A heavy-duty vehicle steer-by-wire system information security control method

Abstract

The application discloses a heavy-load vehicle steer-by-wire system information security control method, which comprises the following steps: establishing a networked control architecture and a dynamics model of the heavy-load vehicle steer-by-wire system; establishing a double-channel communication interruption model under network attack of the heavy-load vehicle steer-by-wire system, and establishing a sensor-controller channel sampling output model; obtaining a system observation state under network attack; designing a robust safety controller based on Tube-MPC to generate a prediction control sequence; sending the prediction control sequence to a steer-by-wire execution mechanism, and designing a control sequence cache compensation mechanism of the steer-by-wire execution mechanism; selecting a corresponding control amount according to the control sequence cache compensation mechanism, and applying the control amount to an actual steering motor to realize angle tracking control of the heavy-load vehicle steer-by-wire system. The application reduces the risk of state estimation error amplification, control input interruption and angle tracking failure, and improves the safe operation ability of the heavy-load vehicle steer-by-wire system under network attack.

Description

Technical Field

[0001] This invention belongs to the field of information security control technology for steer-by-wire systems of heavy-duty vehicles, and specifically relates to an information security control method for steer-by-wire systems of heavy-duty vehicles. Background Technology

[0002] Steer-by-wire systems eliminate the mechanical connection between the steering wheel and the steering actuator. They utilize sensors, controllers, communication networks, and the steering actuator to acquire steering commands, make control decisions, and execute steering angles, providing a technological foundation for trajectory tracking, stability control, and intelligent driving in heavy-duty vehicles. However, steer-by-wire systems are highly dependent on the vehicle's communication network. The sensor-controller and controller-actuator channels can be affected by communication interruptions, data loss, network congestion, or malicious attacks, leading to delays in the transmission of steering status information or control commands, thus impacting vehicle safety.

[0003] In the prior art, there are already relevant solutions for the safety and control performance of steer-by-wire systems. For example, Chinese invention patent application No. CN202080004477.0, entitled "Steering-by-wire system and steering control method," discloses a method to improve the operational safety problem caused by the inconsistency between the steering wheel rotation range and the steering wheel rotation range by setting a steering column limiting device; Chinese invention patent application No. CN201310649410.4, entitled "A steer-by-wire system and control method for electric vehicles," discloses a method to achieve variable transmission ratio, vehicle lateral stability control, and road feel optimization through a steering wheel force feedback module and a front wheel steering module. The above methods mainly improve the performance of steer-by-wire systems from the perspectives of mechanical limiting, force feedback, and front wheel angle control, but they do not adequately consider the problems of missing sampling output and lost control commands caused by denial-of-service attacks.

[0004] Denial-of-service attacks can prevent control systems from acquiring data or sending commands in a timely manner by blocking communication resources or consuming network bandwidth. For networked steering-by-wire systems, when the sensor-controller channel is attacked, the controller struggles to obtain real-time steering status, easily leading to increased state estimation errors. When the controller-actuator channel is attacked, the steering actuator cannot obtain current control commands, easily causing interruptions in control input updates and a decline in corner tracking performance. If both types of channel attacks occur simultaneously or alternately, traditional control methods that rely on continuous sampling and continuous control command transmission cannot guarantee closed-loop security.

[0005] For heavy-duty vehicles, the large vehicle mass, high front axle load, and large steering actuator inertia mean that the steering system needs to overcome greater tire return torque, kingpin drag torque, and mechanism friction torque. Simultaneously, the high center of gravity and load transfer cause time-varying uncertainties in tire vertical load, road surface adhesion utilization, and steering resistance. These disturbances, when caused by denial-of-service attacks resulting in lost sampling information or interrupted control commands, will further amplify state estimation errors, angle tracking errors, and the risk of abrupt changes in actuator control input.

[0006] Therefore, it is necessary to propose an information security control method for the steer-by-wire system of heavy-duty vehicles. Based on considering the sources of disturbance in heavy-duty vehicles, a safety observation and robust control mechanism should be established to address the loss of state information and control commands. Summary of the Invention

[0007] In view of the shortcomings of the prior art, the purpose of this invention is to provide an information security control method for a heavy-duty vehicle steer-by-wire system, so as to solve the technical problems of interruption of state information acquisition, failure of control command transmission, significant steering disturbance of heavy-duty vehicles, and difficulty in guaranteeing the security of closed-loop angle tracking under network denial-of-service attacks.

[0008] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0009] The present invention provides an information security control method for a steer-by-wire system for heavy-duty vehicles, comprising the following steps:

[0010] 1) Establish a networked control architecture and dynamic model for the steer-by-wire system of heavy-duty vehicles;

[0011] 2) Establish a dual-channel communication interruption model for the steer-by-wire system of heavy-duty vehicles under network attacks, and establish a sensor-controller channel sampling output model;

[0012] 3) Design a security status observer to obtain the system's observed status under network attacks;

[0013] 4) Design a robust safety controller based on Tube-MPC to generate predictive control sequences;

[0014] 5) Send the predictive control sequence obtained in step 4) to the steer-by-wire actuator, and design a control sequence buffer compensation mechanism for the steer-by-wire actuator;

[0015] 6) Select the corresponding control quantity according to the control sequence buffer compensation mechanism in step 5) and apply it to the actual steering motor to realize the angle tracking control of the steer-by-wire system of heavy-duty vehicles.

[0016] Further, step 1) specifically includes:

[0017] 11) Establish a networked control architecture for the steer-by-wire system of heavy-duty vehicles. The networked control architecture includes: a steering angle sensor, an electronic control unit, a CAN communication network, and a steer-by-wire actuator. The steer-by-wire actuator includes: a permanent magnet synchronous steering motor, a reduction mechanism, a rack and pinion mechanism, and a front wheel steering mechanism. The steering angle sensor sends a steering angle signal to the electronic control unit through the sensor-controller channel in the CAN communication network. The electronic control unit sends a steering control command to the steer-by-wire actuator through the controller-actuator channel in the CAN communication network. The permanent magnet synchronous steering motor generates a steering driving torque according to the steering control command. The steering driving torque is transmitted to the rack and pinion mechanism through the reduction mechanism, and the rack and pinion mechanism drives the front wheel steering mechanism to rotate through the steering tie rod.

