the reconstructed third operating speed is used as the fourth operating speed after compensating for the FDI attack. This effectively identifies and filters maliciously tampered speed information in the feedback loop, fundamentally ensuring the authenticity of the feedback information and improving the control accuracy of trains under composite network attacks. After multiple iterations of the preceding operations, the target operating speed is obtained. Iterative optimization can continuously correct speed data deviations caused by composite network attacks, forming a high-precision speed control benchmark adapted to the attack scenario. Finally, the train is controlled according to the target operating speed, ensuring that the final executed command accurately drives the train to converge to the correct speed state, effectively improving the control accuracy of trains under composite network attacks. This application can improve the control accuracy of trains under composite network attacks.

[0011] Furthermore, if the condition exists, compensation is applied to the first set point to obtain the second set point, specifically including:

[0012] If a DoS attack is transmitted to the first set point, the acquired historical set point is compensated with a gain to obtain a DoS compensation value, wherein the historical set point is set by a preset expected trajectory.

[0013] The first setpoint is compensated using the DoS compensation value to obtain the second setpoint.

[0014] When a DoS attack is detected targeting the first setpoint, the DoS compensation value is obtained by using historical setpoints generated from a preset desired trajectory to compensate for the gain. This compensation is then applied to the first setpoint, enabling the rapid generation of a second setpoint unaffected by the DoS attack. This ensures the integrity and reliability of the setpoints transmitted to subsequent steps, providing a correct target reference for generating accurate traction control commands, thereby improving the control accuracy of the vehicle under complex network attacks.

[0015] Furthermore, the step of generating a first traction control command based on the second setpoint and the first operating speed specifically includes:

[0016] The first tracking error is calculated based on the second setpoint and the first running speed;

[0017] Based on the first tracking error and the second tracking error, the error change is calculated, wherein the second tracking error is calculated based on the third set point and the third running speed of the previous moment.

[0018] The first tracking error and the error change amount are controlled by gain to generate the first traction control command.

[0019] By calculating the first tracking error based on the second setpoint and the first running speed, and combining it with the second tracking error calculated from the data of the previous moment to calculate the error change, and then performing control gain processing based on the two to generate the first traction control command, the control command can synchronously and accurately respond to the real-time deviation of the speed and its dynamic change trend. This enables the train speed to be quickly and smoothly adjusted initially in a complex attack environment, directly improving the dynamic accuracy of the control.

[0020] Furthermore, the step of combining the discrete nonlinear model and the first traction control command to predict the train's second operating speed at the next moment specifically includes:

[0021] Obtain the preset first speed reference value;

[0022] The first speed reference value and the first traction control command are input into the discrete nonlinear model for prediction to obtain the second operating speed of the train at the next moment.

[0023] By inputting a preset first speed reference and a first traction control command into a discrete nonlinear model to construct a correlation prediction logic between traction and speed, the system can stably output the predicted speed value for the next moment, i.e., the second operating speed, in complex network environments. This avoids the loss of reliable basis for speed prediction due to complex network attacks. At the same time, the discrete nonlinear model can adapt to the dynamic characteristics of the train, ensuring the accuracy of speed prediction and helping to improve the train control precision under complex network attacks.

[0024] Furthermore, the process of reconstructing the data from the second operating speed to obtain the third operating speed specifically includes:

[0025] Obtain the preset second speed reference value and the preset traction force change reference value;

[0026] Using the second speed reference value and the traction force change reference value, the second operating speed is adjusted with an adjustable gain to obtain the third operating speed.

[0027] By using a preset second speed reference value and a preset traction force change reference value, the second operating speed can be reconstructed with adjustable gain, which can quickly obtain speed data close to the real state. The adjustable gain can flexibly adapt to the train's operating characteristics, ensuring that the reconstruction process is accurate and controllable, providing a reliable basis for subsequent train control, and effectively improving the control accuracy of trains under complex network attacks.

[0028] Furthermore, the step of determining whether the train is experiencing an FDI attack targeting the second operating speed based on the third operating speed specifically includes:

[0029] Based on the third operating speed and the fifth operating speed, the speed difference is calculated, wherein the fifth operating speed is the actual received value of the second operating speed after being transmitted through the communication network;

[0030] If the speed difference is greater than a preset detection threshold, the train is determined to have been subjected to the FDI attack.

[0031] By comparing the difference between the reconstructed third operating speed and the transmitted fifth operating speed, FDI attacks can quickly and accurately identify tampering with speed data. The preset detection threshold provides a clear quantitative standard for attack judgment, which can effectively avoid misjudgment of attacks caused by transmission fluctuations or environmental interference, ensure timely identification of FDI attacks, prevent false speed data from entering the control loop, and significantly improve the reliability and accuracy of train control under complex network attacks.

[0032] Secondly, an embodiment of the present invention provides a train control device for complex network attacks, characterized in that it includes a first module, a second module, a third module and a fourth module;

[0033] The first module is used to acquire the first operating speed of the train and the first channel communication data in real time, wherein the first channel communication data is obtained by monitoring the first operating speed during the transmission process of the communication network;

[0034] The second module is used to determine whether the train has been subjected to a DoS attack against the transmission of the first set point based on the communication data of the first channel. If so, the first set point is compensated to obtain a second set point, wherein the first set point is transmitted in the communication network.