[0018] 12) Establish a dynamic model of the steer-by-wire system for heavy-duty vehicles. The dynamic model includes: steering motor voltage equations, steering motor dynamic equations, rack and pinion dynamic equations, and front wheel dynamic equations; wherein,

[0019] The voltage equation for the steering motor is as follows:

[0020] ;

[0021] The dynamic equations of the steering motor are as follows:

[0022] ;

[0023] The dynamic equations of the gear and rack are as follows:

[0024] ;

[0025] The front wheel dynamics equations are as follows:

[0026] ;

[0027] In the formula, This represents the equivalent control current of the steering motor. This represents the rate of change of the equivalent control current of the steering motor. This represents the equivalent control voltage of the steering motor. This represents the equivalent resistance of the steering motor. This represents the equivalent inductance of the steering motor. This represents the back electromotive force coefficient of the steering motor. Indicates the angle of the steering motor shaft. Indicates the angular velocity of the steering motor shaft. This indicates the angular acceleration of the steering motor shaft. This represents the moment of inertia of the steering motor. This represents the coefficient of viscous friction of the steering motor. This indicates the torque coefficient of the steering motor. This indicates the load torque at the shaft end of the steering motor. Indicates the displacement of the steering rack. Indicates the speed of the steering rack. Indicates the acceleration of the steering rack. Indicates the equivalent mass of the steering rack. This represents the equivalent damping coefficient of the gear and rack mechanism. Indicates the rack and pinion load torque. Indicates the gear ratio of the reduction mechanism. This indicates the radius of the pinion in a gear and rack mechanism. Indicates the front wheel steering angle. Indicates the steering angular velocity of the front wheels. This represents the front wheel steering angular acceleration. This represents the equivalent moment of inertia of the front wheel steering mechanism. This represents the equivalent viscous friction coefficient of the front wheel steering mechanism. Indicates the steering tie rod transmission ratio. This indicates the tire return torque;

[0028] 13) Establish the expressions for the load torque at the shaft end of the steering motor and the load torque of the rack;

[0029] The load torque at the end of the steering motor shaft is determined by the relative deformation between the steering motor shaft angle and the rack displacement obtained through equivalent conversion by the reduction mechanism, and is expressed as:

[0030] ;

[0031] The rack load torque is determined by the relative deformation between the steering rack displacement and the front wheel steering angle obtained through equivalent conversion via the steering tie rod, and is expressed as:

[0032] ;

[0033] In the formula, This indicates the equivalent torsional stiffness at the end of the steering motor shaft. Indicates the equivalent rack stiffness of a gear and rack mechanism;

[0034] 14) Define the system state, control input, and tire return torque input; let Representing continuous time, the system state vector, control input, and tire self-aligning torque input are respectively expressed as:

[0035] ;

[0036] ;

[0037] ;

[0038] In the formula, Representing continuous time The system state vector under the following conditions Representing continuous time The steering motor control voltage input is below. Representing continuous time Input the tire return torque;

[0039] Select front wheel steering angle As the measurement output, a continuous state-space model is established as follows:

[0040] ;

[0041] ;

[0042] In the formula, Represents the system state vector in continuous time. The first derivative, Representing continuous time The measurement output below, Indicates external disturbance. Indicates measurement noise; the external disturbance This includes equivalent disturbances caused by the large overall mass of heavy-duty vehicles, high front axle load, large inertia of steering actuators, changes in tire return torque, changes in kingpin resistance torque, gear and rack friction, load transfer, and changes in road surface adhesion.

[0043] ;

[0044] ; ; ;

[0045] In the formula, This represents the state matrix of the steering-by-wire system. This represents the steering motor control input matrix. This represents the tire return torque input matrix. This represents the output matrix of the front wheel steering angle measurement. This represents the external perturbation input matrix. Represents the measurement noise input matrix;

[0046] 15) Discretize the continuous state-space model established in step 14) to obtain the discrete dynamic model required for controller design; let If we denote the discrete sampling time, then the discrete dynamic model can be expressed as:

[0047] ;

[0048] ;

[0049] In the formula, Indicates the first The system state vector at each sampling time. Indicates the first The system state vector at each sampling time. Indicates the first Steering motor control voltage input at each sampling time, Indicates the first The tire return torque input at each sampling time. Indicates the first External disturbances at each sampling time Indicates the first Measurement noise at each sampling time, Indicates the first Measurement output at each sampling time, Represents the state matrix of a discrete system. This represents the discrete steering motor control input matrix. This represents the discrete tire return torque input matrix. This represents the discrete external disturbance input matrix. Represents the discrete measurement output matrix. Represents the discrete measurement noise input matrix; , , , , , Indicates the controller sampling period. .

[0050] Furthermore, step 2) of establishing a dual-channel communication interruption model under a network attack on the steer-by-wire system of a heavy-duty vehicle specifically includes:

[0051] The impact of network attacks on the dual-channel signal transmission process is analyzed, and a model is developed to address the communication interruption problem caused by denial-of-service attacks, resulting in a dual-channel communication interruption model.

[0052] The dual-channel communication interruption model includes: a communication interruption state function, a communication interruption frequency boundary, and a communication interruption duration boundary;

[0053] Define communication interruption status function ,in This indicates that the communication link is in an interrupted state due to a denial-of-service attack. This indicates that the communication link is in a normal communication state. The function represents the normal communication state, and ;when At that time, the communication link was normal; when At that time, the communication link was interrupted;

[0054] Establish the communication interruption frequency boundary in the dual-channel communication interruption model: for any time interval , This indicates the start time of the time interval. This represents the end time of the time interval, and satisfies... ;set up This indicates the number of communication interruptions that occurred within the given time interval. The number of communication interruptions is defined by the following formula:

[0055] ;

[0056] In the formula, This represents the communication interruption frequency offset constant. This represents the lower bound of the average communication interruption interval; thus, it constrains the frequency of dual-channel communication interruptions per unit time.

[0057] Establish the communication interruption duration boundary in the dual-channel communication interruption model: Let Indicates time interval Given the set of durations of internal communication interruptions, the cumulative duration of communication interruptions is limited by the following formula:

[0058] ;

[0059] In the formula, This represents a constant indicating the duration offset of the communication interruption. This represents the average communication interruption duration constraint parameter; thus, it constrains the cumulative duration of dual-channel communication interruptions within a given time interval.

[0060] Furthermore, step 2) of establishing the sensor-controller channel sampling output model specifically includes:

[0061] set up The sensor-controller channel sampling output model under the influence of communication interruption caused by a denial-of-service attack at the sampling time is expressed as follows:

[0062] ;

[0063] In the formula, Indicates the controller at the sampling time The actual sampled output, This represents the output matrix of the front wheel steering angle measurement. Indicates the system status. Indicates the measurement noise term; when hour, The sensor-controller channel is in a communication interruption state, and the controller cannot obtain the current valid steering angle sampling output; when hour, The sensor-controller channel is in normal communication mode, and the controller obtains normal steering angle sampling output.

[0064] Furthermore, step 3) specifically includes:

[0065] 31) Based on the dynamic model established in step 1) and the sensor-controller channel sampling output model established in step 2), design a safety state observer to reconstruct the system state under conditions of sampling output loss, measurement noise, and steering disturbances from heavy-duty vehicles; the safety state observer is represented as:

[0066] ;

[0067] In the formula, Indicates continuous time. Indicates the sampling time. Representing continuous time The system's observed state vector is as follows. Represents the system's observed state vector in continuous time. The first derivative, Indicates the sampling time The corresponding system observation state vector, Indicates the sampling time The corresponding output estimate, Indicates the controller at the sampling time The sampling output obtained through the sensor-controller channel This represents the gain matrix of the safety state observer;

[0068] 32) Define the state estimation error : and define the sampling delay : Combining the sensor-controller channel sampling output model under the influence of communication interruption caused by a denial-of-service attack, the system equation for the state estimation error is obtained as follows:

[0069] ;

[0070] In the formula, Indicates the sampling interval. This indicates steering disturbance of a heavily loaded vehicle. This represents the state estimation error corresponding to the delayed sampling time. Represents the measurement noise input matrix. This represents the measurement noise corresponding to the delayed sampling time.

[0071] 33) Solve for the observer gain matrix using the Lyapunov method; if a symmetric positive definite matrix exists... , Invertible matrix ,matrix Given the associated relaxation matrix and positive scalar, if the linear matrix inequality holds, then the estimation error system is uniformly and eventually bounded. The linear matrix inequality is as follows:

[0072] ;

[0073] ;

[0074] ;

[0075] in, This represents the transpose term in a symmetric matrix, determined by symmetry, where the lower triangular partition of the matrix is ​​equal to the transpose of the upper triangular partition. The matrix elements in the linear matrix inequality are defined as follows:

[0076] , , , , , , , , , , , , , , , , , , , , , , , , ;

[0077] In the formula, Representation matrix Symmetric operators, Let represent a matching matrix of arbitrary dimensions, and , and It is a symmetric positive definite matrix. It is an invertible matrix. , , , , , , , and Let the variables be the matrix to be solved. , , , , and It is a positive scalar;

[0078] 34) Based on the solution of the linear matrix inequality, the gain matrix of the safety state observer is obtained as follows:

[0079] ;

[0080] In the formula, Used to determine the feedback effect of the sampling output deviation in the safety status observer in step 31) on the observation status update;

[0081] 35) The gain matrix of the safety state observer obtained in step 34) Substitute the security status observer established in step 31) to obtain the system observation status under network attack online. ;

[0082] When the sensor-controller channel is in normal communication mode, the controller outputs based on the current sampling. With output estimate The deviation between the observation states is corrected; when the sensor-controller channel is attacked by a denial-of-service attack and the sampling output is interrupted, the controller recursively updates the observation state based on the dynamic model of the heavy-duty vehicle's steer-by-wire system, historical control inputs, and the observation state at the previous moment.

[0083] Further, step 4) specifically includes:

[0084] 41) Based on the discrete dynamics model established in step 15), the dual-channel communication interruption model established in step 2), and the system observation state under network attack obtained in step 3), design a robust security control algorithm based on Tube-MPC; the robust security control algorithm calculates the optimal control sequence in the future prediction time domain during each rolling optimization.

[0085] 42) Establish the equations of the controlled system under a controller-actuator channel attack; set the actuator at time... The actual control quantity used is The equation of the controlled system is then expressed as:

[0086] ;

[0087] 43) Establish a nominal prediction system: Set the nominal state as The nominal output is The nominal control quantity is The nominal system equations are then expressed as:

[0088] ;

[0089] In the formula, The system observation state obtained in step 3) For the first The nominal state at each sampling time. This indicates the nominal tire return torque input;

[0090] 44) Define the error between the output of the controlled system and the output of the nominal system. , The Tube-MPC control law is constructed as follows:

[0091] ;

[0092] In the formula, Represents the error feedback control matrix;

[0093] 45) Based on the actual controlled system and the nominal system, the error dynamics are obtained as follows:

[0094] ;

[0095] 46) Define the closed-loop state transition matrix of the error equation. Perturbation input matrix and define lumped disturbance ,in , , Indicates the first The lumped perturbation vector at each sampling time. This represents the deviation between the nominal control quantity and the actual control quantity used by the actuator. The deviation between the nominal tire return torque input and the actual tire return torque input can be represented by the following dynamic error formula:

[0096] ;

[0097] 47) Constructing a lumped disturbance set and robust positive invariant set ;like make If it is stable, then there exists a robust positive invariant set. The following conditions must be met:

[0098] ;

[0099] 48) Tighten the output constraints and control input constraints based on the robust positive invariant set; set the original output constraint set as... The original set of control input constraints is Then the tightening constraint set is:

[0100] ;

[0101] 49) Establish the set of reachable steady states of the nominal system: Define the nominal steady state as... , Represents the nominal steady-state vector. Let the nominal steady-state control input vector be denoted; then there exists a matrix . Make:

[0102] ;

[0103] ;

[0104] In the formula, Represents steady-state parameters. , Indicated by steady-state parameters to the nominal steady-state vector The mapping matrix, Indicated by steady-state parameters To nominal steady-state output The mapping matrix;

[0105] 410) Construct the prediction output model, and express the prediction output as follows:

[0106] ;

[0107] In the formula, Indicates the first The nominal prediction output sequence constructed at each sampling time point This represents the state prediction matrix from the current system observation state to the predicted output sequence. Indicates the first The system observation state obtained by the safety state observer at each sampling time. This represents the control prediction matrix from the nominal control sequence to the predicted output sequence. Indicates the first The nominal control sequence constructed at each sampling time. This represents the external input prediction matrix used to predict the tire self-alignment torque sequence and the predicted output sequence. This represents the predicted tire self-alignment torque sequence; , ,in, Indicates the first The prediction at the nth sampling time Nominal output at each sampling time, Indicates the first The prediction at the nth sampling time Nominal control quantity at each sampling time Indicates the prediction time domain, Indicates control over the time domain;

[0108] 411) The objective function for Tube-MPC optimization is constructed as follows:

[0109] ;

[0110] In the formula, This represents the objective function for Tube-MPC optimization. Indicates the nominal output variable. Indicates the nominal control variable. This represents the output estimate obtained from the safety state observer. Indicates a given output target value. Indicates the first The nominal output at each sampling time, Indicates the first The nominal control quantity at each sampling time. Indicates the nominal steady-state output. This represents the nominal steady-state control input. This represents the nominal output at the end of the prediction time domain. Indicates the first The reference steering angle at each sampling time. and These represent the prediction time domain and the control time domain, respectively. , , and Represents the weight matrix. , , and They represent the weight matrices respectively. , , and Weighted quadratic norm;

[0111] The weight matrix The Lyapunov equations are satisfied as follows:

[0112] ;

[0113] 412) Considering the safety constraints of the steer-by-wire system, constraints are imposed on the nominal control voltage and its rate of change, tire return torque, and front wheel steering angle, as follows:

[0114] ;

[0115] ;

[0116] ;

[0117] In the formula, Indicates the first The nominal control voltage at each sampling time. This indicates the lower limit of the nominal control voltage. Indicates the nominal upper limit of the control voltage. Indicates the first The nominal control voltage change rate at each sampling time This indicates the lower limit of the nominal control voltage change rate. This indicates the upper limit of the nominal control voltage change rate. Indicates the first The nominal tire return torque input at each sampling time. This indicates the lower limit of the tire return torque input. This indicates the upper limit of the tire return torque input. Indicates the first The nominal output at each sampling time, where the nominal output is the front wheel steering angle. Indicates the lower limit of the front wheel steering angle. Indicates the upper limit of the front wheel steering angle;

[0118] 413) Under tightening and terminal constraints, the rolling optimization problem is constructed as follows:

[0119] ;

[0120] ;

[0121] In the formula, This represents the objective function for Tube-MPC optimization. This represents the output estimate obtained from the safety state observer. Indicates a given output target value. Indicates the first The nominal output at each sampling time, This represents the nominal output constraint set obtained after tightening the front wheel steering angle safety constraint. This represents the set of nominal control constraints obtained after tightening the safety constraints on the control voltage and its rate of change. This represents the nominal output at the end of the prediction time domain. Represents the set of terminal constraints;

[0122] 414) Obtain the nominal optimal control sequence by solving the rolling optimization problem. Furthermore, the predictive control sequence used for the actuator-side buffer is obtained as follows:

[0123] ;

[0124] in, , ; Indicates the first The nominal optimal control sequence obtained at each sampling time point. Indicates the first The prediction at the nth sampling time The nominal optimal control quantity at each sampling time. , Indicates control over the time domain; This represents the predictive control sequence used to send to and buffer the steering actuator by steer-by-wire. Indicates the first The prediction at the nth sampling time The actual executable control quantity at each sampling time; Represented by the error feedback gain matrix The expanded block diagonal feedback gain matrix, express 3D identity matrix Represents the Kronecker product. This represents the output estimate obtained from the safety state observer. Indicates the output from the current measurement. and reference output The nominal optimal output is determined.

[0125] Furthermore, step 5) specifically includes:

[0126] 51) Establish a control sequence cache compensation mechanism for the steering-by-wire actuator. The control sequence cache compensation mechanism is used to update the predictive control sequence cache when the controller-actuator channel is communicating normally, and to call the cached predictive control quantity when the controller-actuator channel is interrupted due to a denial-of-service attack. The control sequence cache compensation mechanism includes predictive control sequence transmission, predictive control sequence caching, communication status judgment, cache sequence update, and cached control quantity invocation.

[0127] 52) Based on the predictive control sequence obtained in step 4). The predictive control sequence is sent to the steering-by-wire actuator via the controller-actuator channel, and the most recently successfully received predictive control sequence is stored in the steering-by-wire actuator.

[0128] 53) Determine whether the controller-actuator channel is communicating normally based on the detection results of denial-of-service attacks. Specifically, this involves detecting denial-of-service attacks through control message reception timestamps, heartbeat message timeouts, continuous packet loss counts, message sequence number changes, control command update flags, controller acknowledgment signals, abnormal CAN bus occupancy rates, or communication cycle timeouts.

[0129] If no control command is received from the controller within the set communication period, or the number of consecutive packet losses exceeds the set threshold, or the heartbeat message times out, or the control command update flag does not change, it is determined that there is a denial-of-service attack or communication interruption in the controller-actuator channel; otherwise, it is determined that the controller-actuator channel is in normal communication state.