[0035] The third module is used to generate a first traction control command based on the second setpoint and the first operating speed, and predict the second operating speed of the train at the next moment by combining the discrete nonlinear model and the first traction control command. The second operating speed is then reconstructed to obtain a third operating speed. Based on the third operating speed, it is determined whether the train has experienced an FDI attack targeting the second operating speed. If so, the third operating speed is used as the fourth operating speed after compensation for the FDI attack, until a preset iteration termination condition is met to obtain the target operating speed.

[0036] The fourth module is used to control the train based on the target operating speed.

[0037] In this way, the first module acquires the train's first operating speed and the first channel communication data monitored during its transmission through the communication network in real time, providing real and reliable physical state input and original network state information for subsequent control decisions, thereby improving the baseline accuracy of the control system from the source. The second module determines whether the train is experiencing a DoS attack targeting the first setpoint in the communication network transmission based on the first channel communication data, and compensates for the affected first setpoint to obtain a second setpoint, which can effectively resist the interruption of control commands due to data packet loss, providing a foundation for maintaining control accuracy. The third module generates traction control commands based on the second setpoint and the first operating speed, making the traction control commands closely aligned with the current operating state and the compensated target, directly contributing to improving control accuracy. Combining the traction control commands and discrete nonlinear models, the second speed of the train at the next moment is predicted. The system first calculates the train's operating speed and then reconstructs the data from the second operating speed to obtain a third operating speed. Based on this third operating speed, it determines whether an FDI attack targeting the second operating speed exists. If so, the reconstructed third operating speed is used as the fourth operating speed after compensating for the FDI attack. This effectively identifies and filters out maliciously tampered speed information in the feedback loop, fundamentally ensuring the authenticity of the feedback information and improving the control accuracy of trains under composite network attacks. After multiple iterations of the preceding operations, the target operating speed is obtained. Iterative optimization can continuously correct speed data deviations caused by composite network attacks, forming a high-precision speed control benchmark adapted to the attack scenario. Finally, the fourth module controls the train according to the target operating speed, ensuring that the final executed command accurately drives the train to converge to the correct speed state, effectively improving the control accuracy of trains under composite network attacks.

[0038] Thirdly, another embodiment of the present invention also provides a terminal device, including: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other through the communication bus;

[0039] The memory is used to store at least one executable instruction that causes the processor to perform operations of a train control method for a complex network attack.

[0040] Fourthly, another embodiment of the present invention provides a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device or apparatus containing the computer-readable storage medium to perform a train control method for a complex network attack.

[0041] Fifthly, another embodiment of the present invention provides a computer program product, including a computer program or instructions, which, when executed by a communication device, implements a train control method for a complex network attack. Attached Figure Description

[0042] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0043] Figure 1 This is a flowchart illustrating one embodiment of a train control method for a complex network attack provided in this application;

[0044] Figure 2 This is a schematic diagram illustrating the variation of additional resistance per unit mass in special operating sections of the train provided in this application;

[0045] Figure 3 This is a schematic diagram of the expected trajectory of the output of the discrete nonlinear model that satisfies the convergence condition provided in this application;

[0046] Figure 4 This is an iterative flowchart of the train speed control method against composite network attacks provided in this application;

[0047] Figure 5 This is a schematic diagram of the structure of a train control device for a complex network attack provided in this application. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0049] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0050] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0051] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0052] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0053] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0054] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0055] In the field of train control, the Automatic Train Operation (ATO) system is the core of precise speed control and automatic operation of high-speed trains. Its accuracy directly affects energy consumption, punctuality, comfort, and safety, especially in fixed-line reciprocating operations where it must strictly track the desired trajectory to ensure safety and energy conservation. Existing technologies have limitations: First, train repetitive speed tracking schemes (such as PID and ILC) are mostly based on ideal communication environments or only defend against single network attacks; second, they lack flexible mechanisms to deal with combined DoS and FDI attacks. These attacks cause data loss and distortion respectively, and their combined effect renders controller information ineffective, ultimately leading to a significant decrease in control accuracy.

[0056] See Figure 1 To improve the control accuracy of trains under complex network attacks, an embodiment of the present invention provides a train control method for complex network attacks, including steps S101 to S104.

[0057] Step S101: Real-time acquisition of the train's first operating speed and first channel communication data, wherein the first channel communication data is obtained by monitoring the first operating speed during transmission over the communication network;

[0058] In some embodiments, the train's first operating speed and first channel communication data are acquired in real time. The first channel communication data is obtained by monitoring the first operating speed during the transmission process of the communication network. Specifically, this includes: acquiring the instantaneous speed of the train in real time at a preset sampling interval (e.g., 1 second / time) using a train-mounted speed sensor (such as a wheel axle speed sensor or Doppler radar), and converting it into a digital signal to obtain the first operating speed; simultaneously, deploying a data monitoring module in the communication link from the sensor to the ATO controller to transmit the first operating speed, capturing in real time the data packet transmission timestamp, reception timestamp, loss marker, transmission delay value, and channel occupancy rate during the transmission process from the sensor terminal to the ATO controller communication interface, extracting network status data directly related to the transmission of the first operating speed, and integrating it into the first channel communication data; finally, associating and binding the first operating speed with the first channel communication data of the corresponding transmission period through timestamp matching, performing a rationality check on the first operating speed and an integrity check on the first channel communication data, and finally outputting synchronized and valid first operating speed and first channel communication data.

[0059] Step S102: Based on the communication data of the first channel, determine whether the train has been subjected to a DoS attack targeting the transmission of the first set point. If so, compensate the first set point to obtain the second set point, wherein the first set point is transmitted in the communication network.