[0130] When the controller-actuator channel is communicating normally, the steering actuator uses the currently received predictive control sequence and updates the buffer; when a denial-of-service attack occurs in the controller-actuator channel, the steering actuator retains the most recently successfully received predictive control sequence and calls the corresponding predictive control quantity according to the number of interrupt steps.

[0131] 54) Perform control sequence buffer compensation based on the communication state obtained in step 53):

[0132] When the controller-actuator channel is in normal communication mode, the steering-by-wire actuator adopts the currently received predictive control sequence and updates the current predictive control sequence to the latest cached sequence;

[0133] When the controller-actuator channel is interrupted by a denial-of-service attack, the steering actuator with drive by drive retains the most recently successfully received predictive control sequence and calls the corresponding buffered predictive control quantity according to the number of communication interruption steps.

[0134] 55) Settings This indicates the time of the most recent successful reception of the predictive control sequence. Let represent the number of steps after the controller-actuator channel interruption. Then, the buffered predictive control quantity invoked by the steering-by-wire actuator in the communication interruption state is expressed as:

[0135] ;

[0136] In the formula, Indicates the steering-by-wire actuator in the first Each sampling time point is based on the control quantity actually used according to the most recently successfully received predictive control sequence. Indicates the time of the most recent successful communication. The number of communication interruption steps in the generated and cached predictive control sequence The control quantity.

[0137] Furthermore, step 6) specifically includes:

[0138] 61) Select the control quantity that actually acts on the steering actuator at the current moment according to the control sequence buffer compensation mechanism in step 5); when the controller-actuator channel is communicating normally, use the first control quantity of the current predicted control sequence; when the controller-actuator channel is under denial-of-service attack, use the corresponding control quantity in the predicted control sequence buffered by the steering actuator.

[0139] 62) Based on the control quantity selection results in step 61), construct the control quantity selection rule that actually acts on the steering motor, expressed as:

[0140] ;

[0141] In the formula, This represents the set of times a denial-of-service attack occurred. Represents the set of normal communication times. This represents the first control variable obtained through rolling optimization at the current moment. This represents the corresponding control quantity in the predictive control sequence buffered by the steering-by-wire actuator;

[0142] 63) The control quantity output by the actual control law is applied to the actual steering motor, so that the steering motor drives the steer-by-wire actuator to complete the corner tracking; thus, when the sensor-controller channel sampling output is lost, the system observation state is provided by the safety state observer; when the controller-actuator channel control command is interrupted, the execution control quantity is provided by the predictive control sequence buffer compensation mechanism, so as to realize the corner tracking control of the steer-by-wire system of heavy-duty vehicles under denial-of-service attack.

[0143] The beneficial effects of this invention are:

[0144] 1. This invention addresses the problem of lost sampling output in the sensor-controller channel by designing a safety state observer. It incorporates measurement noise, external disturbances, and sampling interruptions caused by denial-of-service attacks into the estimation error system. By solving the observer gain using linear matrix inequalities, the system can still obtain a reliable observation state even when sampling output is intermittently lost, providing a state basis for subsequent robust safety control.

[0145] 2. This invention incorporates steering disturbances caused by factors such as the large mass of heavy-duty vehicles, high front axle load, large inertia of steering actuators, changes in tire return torque, load transfer, and changes in road surface adhesion into the system model. The Tube-MPC method is used to constrain the deviation between the actual system and the nominal system within a robust positive invariant set, thereby improving the adaptability of the steer-by-wire system to time-varying disturbances and parameter perturbations of heavy-duty vehicles.

[0146] 3. The present invention generates a predictive control sequence on the controller side and sends the predictive control sequence to the steer-by-wire actuator for caching. When the controller-actuator channel is attacked by a denial-of-service attack and the current control command is interrupted, the actuator can select the corresponding control quantity according to the cached sequence and drive the actual steering motor, avoiding the direct loss of control input, thereby improving the continuity and safety of corner tracking.

[0147] 4. This invention establishes defense mechanisms for the loss of state information in the sensor-controller channel and the interruption of control commands in the controller-actuator channel. Through the coordinated work of the safety state observer, the Tube-MPC robust safety controller, and the steer-by-wire actuator end-buffer compensation mechanism, communication interruption, measurement noise, and steering disturbances of heavy-duty vehicles are jointly processed, thereby reducing the risk of amplified state estimation errors, control input interruption, and angle tracking failure, and improving the safe operation capability of the steer-by-wire system of heavy-duty vehicles in a network attack environment. Attached Figure Description

[0148] Figure 1 This is a schematic diagram of the method of the present invention. Detailed Implementation

[0149] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to embodiments and accompanying drawings. The content mentioned in the embodiments is not intended to limit the present invention.

[0150] Reference Figure 1 As shown, the present invention provides an information security control method for a heavy-duty vehicle steer-by-wire system, comprising the following steps:

[0151] 1) Establish a networked control architecture and dynamic model for the steer-by-wire system of heavy-duty vehicles, specifically including:

[0152] 11) Establish a networked control architecture for the steer-by-wire system of heavy-duty vehicles. The networked control architecture includes: a steering angle sensor, an electronic control unit, a CAN communication network, and a steer-by-wire actuator. The steer-by-wire actuator includes: a permanent magnet synchronous steering motor, a reduction mechanism, a rack and pinion mechanism, and a front wheel steering mechanism. The steering angle sensor sends a steering angle signal to the electronic control unit through the sensor-controller channel in the CAN communication network. The electronic control unit sends a steering control command to the steer-by-wire actuator through the controller-actuator channel in the CAN communication network. The permanent magnet synchronous steering motor generates a steering driving torque according to the steering control command. The steering driving torque is transmitted to the rack and pinion mechanism through the reduction mechanism, and the rack and pinion mechanism drives the front wheel steering mechanism to rotate through the steering tie rod.

[0153] 12) Establish a dynamic model of the steer-by-wire system for heavy-duty vehicles. The dynamic model includes: steering motor voltage equations, steering motor dynamic equations, rack and pinion dynamic equations, and front wheel dynamic equations; wherein,

[0154] The voltage equation for the steering motor is as follows:

[0155] ;

[0156] The dynamic equations of the steering motor are as follows:

[0157] ;

[0158] The dynamic equations of the gear and rack are as follows:

[0159] ;

[0160] The front wheel dynamics equations are as follows:

[0161] ;

[0162] In the formula, This represents the equivalent control current of the steering motor. This represents the rate of change of the equivalent control current of the steering motor. This represents the equivalent control voltage of the steering motor. This represents the equivalent resistance of the steering motor. This represents the equivalent inductance of the steering motor. This represents the back electromotive force coefficient of the steering motor. Indicates the angle of the steering motor shaft. Indicates the angular velocity of the steering motor shaft. This indicates the angular acceleration of the steering motor shaft. This represents the moment of inertia of the steering motor. This represents the coefficient of viscous friction of the steering motor. This indicates the torque coefficient of the steering motor. This indicates the load torque at the shaft end of the steering motor. Indicates the displacement of the steering rack. Indicates the speed of the steering rack. Indicates the acceleration of the steering rack. Indicates the equivalent mass of the steering rack. This represents the equivalent damping coefficient of the gear and rack mechanism. Indicates the rack and pinion load torque. Indicates the gear ratio of the reduction mechanism. This indicates the radius of the pinion in a gear and rack mechanism. Indicates the front wheel steering angle. Indicates the steering angular velocity of the front wheels. This represents the front wheel steering angular acceleration. This represents the equivalent moment of inertia of the front wheel steering mechanism. This represents the equivalent viscous friction coefficient of the front wheel steering mechanism. Indicates the steering tie rod transmission ratio. This indicates the tire return torque;

[0163] 13) Establish the expressions for the load torque at the shaft end of the steering motor and the load torque of the rack;

[0164] The load torque at the end of the steering motor shaft is determined by the relative deformation between the steering motor shaft angle and the rack displacement obtained through equivalent conversion by the reduction mechanism, and is expressed as:

[0165] ;

[0166] The rack load torque is determined by the relative deformation between the steering rack displacement and the front wheel steering angle obtained through equivalent conversion via the steering tie rod, and is expressed as:

[0167] ;

[0168] In the formula, This indicates the equivalent torsional stiffness at the end of the steering motor shaft. Indicates the equivalent rack stiffness of a gear and rack mechanism;

[0169] 14) Define the system state, control input, and tire return torque input; let Representing continuous time, the system state vector, control input, and tire self-aligning torque input are respectively expressed as:

[0170] ;

[0171] ;

[0172] ;

[0173] In the formula, Representing continuous time The system state vector under the following conditions Representing continuous time The steering motor control voltage input is below. Representing continuous time Input the tire return torque;

[0174] Select front wheel steering angle As the measurement output, a continuous state-space model is established as follows:

[0175] ;

[0176] ;

[0177] In the formula, Represents the system state vector in continuous time. The first derivative, Representing continuous time The measurement output below, Indicates external disturbance. Indicates measurement noise; the external disturbance This includes equivalent disturbances caused by the large overall mass of heavy-duty vehicles, high front axle load, large inertia of steering actuators, changes in tire return torque, changes in kingpin resistance torque, gear and rack friction, load transfer, and changes in road surface adhesion.

[0178] ;

[0179] ; ; ;

[0180] In the formula, This represents the state matrix of the steering-by-wire system. This represents the steering motor control input matrix. This represents the tire return torque input matrix. This represents the output matrix of the front wheel steering angle measurement. This represents the external perturbation input matrix. Represents the measurement noise input matrix;

[0181] 15) Discretize the continuous state-space model established in step 14) to obtain the discrete dynamic model required for controller design; let If we denote the discrete sampling time, then the discrete dynamic model can be expressed as:

[0182] ;

[0183] ;

[0184] In the formula, Indicates the first The system state vector at each sampling time. Indicates the first The system state vector at each sampling time. Indicates the first Steering motor control voltage input at each sampling time, Indicates the first The tire return torque input at each sampling time. Indicates the first External disturbances at each sampling time Indicates the first Measurement noise at each sampling time, Indicates the first Measurement output at each sampling time, Represents the state matrix of a discrete system. This represents the discrete steering motor control input matrix. This represents the discrete tire return torque input matrix. This represents the discrete external disturbance input matrix. Represents the discrete measurement output matrix. Represents the discrete measurement noise input matrix; , , , , , Indicates the controller sampling period. .

[0185] 2) Establish a dual-channel communication interruption model for the steer-by-wire system of heavy-duty vehicles under network attacks, and establish a sensor-controller channel sampling output model;

[0186] Specifically, step 2) involves establishing a dual-channel communication interruption model under a network attack on the steer-by-wire system of a heavy-duty vehicle, which includes:

[0187] The impact of network attacks on the dual-channel signal transmission process is analyzed, and a model is developed to address the communication interruption problem caused by denial-of-service attacks, resulting in a dual-channel communication interruption model.

[0188] The network attacks include: data tampering, fake data injection, replay attacks, and denial-of-service attacks; among them, data tampering, fake data injection, and replay attacks affect the authenticity, integrity, and timing consistency of transmitted data; denial-of-service attacks prevent steering angle sampling output or steering control commands from reaching the receiving end on time by blocking communication resources, occupying communication bandwidth, or interfering with message transmission.

[0189] The dual channels include a sensor-controller channel and a controller-actuator channel. When the sensor-controller channel is attacked by a network, the steering angle sampling output will not be transmitted to the controller. When the controller-actuator channel is attacked by a network, the current control command will not be transmitted to the steering-by-wire actuator.