[0060] In some embodiments, determining whether a DoS attack has occurred on the train targeting a first set point transmission is based on the first channel communication data, wherein the first set point is transmitted in the communication network. Specifically, this includes: extracting network status information related to the first set point transmission from the first channel communication data, comparing the extracted network status information with preset attack judgment criteria, and determining that a DoS attack has occurred targeting the first set point transmission if the network status information meets preset DoS attack characteristic conditions (such as data transmission integrity or timeliness not meeting preset requirements); otherwise, determining that no attack has occurred.

[0061] It should be noted that in the first iteration, the first setpoint is set directly through the preset expected trajectory. In each iteration, the first setpoint is updated for the next iteration. This first setpoint is transmitted in the communication network along with the first running speed, and therefore is vulnerable to DoS attacks during transmission.

[0062] For example, the formula for setting the first set point by pre-setting the desired trajectory in the first iteration is as follows:

[0063] y r (t,0)=y R (t);

[0064] In the formula, y R (t) represents the velocity value at time t in the preset desired trajectory; y r (t,0) is the initial first setpoint.

[0065] For example, considering the randomness and sporadic nature of network attacks, attackers can successfully disrupt data transmission with a certain probability. Therefore, an independent random variable γ1(t,k) is introduced to model the DoS attack. γ1(t,k) follows a Bernoulli distribution, and the probability distribution function describes the action on the first set point y. r DoS attack on (t,k).

[0066] For example, the probability distribution function of DoS is:

[0067]

[0068] In the formula, γ1(t,k)=0 indicates that a DoS attack occurred at time t in the k-th iteration; γ1(t,k)=1 indicates that no DoS attack occurred. This represents the probability that a DoS attack did not occur.

[0069] In some embodiments, the step of compensating the first setpoint to obtain the second setpoint if it exists specifically includes: if a DoS attack is transmitted to the first setpoint, then a compensation gain is applied to the acquired historical setpoint to obtain a DoS compensation value, wherein the historical setpoint is set by a preset expected trajectory; the first setpoint is compensated using the DoS compensation value to obtain the second setpoint. Specifically, if a DoS attack is detected transmitted to the first setpoint, the current iteration stage is first determined to obtain the corresponding historical setpoint: for the first iteration (k=1), since there is no previous iteration data, the historical setpoint is directly determined based on the speed value corresponding to the current running time in the preset expected trajectory; for subsequent iterations (k≥2), the historical setpoint is the setpoint updated during the historical iteration process; then, using the historical setpoint, according to a preset random DoS attack compensation mechanism based on historical data prediction, the DoS compensation value is first calculated, and then the first setpoint is compensated using the DoS compensation value to obtain the second setpoint.

[0070] In some embodiments, the relevant formulas for the random DoS attack compensation mechanism based on historical data prediction specifically include:

[0071]

[0072] In the formula, t is the time index; k is the iteration index; This is the DoS compensation value; This is the second setpoint for generation K-1; This represents the change in the second setpoint for generations k-1 and k-2. K1 represents the change in the second setpoint for generations k-2 and k-3; K2 represents the first-order compensation gain, K1 > 0; K2 represents the second-order compensation gain, K2 > 0; y r (t,k) is the first setpoint of the current iteration; γ1(t,k) represents the probability that a DoS attack occurred at time t in the k-th iteration.

[0073] It should be noted that, in the first iteration, the formula for the random DoS attack compensation mechanism based on historical data prediction above, It can be determined based on a preset expected trajectory.

[0074] When a DoS attack is detected targeting the first setpoint, the DoS compensation value is obtained by using historical setpoints generated from a preset desired trajectory to compensate for the gain. This compensation is then applied to the first setpoint, enabling the rapid generation of a second setpoint unaffected by the DoS attack. This ensures the integrity and reliability of the setpoints transmitted to subsequent steps, providing a correct target reference for generating accurate traction control commands, thereby improving the control accuracy of the vehicle under complex network attacks.

[0075] Step S103: Based on the second setpoint and the first running speed, generate a first traction control command. Combine the discrete nonlinear model and the first traction control command to predict the second running speed of the train at the next moment. Reconstruct the data of the second running speed to obtain a third running speed. Based on the third running speed, determine whether the train has experienced an FDI attack targeting the second running speed. If so, use the third running speed as the fourth running speed after compensating for the FDI attack. Continue until the preset iteration termination condition is met to obtain the target running speed.

[0076] In some embodiments, generating a first traction control command based on the second setpoint and the first operating speed specifically includes: calculating a first tracking error based on the second setpoint and the first operating speed; calculating an error change based on the first tracking error and a second tracking error, wherein the second tracking error is calculated based on the acquired third setpoint and the third operating speed at the previous moment; applying a control gain to the first tracking error and the error change to generate the first traction control command. Specifically, the difference between the second setpoint and the first operating speed is calculated to obtain the first tracking error; the difference between the acquired third setpoint and the third operating speed at the previous moment is calculated to obtain the second tracking error; the difference between the first tracking error and the second tracking error is calculated to obtain the error change; and then a control gain is applied to the first tracking error and the error change to generate the first traction control command.

[0077] In some embodiments, the formula for generating the first traction control command based on the second setpoint and the first operating speed specifically includes:

[0078] The formula for calculating tracking error is:

[0079]

[0080] The control gain formula is:

[0081]

[0082] In the formula, This is the second set point; The first operating speed; This is the first tracking error; The second tracking error is represented by u(t,k), which is the first traction control command; K is the second tracking error. p K is the proportional (P-type) control gain coefficient. d This is the differential (D-type) control gain coefficient.

[0083] It should be noted that the control gain is achieved using a proportional-derivative (PD) feedback controller within the ATO system.

[0084] By calculating the first tracking error based on the second setpoint and the first running speed, and combining it with the second tracking error calculated from the data of the previous moment to calculate the error change, and then performing control gain processing based on the two to generate the first traction control command, the control command can synchronously and accurately respond to the real-time deviation of the speed and its dynamic change trend. This enables the train speed to be quickly and smoothly adjusted initially in a complex attack environment, directly improving the dynamic accuracy of the control.

[0085] For example, the process of constructing a discrete nonlinear model specifically includes:

[0086] First, a dynamic model for a CRH 2-A train operation system is determined:

[0087]

[0088] In the formula, t∈[0,T] is the train travel time; v(t) is the train speed. F is the acceleration of the train; F(t) is the traction force of the train; f b (v(t)) represents the basic resistance per unit mass; f a (v(t)) represents the additional resistance per unit mass of the train in special operating sections (such as curves, slopes, and tunnels); a(t), b(t), and c(t) are the basic resistance coefficients.

[0089] The condition for satisfying the basic drag coefficient is as follows:

[0090] a(t)=[2977+275sin(0.0037t)] / M;

[0091] b(t)=[25.17+2.5sin(0.0037t)] / M;

[0092] c(t)=[0.3864+0.04sin(0.0037t)] / M;

[0093] In the formula, t∈[0,T] is the train travel time; M is the total mass of the train;

[0094] By integrating the iterative index k into the dynamic model of a class of CRH 2-A train operation systems, a high-speed train dynamic model oriented towards iterative learning control is obtained:

[0095]

[0096] In the formula, t∈[0,T] is the train travel time; k∈N represents the number of iterations; v(t,k) is the train speed. Let F(t,k) be the acceleration of the train; F(t,k) be the traction force of the train; f b (v(t,k)) represents the basic resistance per unit mass; f a (v(t,k)) represents the additional resistance per unit mass of the train in special operating sections (such as curves, slopes, and tunnels); a(t), b(t), and c(t) are the basic resistance coefficients.

[0097] Let the sampling interval be t s =1(s), according to The derivative is defined, and the high-speed train dynamics model for iterative learning control is discretized into a discrete dynamic system using the Euler method:

[0098]

[0099] In the formula, the output y(t,k) of the discrete dynamic system is a sampling sequence of the train speed v(t,k); the input u(t,k) of the discrete dynamic system is a sampling sequence of the train traction force F(t,k); a(t), b(t), and c(t) are the basic drag coefficients; f a (y(t,k)) represents the additional resistance per unit mass of the train in special operating sections (such as curves, slopes, and tunnels);

[0100] Let f(y(t,k),u(t,k))=y(t,k)+u(t,k)-[a(t)+b(t)y(t,k)+c(t)y 2 (t,k)]-f a (y(t,k)), then the discrete nonlinear model is obtained:

[0101] y(t+1,k)=f(y(t,k),...,y(tn y ,k),u(t,k),...,u(tn u ,k));

[0102] In the formula, k∈N represents the number of iterations; t∈Z T The length of the train journey is represented by u(t,k)∈R; u(t,k)∈R represents the input traction control command (including the first traction control command obtained in the current iteration); y(t,k)∈R represents the running speed predicted by the discrete nonlinear model, where y(t+1,k) is the second running speed output by the discrete nonlinear model in the current iteration; n y The output order of the discrete nonlinear model is given, where the future output y(t+1,k) depends on the past n... y +1 output at time step; n uThe input order is the discrete nonlinear model, and the future output y(t+1,k) of the model depends on the past n u The input at time +1; f(·) is a nonlinear function.

[0103] For further explanation, please refer to Figure 2 , Figure 2 This application provides a schematic diagram illustrating the variation of additional resistance per unit mass in special train operating sections, showing the additional resistance f per unit mass in special train operating sections (such as curves, slopes, and tunnels). a (v(t)) curve as a function of time.

[0104] In some embodiments, the step of combining the discrete nonlinear model and the first traction control command to predict the second operating speed of the train at the next moment specifically includes: obtaining a preset first speed reference value; inputting the first speed reference value and the first traction control command into the discrete nonlinear model for prediction to obtain the second operating speed of the train at the next moment.

[0105] Specifically, in the first iteration, a preset first speed reference value is obtained first. The first traction control command and the first speed reference value obtained in this iteration are directly input into the discrete nonlinear model to obtain the predicted second running speed of the train at the next moment. In subsequent iterations, the first traction control command obtained in the current iteration (nth generation), the traction control command obtained in the previous iteration (n-1 generations), and the train running speed predicted in the previous iteration (n-1 generations) are all directly input into the discrete nonlinear model to obtain the predicted second running speed of the train at the next moment in the current iteration.

[0106] By inputting a preset first speed reference and a first traction control command into a discrete nonlinear model to construct a correlation prediction logic between traction and speed, the system can stably output the predicted speed value for the next moment, i.e., the second operating speed, in complex network environments. This avoids the loss of reliable basis for speed prediction due to complex network attacks. At the same time, the discrete nonlinear model can adapt to the dynamic characteristics of the train, ensuring the accuracy of speed prediction and helping to improve the train control precision under complex network attacks.

[0107] In some embodiments, data reconstruction is performed on the second operating speed to obtain a third operating speed, specifically including: obtaining a preset second speed reference value and a preset traction force change reference value; and applying an adjustable gain to the second operating speed using the second speed reference value and the traction force change reference value to obtain the third operating speed. Specifically, in the first iteration, the preset second speed reference value and the preset traction force change reference value are first obtained, and then the second operating speed is processed with an adjustable gain using these two data to obtain the third operating speed. In subsequent iterations, the operating speed obtained after data reconstruction in the previous iteration and the traction force change calculated in the previous iteration are obtained, and then the adjustable gain is processed using these two data to obtain the third operating speed.

[0108] In some embodiments, the formula for reconstructing the second operating speed to obtain the third operating speed specifically includes:

[0109] The formula for data reconstruction:

[0110]

[0111] In the formula, K y K3 and K4 are both adjustable gain coefficients; The running speed obtained after data reconstruction in the previous iteration; Δu(t,k-1) is the pseudo-partial derivative estimate for the (k-1)th iteration at time t; Δu(t,k-1) is the change in traction force calculated in the previous iteration. and y R (t+1) represents the velocity value corresponding to the preset expected trajectory at the next moment.

[0112] It should be noted that if this is the first iteration, then in the above data reconstruction formula, The second speed reference value is preset, and Δu(t,k-1) is the preset traction force change reference value. and Specific values ​​can be preset as needed.

[0113] By using a preset second speed reference value and a preset traction force change reference value, the second operating speed can be reconstructed with adjustable gain, which can quickly obtain speed data close to the real state. The adjustable gain can flexibly adapt to the train's operating characteristics, ensuring that the reconstruction process is accurate and controllable, providing a reliable basis for subsequent train control, and effectively improving the control accuracy of trains under complex network attacks.

[0114] In some embodiments, determining whether the train has been subjected to an FDI attack targeting the transmission of the second operating speed based on the third operating speed specifically includes: calculating a speed difference based on the third operating speed and a fifth operating speed, wherein the fifth operating speed is the actual received value of the second operating speed after transmission through the communication network; if the speed difference is greater than a preset detection threshold, it is determined that the train has been subjected to the FDI attack. Specifically, an independent random variable γ2(t,k) is first introduced, and an FDI attack is modeled using a probability distribution function. The second operating speed of the train predicted in the current iteration of the discrete nonlinear model is transmitted through the communication network. The actual received value after transmission is recorded as the fifth operating speed. The speed difference is calculated between the third operating speed obtained after data reconstruction of the second operating speed and the fifth operating speed that may be subjected to an FDI attack after transmission through the communication network. If the speed difference is greater than a preset detection threshold, it indicates that the train has been subjected to an FDI attack targeting the transmission of the second operating speed during transmission; if it is less than or equal to the preset threshold, it indicates that the train has not been subjected to an FDI attack.

[0115] It should be noted that γ2(t,k) follows a Bernoulli distribution.

[0116] In some embodiments, the formula for determining whether an FDI attack targeting the second operating speed has occurred based on the third operating speed specifically includes:

[0117] The expression for the fifth running speed is:

[0118]

[0119] In the formula, y(t+1,k) is the second running speed of the train at the next moment predicted by the discrete nonlinear model; γ2(t,k)=0 indicates that no FDI attack occurred at time t in the k-th iteration, and γ2(t,k)=1 indicates that an FDI attack occurred; D(t+1,k) is the FDI attack signal.

[0120] The probability distribution function of an FDI attack is:

[0121]

[0122] In the formula, Let γ2(t,k) = 0 represent the probability of an FDI attack occurring, where γ2(t,k) = 0 indicates that no FDI attack occurred at time t in the k-th iteration, and γ2(t,k) = 1 indicates that an FDI attack occurred.

[0123] By comparing the difference between the reconstructed third operating speed and the transmitted fifth operating speed, FDI attacks can quickly and accurately identify tampering with speed data. The preset detection threshold provides a clear quantitative standard for attack judgment, which can effectively avoid misjudgment of attacks caused by transmission fluctuations or environmental interference, ensure timely identification of FDI attacks, prevent false speed data from entering the control loop, and significantly improve the reliability and accuracy of train control under complex network attacks.

[0124] In some embodiments, if present, the third operating speed is used as the fourth operating speed after compensation for the FDI attack, until a preset iteration termination condition is met to obtain the target operating speed. Specifically, if an FDI attack targeting the transmission of the second operating speed is detected, the fifth operating speed attacked during the transmission is discarded, and the third operating speed obtained by data reconstruction is used as the fourth operating speed after compensation for the FDI attack. If the preset iteration termination condition has been met (e.g., the number of iterations reaches 50), this fourth operating speed is directly used as the final target operating speed. If the preset iteration termination condition is not met, the fourth operating speed is used as the first operating speed in the next iteration to update the first setpoint in the next round of iterations, and the above steps S101 to S103 are continued to be iterated until the preset iteration termination condition is met. Then, the operating speed after compensation for the FDI attack in the last iteration is used as the final target operating speed.

[0125] For example, if there is no FDI attack targeting the second running speed, the fifth running speed that was not attacked during the transmission is directly used as the final running speed obtained in this iteration. If this iteration is the last iteration, this fifth running speed is used as the target running speed.

[0126] In some embodiments, if present, the third running speed is used as the fourth running speed after compensation for the FDI attack, until a preset iteration termination condition is met, to obtain the relevant formula for the target running speed, specifically including:

[0127] Fourth operating speed:

[0128]

[0129] In the formula, γ2(t,k)=0 indicates that no FDI attack occurred at time t in the k-th iteration, and γ2(t,k)=1 indicates that an FDI attack occurred; y(t+1,k) is the fifth running speed during the transmission process; This represents the fourth running speed in the current iteration; y p (t+1,k) represents the third running speed after data reconstruction.

[0130] For example, in each iteration, the parameters are updated, specifically including:

[0131] Discrete nonlinear models belong to SISO nonlinear nonaffine systems. After the current iteration ends, based on the running speed reconstructed by the data, the tracking problem of the SISO nonlinear nonaffine system can be transformed into the following system:

[0132]

[0133] In the formula, u(t,k)∈R and respectively This represents the input traction control command and the reconstructed operating speed after being attacked by FDI during transmission; k∈N represents the number of iterations; t∈Z T Indicates the length of train travel time; n y Output the order of the model; n u Input the order of the model, n u ,n y ∈N; f(·) is a nonlinear function;

[0134] Then, four hypotheses are proposed for the system. Hypothesis 1 is that the initial output of each iteration is bounded, i.e. |y(0,k)|≤y0, where y0 is a positive constant. Assumption 2 states that the partial derivatives of the nonlinear function f(·) with respect to the control input u(t,k) are continuous. Assumption 3 states that the model satisfies the generalized Lipschitz condition, i.e., for t∈Z... T And k∈N, if Δu(t,k)≠0, then It is a finite positive number. Assume that 4 is the FDI attack signal D(t,k) and it is bounded.

[0135] If the general form of a discrete nonlinear dynamic system satisfies assumptions 2 and 3, then when Δu(t,k)≠0, and The IDL data model can be obtained:

[0136]

[0137] Δu(t,k)=u(t,k)-u(t,k-1);

[0138] In the formula, k∈N represents the number of iterations; t∈Z T This represents the train's travel time; Δu(t,k) represents the change in traction force. The running speed after data reconstruction in the current iteration; The running speed after data reconstruction in the previous iteration; u(t,k) is the traction control command obtained in the current iteration; u(t,k-1) is the traction control command obtained in the previous iteration; Let be the pseudo-partial derivative at time t for the k-th iteration. Bounded;

[0139] The algorithm for estimating pseudo-partial derivatives is as follows:

[0140]

[0141] In the formula, η1∈[0,1] is the pseudo-partial derivative estimate from the previous iteration; Δu(t,k-1) is the step size factor; μ1>0 is the change in traction force obtained from the previous iteration; μ1>0 is the weighting factor. This represents the difference in running speed obtained after reconstruction in the previous iteration;

[0142] Update the first setting point:

[0143]

[0144] In the formula, The setpoint is the compensated point from the previous iteration; θ(t,k) is the adaptive parameter; α is the adjustable parameter; y R (t+1) represents the velocity value corresponding to the preset desired trajectory at the next moment; The running speed after data reconstruction in the previous iteration;

[0145] Update adaptive parameters:

[0146]

[0147] In the formula, θ(t,k-1) is the adaptive parameter in the previous iteration; η2∈[0,1] is the step size factor; μ2 is the weight factor; K p For P-type control gain; K d D-type controlled gain; This is the estimated value of the pseudo-partial derivative obtained in this iteration; α represents the change in tracking error at the setpoint calculated in the previous iteration; α is an adjustable parameter. The running speed after data reconstruction in the current iteration; The running speed after data reconstruction in the previous iteration;

[0148] If |θ(t,k)|≤ε2 or sign(θ(t,k))≠sign(θ(t,0)), the adaptive parameter reset algorithm is as follows to ensure that the parameter θ(t,k) can be updated normally:

[0149] θ(t,k)=θ(t,0);

[0150] In the formula, θ(t,0) is the initial value of θ(t,k);

[0151] if Or |Δu(t,k)|≤ε1, pseudo-partial derivative estimate The reset algorithm is as follows:

[0152]

[0153] In the formula, yes The initial value.

[0154] For example, if the general form of the discrete nonlinear model satisfies the assumptions 1-3 mentioned above, and the parameter α is chosen appropriately to satisfy the following convergence condition:

[0155] sign(α) = sign(θ(t,0));

[0156]

[0157] In the formula, For pseudo-partial derivative estimates The upper bound of d; θ K is the upper bound of the adaptive parameter θ(t,k); p For P-type control gain; K d The gain is a D-type control; θ(t,0) is the initial value of the adaptive parameter θ(t,k);

[0158] The tracking error in the general form of the discrete nonlinear model

[0159] For example, please refer to Figure 3 , Figure 3 This is a schematic diagram of the expected trajectory of the discrete nonlinear model output under the convergence condition provided in this application. According to the convergence condition, if an appropriate α = 0.03 is chosen, the initial value of the adaptive parameter θ(t,0) = 1, and the initial pseudo-partial derivative estimate is... Assume that at any time t, the probability of a DoS attack occurring is... The probability of an FDI attack occurring is Both types of attacks occur according to independent Bernoulli distributions. At runtime t∈[0,1,…,3400], the expected trajectory output by the discrete nonlinear model is as follows: Figure 3 As shown.

[0160] Step S104: Control the train based on the target operating speed;

[0161] In some embodiments, controlling the train based on the target operating speed specifically includes: sending the obtained target operating speed as a control command of the train ATO system to the train traction and braking execution unit; the execution unit dynamically adjusts the output of traction or braking force according to the deviation between the target operating speed and the current actual speed, thereby achieving precise speed control of the train.

[0162] For further explanation, please refer to Figure 4 , Figure 4 This is an iterative flowchart of the train speed control method against composite network attacks provided in this application. Specifically, it includes: first, updating the first setpoint; then, executing step S102, determining whether the first setpoint transmission has encountered a DoS attack; if so, performing DoS attack compensation; otherwise, no compensation is needed, thus obtaining the first setpoint (if attacked, it is the compensated setpoint; if not attacked, it is the original setpoint); then, executing step S103, calculating the control input (i.e., the first traction control command) based on the first setpoint, and then inputting the first traction control command into the high-speed train ATO system. The system outputs a discrete nonlinear model (i.e., the train's speed predicted by the model for the next moment). Then, it performs FDI attack detection on the transmission of the system output. That is, it first reconstructs the system output data, and then judges it by a preset detection threshold. If an FDI attack exists, the reconstructed data is used to compensate the system output that was attacked during transmission, resulting in a compensated system output. This compensated system output is then used to update the first setpoint for the next iteration, thus realizing the iterative process. The compensated system output obtained in the last iteration is used as the target speed to control the train.

[0163] By acquiring the train's initial operating speed and the first channel communication data monitored during its transmission through the communication network in real time, this provides accurate and reliable physical state input and raw network state information for subsequent control decisions, improving the baseline accuracy of the control system from the source. Based on the first channel communication data, it determines whether the train is experiencing a DoS attack targeting the first setpoint in the communication network transmission, and compensates for the affected first setpoint to obtain a second setpoint. This effectively resists control command interruptions caused by packet loss, providing a foundation for maintaining control accuracy. Based on the second setpoint and the first operating speed, traction control commands are generated, ensuring that the traction control commands closely match the current operating state and the compensated target, directly contributing to improved control accuracy. Finally, by combining the traction control commands and a discrete nonlinear model, the train's second operating speed at the next moment is predicted. The system reconstructs the second operating speed to obtain a third operating speed. Based on this third operating speed, it determines whether an FDI attack targeting the second operating speed exists. If so, the reconstructed third operating speed is used as the fourth operating speed after compensating for the FDI attack. This effectively identifies and filters maliciously tampered speed information in the feedback loop, fundamentally ensuring the authenticity of the feedback information and improving the control accuracy of trains under composite network attacks. After multiple iterations of the preceding operations, the target operating speed is obtained. Iterative optimization can continuously correct speed data deviations caused by composite network attacks, forming a high-precision speed control benchmark adapted to the attack scenario. Finally, the train is controlled according to the target operating speed, ensuring that the final executed command accurately drives the train to converge to the correct speed state, effectively improving the control accuracy of trains under composite network attacks. This application can improve the control accuracy of trains under composite network attacks.

[0164] See Figure 5 Based on the above method embodiments, corresponding device embodiments are provided;

[0165] An embodiment of the present invention provides a train control device for complex network attacks, characterized in that it includes a first module 100, a second module 200, a third module 300 and a fourth module 400;

[0166] The first module 100 is used to acquire the first operating speed of the train and the first channel communication data in real time, wherein the first channel communication data is obtained by monitoring the first operating speed during the transmission process of the communication network;

[0167] The second module 200 is used to determine whether the train has been subjected to a DoS attack against the transmission of the first set point based on the communication data of the first channel. If so, the first set point is compensated to obtain a second set point, wherein the first set point is transmitted in the communication network.

[0168] The third module 300 is used to generate a first traction control command based on the second set point and the first running speed, predict the second running speed of the train at the next moment by combining the discrete nonlinear model and the first traction control command, reconstruct the data of the second running speed to obtain a third running speed, determine whether the train has been subjected to an FDI attack targeting the second running speed based on the third running speed, and if so, use the third running speed as the fourth running speed after compensating for the FDI attack, until the preset iteration termination condition is met to obtain the target running speed;

[0169] The fourth module 400 is used to control the train based on the target operating speed.

[0170] In this way, the first module acquires the train's first operating speed and the first channel communication data monitored during its transmission through the communication network in real time, providing real and reliable physical state input and original network state information for subsequent control decisions, thereby improving the baseline accuracy of the control system from the source. The second module determines whether the train is experiencing a DoS attack targeting the first setpoint in the communication network transmission based on the first channel communication data, and compensates for the affected first setpoint to obtain a second setpoint, which can effectively resist the interruption of control commands due to data packet loss, providing a foundation for maintaining control accuracy. The third module generates traction control commands based on the second setpoint and the first operating speed, making the traction control commands closely aligned with the current operating state and the compensated target, directly contributing to improving control accuracy. Combining the traction control commands and discrete nonlinear models, the second speed of the train at the next moment is predicted. The system first calculates the train's operating speed and then reconstructs the data from the second operating speed to obtain a third operating speed. Based on this third operating speed, it determines whether an FDI attack targeting the second operating speed exists. If so, the reconstructed third operating speed is used as the fourth operating speed after compensating for the FDI attack. This effectively identifies and filters out maliciously tampered speed information in the feedback loop, fundamentally ensuring the authenticity of the feedback information and improving the control accuracy of trains under composite network attacks. After multiple iterations of the preceding operations, the target operating speed is obtained. Iterative optimization can continuously correct speed data deviations caused by composite network attacks, forming a high-precision speed control benchmark adapted to the attack scenario. Finally, the fourth module controls the train according to the target operating speed, ensuring that the final executed command accurately drives the train to converge to the correct speed state, effectively improving the control accuracy of trains under composite network attacks.

[0171] It is understood that the above-described device embodiments correspond to the method embodiments of the present invention, and can implement the train control method against complex network attacks provided by any of the above-described method embodiments of the present invention.

[0172] It should be noted that the device embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can specifically be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0173] Based on the above-described embodiment of a train control method for a complex network attack, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements a train control method for a complex network attack according to any embodiment of the present invention.

[0174] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the terminal device.

[0175] The terminal device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0176] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0177] Based on the above-described method embodiments, another embodiment of the present invention provides a computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute a train control method for a complex network attack as described in any of the above-described method embodiments of the present invention.

[0178] The modules / units integrated in the device / terminal equipment, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0179] Based on the above-described method embodiments, another embodiment of the present invention provides a computer program product, including a computer program or instructions, which, when executed by a communication device, implements a train control method for a complex network attack according to any embodiment of the present invention.

[0180] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A train control method against a complex network attack, characterized in that, include: The train's first operating speed and first channel communication data are acquired in real time, wherein the first channel communication data is obtained by monitoring the first operating speed during transmission over the communication network; Based on the communication data of the first channel, it is determined whether the train has been subjected to a DoS attack targeting the transmission of the first set point. If so, the first set point is compensated to obtain the second set point, wherein the first set point is transmitted in the communication network. Based on the second setpoint and the first operating speed, a first traction control command is generated. Combining the discrete nonlinear model and the first traction control command, the second operating speed of the train at the next moment is predicted. The second operating speed is reconstructed to obtain a third operating speed. Based on the third operating speed, it is determined whether the train has experienced an FDI attack targeting the second operating speed. If so, the third operating speed is used as the fourth operating speed after compensation for the FDI attack. This process continues until a preset iteration termination condition is met to obtain the target operating speed. The train is controlled based on the target operating speed.

2. The train control method against composite network attacks as described in claim 1, characterized in that, If the condition exists, the first set point is compensated to obtain the second set point, specifically including: If a DoS attack is transmitted to the first set point, the acquired historical set point is compensated with a gain to obtain a DoS compensation value, wherein the historical set point is set by a preset expected trajectory. The first setpoint is compensated using the DoS compensation value to obtain the second setpoint.

3. The train control method against composite network attacks as described in claim 1, characterized in that, The generation of the first traction control command based on the second setpoint and the first operating speed specifically includes: The first tracking error is calculated based on the second setpoint and the first running speed; Based on the first tracking error and the second tracking error, the error change is calculated, wherein the second tracking error is calculated based on the third set point and the third running speed of the previous moment. The first tracking error and the error change amount are controlled by gain to generate the first traction control command.

4. The train control method against composite network attacks as described in claim 1, characterized in that, The method of combining the discrete nonlinear model and the first traction control command to predict the train's second operating speed at the next moment specifically includes: Obtain the preset first speed reference value; Based on the first traction control command and the second traction control command, the change in the first traction force is calculated; The first change in traction force and the first speed reference value are input into the discrete nonlinear model for prediction to obtain the second operating speed of the train at the next moment.

5. The train control method against composite network attacks as described in claim 1, characterized in that, The process of reconstructing the data from the second operating speed to obtain the third operating speed specifically includes: Obtain the preset second speed reference value and the preset traction force change reference value; Using the second speed reference value and the traction force change reference value, the second operating speed is adjusted with an adjustable gain to obtain the third operating speed.

6. The train control method against composite network attacks as described in claim 1, characterized in that, The determination of whether the train is experiencing an FDI attack targeting the second operating speed based on the third operating speed specifically includes: Based on the third operating speed and the fifth operating speed, the speed difference is calculated, wherein the fifth operating speed is the actual received value of the second operating speed after being transmitted through the communication network; If the speed difference is greater than a preset detection threshold, the train is determined to have been subjected to the FDI attack.

7. A train control device for composite network attacks, characterized in that, It includes Module 1, Module 2, Module 3, and Module 4; The first module is used to acquire the first operating speed of the train and the first channel communication data in real time, wherein the first channel communication data is obtained by monitoring the first operating speed during the transmission process of the communication network; The second module is used to determine whether the train has been subjected to a DoS attack against the transmission of the first set point based on the communication data of the first channel. If so, the first set point is compensated to obtain a second set point, wherein the first set point is transmitted in the communication network. The third module is used to generate a first traction control command based on the second setpoint and the first operating speed, and predict the second operating speed of the train at the next moment by combining the discrete nonlinear model and the first traction control command. The second operating speed is then reconstructed to obtain a third operating speed. Based on the third operating speed, it is determined whether the train has experienced an FDI attack targeting the second operating speed. If so, the third operating speed is used as the fourth operating speed after compensation for the FDI attack, until a preset iteration termination condition is met to obtain the target operating speed. The fourth module is used to control the train based on the target operating speed.

8. A terminal device, characterized in that, include: The processor, memory, communication interface, and communication bus are provided, wherein the processor, memory, and communication interface communicate with each other via the communication bus. The memory is used to store at least one executable instruction that causes the processor to perform the operation of a train control method for a complex network attack as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device or apparatus containing the computer-readable storage medium to perform a train control method for a complex network attack as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by the communication device, they implement a train control method for a complex network attack as described in any one of claims 1 to 6.

Images