[0190] The dual-channel communication interruption model includes: a communication interruption state function, a communication interruption frequency boundary, and a communication interruption duration boundary;

[0191] Define communication interruption status function ,in This indicates that the communication link is in an interrupted state due to a denial-of-service attack. This indicates that the communication link is in a normal communication state. The function represents the normal communication state, and ;when At that time, the communication link was normal; when At that time, the communication link was interrupted;

[0192] Establish the communication interruption frequency boundary in the dual-channel communication interruption model: for any time interval , This indicates the start time of the time interval. This represents the end time of the time interval, and satisfies... ;set up This indicates the number of communication interruptions that occurred within the given time interval. The number of communication interruptions is defined by the following formula:

[0193] ;

[0194] In the formula, This represents the communication interruption frequency offset constant. This represents the lower bound of the average communication interruption interval; thus, it constrains the frequency of dual-channel communication interruptions per unit time.

[0195] Establish the communication interruption duration boundary in the dual-channel communication interruption model: Let Indicates time interval Given the set of durations of internal communication interruptions, the cumulative duration of communication interruptions is limited by the following formula:

[0196] ;

[0197] In the formula, This represents a constant indicating the duration offset of the communication interruption. This represents the average communication interruption duration constraint parameter; thus, it constrains the cumulative duration of dual-channel communication interruptions within a given time interval.

[0198] Specifically, step 2) of establishing the sensor-controller channel sampling output model includes:

[0199] set up The sensor-controller channel sampling output model under the influence of communication interruption caused by a denial-of-service attack at the sampling time is expressed as follows:

[0200] ;

[0201] In the formula, Indicates the controller at the sampling time The actual sampled output, This represents the output matrix of the front wheel steering angle measurement. Indicates the system status. Indicates the measurement noise term; when hour, The sensor-controller channel is in a communication interruption state, and the controller cannot obtain the current valid steering angle sampling output; when hour, The sensor-controller channel is in normal communication mode, and the controller obtains normal steering angle sampling output.

[0202] 3) Design a security status observer to obtain the system's observed status under network attacks, specifically including:

[0203] 31) Based on the dynamic model established in step 1) and the sensor-controller channel sampling output model established in step 2), design a safety state observer to reconstruct the system state under conditions of sampling output loss, measurement noise, and steering disturbances from heavy-duty vehicles; the safety state observer is represented as:

[0204] ;

[0205] In the formula, Indicates continuous time. Indicates the sampling time. Representing continuous time The system's observed state vector is as follows. Represents the system's observed state vector in continuous time. The first derivative, Indicates the sampling time The corresponding system observation state vector, Indicates the sampling time The corresponding output estimate, Indicates the controller at the sampling time The sampling output obtained through the sensor-controller channel This represents the gain matrix of the safety state observer;

[0206] 32) Define the state estimation error : and define the sampling delay : Combining the sensor-controller channel sampling output model under the influence of communication interruption caused by a denial-of-service attack, the system equation for the state estimation error is obtained as follows:

[0207] ;

[0208] In the formula, Indicates the sampling interval. This indicates steering disturbance of a heavily loaded vehicle. This represents the state estimation error corresponding to the delayed sampling time. Represents the measurement noise input matrix. This represents the measurement noise corresponding to the delayed sampling time.

[0209] 33) Solve for the observer gain matrix using the Lyapunov method; if a symmetric positive definite matrix exists... , Invertible matrix ,matrix Given the associated relaxation matrix and positive scalar, if the linear matrix inequality holds, then the estimation error system is uniformly and eventually bounded. The linear matrix inequality is as follows:

[0210] ;

[0211] ;

[0212] ;

[0213] in, This represents the transpose term in a symmetric matrix, determined by symmetry, where the lower triangular partition of the matrix is ​​equal to the transpose of the upper triangular partition. The matrix elements in the linear matrix inequality are defined as follows:

[0214] , , , , , , , , , , , , , , , , , , , , , , , , ;

[0215] In the formula, Representation matrix Symmetric operators, Let represent a matching matrix of arbitrary dimensions, and , and It is a symmetric positive definite matrix. It is an invertible matrix. , , , , , , , and Let the variables be the matrix to be solved. , , , , and It is a positive scalar;

[0216] 34) Based on the solution of the linear matrix inequality, the gain matrix of the safety state observer is obtained as follows:

[0217] ;

[0218] In the formula, Used to determine the feedback effect of the sampling output deviation in the safety status observer in step 31) on the observation status update;

[0219] 35) The gain matrix of the safety state observer obtained in step 34) Substitute the security status observer established in step 31) to obtain the system observation status under network attack online. ;

[0220] When the sensor-controller channel is in normal communication mode, the controller outputs based on the current sampling. With output estimate The deviation between the observation states is corrected; when the sensor-controller channel is attacked by a denial-of-service attack and the sampling output is interrupted, the controller recursively updates the observation state based on the dynamic model of the heavy-duty vehicle's steer-by-wire system, historical control inputs, and the observation state at the previous moment.

[0221] 4) Design a robust safety controller based on Tube-MPC to generate predictive control sequences, specifically including:

[0222] 41) Based on the discrete dynamics model established in step 15), the dual-channel communication interruption model established in step 2), and the system observation state under network attack obtained in step 3), design a robust security control algorithm based on Tube-MPC; the robust security control algorithm calculates the optimal control sequence in the future prediction time domain during each rolling optimization.

[0223] 42) Establish the equations of the controlled system under a controller-actuator channel attack; set the actuator at time... The actual control quantity used is The equation of the controlled system is then expressed as:

[0224] ;

[0225] 43) Establish a nominal prediction system: Set the nominal state as The nominal output is The nominal control quantity is The nominal system equations are then expressed as:

[0226] ;

[0227] In the formula, The system observation state obtained in step 3) For the first The nominal state at each sampling time. This indicates the nominal tire return torque input;

[0228] 44) Define the error between the output of the controlled system and the output of the nominal system. , The Tube-MPC control law is constructed as follows:

[0229] ;

[0230] In the formula, Represents the error feedback control matrix;

[0231] 45) Based on the actual controlled system and the nominal system, the error dynamics are obtained as follows:

[0232] ;

[0233] 46) Define the closed-loop state transition matrix of the error equation. Perturbation input matrix and define lumped disturbance ,in , , Indicates the first The lumped perturbation vector at each sampling time. This represents the deviation between the nominal control quantity and the actual control quantity used by the actuator. The deviation between the nominal tire return torque input and the actual tire return torque input can be represented by the following dynamic error formula:

[0234] ;

[0235] 47) Constructing a lumped disturbance set and robust positive invariant set ;like make If it is stable, then there exists a robust positive invariant set. The following conditions must be met:

[0236] ;

[0237] 48) Tighten the output constraints and control input constraints based on the robust positive invariant set; set the original output constraint set as... The original set of control input constraints is Then the tightening constraint set is:

[0238] ;

[0239] 49) Establish the set of reachable steady states of the nominal system: Define the nominal steady state as... , Represents the nominal steady-state vector. Let the nominal steady-state control input vector be denoted; then there exists a matrix . Make:

[0240] ;

[0241] ;

[0242] In the formula, Represents steady-state parameters. , Indicated by steady-state parameters to the nominal steady-state vector The mapping matrix, Indicated by steady-state parameters To nominal steady-state output The mapping matrix;

[0243] 410) Construct the prediction output model, and express the prediction output as follows:

[0244] ;

[0245] In the formula, Indicates the first The nominal prediction output sequence constructed at each sampling time point This represents the state prediction matrix from the current system observation state to the predicted output sequence. Indicates the first The system observation state obtained by the safety state observer at each sampling time. This represents the control prediction matrix from the nominal control sequence to the predicted output sequence. Indicates the first The nominal control sequence constructed at each sampling time. This represents the external input prediction matrix used to predict the tire self-alignment torque sequence and the predicted output sequence. This represents the predicted tire self-alignment torque sequence; , ,in, Indicates the first The prediction at the nth sampling time Nominal output at each sampling time, Indicates the first The prediction at the nth sampling time Nominal control quantity at each sampling time Indicates the prediction time domain, Indicates control over the time domain;

[0246] 411) The objective function for Tube-MPC optimization is constructed as follows:

[0247] ;

[0248] In the formula, This represents the objective function for Tube-MPC optimization. Indicates the nominal output variable. Indicates the nominal control variable. This represents the output estimate obtained from the safety state observer. Indicates a given output target value. Indicates the first The nominal output at each sampling time, Indicates the first The nominal control quantity at each sampling time. Indicates the nominal steady-state output. This represents the nominal steady-state control input. This represents the nominal output at the end of the prediction time domain. Indicates the first The reference steering angle at each sampling time. and These represent the prediction time domain and the control time domain, respectively. , , and Represents the weight matrix. , , and They represent the weight matrices respectively. , , and Weighted quadratic norm;

[0249] The weight matrix The Lyapunov equations are satisfied as follows:

[0250] ;

[0251] 412) Considering the safety constraints of the steer-by-wire system, constraints are imposed on the nominal control voltage and its rate of change, tire return torque, and front wheel steering angle, as follows:

[0252] ;

[0253] ;

[0254] ;

[0255] In the formula, Indicates the first The nominal control voltage at each sampling time. This indicates the lower limit of the nominal control voltage. Indicates the nominal upper limit of the control voltage. Indicates the first The nominal control voltage change rate at each sampling time This indicates the lower limit of the nominal control voltage change rate. This indicates the upper limit of the nominal control voltage change rate. Indicates the first The nominal tire return torque input at each sampling time. This indicates the lower limit of the tire return torque input. This indicates the upper limit of the tire return torque input. Indicates the first The nominal output at each sampling time, where the nominal output is the front wheel steering angle. Indicates the lower limit of the front wheel steering angle. Indicates the upper limit of the front wheel steering angle;

[0256] 413) Under tightening and terminal constraints, the rolling optimization problem is constructed as follows:

[0257] ;

[0258] ;

[0259] In the formula, This represents the objective function for Tube-MPC optimization. This represents the output estimate obtained from the safety state observer. Indicates a given output target value. Indicates the first The nominal output at each sampling time, This represents the nominal output constraint set obtained after tightening the front wheel steering angle safety constraint. This represents the set of nominal control constraints obtained after tightening the safety constraints on the control voltage and its rate of change. This represents the nominal output at the end of the prediction time domain. Represents the set of terminal constraints;

[0260] 414) Obtain the nominal optimal control sequence by solving the rolling optimization problem. Furthermore, the predictive control sequence used for the actuator-side buffer is obtained as follows:

[0261] ;

[0262] in, , ; Indicates the first The nominal optimal control sequence obtained at each sampling time point. Indicates the first The prediction at the nth sampling time The nominal optimal control quantity at each sampling time. , Indicates control over the time domain; This represents the predictive control sequence used to send to and buffer the steering actuator by steer-by-wire. Indicates the first The prediction at the nth sampling time The actual executable control quantity at each sampling time; Represented by the error feedback gain matrix The expanded block diagonal feedback gain matrix, express 3D identity matrix Represents the Kronecker product. This represents the output estimate obtained from the safety state observer. Indicates the output from the current measurement. and reference output The nominal optimal output is determined.

[0263] 5) Send the predictive control sequence obtained in step 4) to the steer-by-wire actuator, and design a control sequence buffer compensation mechanism for the steer-by-wire actuator, specifically including:

[0264] 51) Establish a control sequence cache compensation mechanism for the steering-by-wire actuator. The control sequence cache compensation mechanism is used to update the predictive control sequence cache when the controller-actuator channel is communicating normally, and to call the cached predictive control quantity when the controller-actuator channel is interrupted due to a denial-of-service attack. The control sequence cache compensation mechanism includes predictive control sequence transmission, predictive control sequence caching, communication status judgment, cache sequence update, and cached control quantity invocation.

[0265] 52) Based on the predictive control sequence obtained in step 4). The predictive control sequence is sent to the steering-by-wire actuator via the controller-actuator channel, and the most recently successfully received predictive control sequence is stored in the steering-by-wire actuator.

[0266] 53) Determine whether the controller-actuator channel is communicating normally based on the detection results of denial-of-service attacks. Specifically, this involves detecting denial-of-service attacks through control message reception timestamps, heartbeat message timeouts, continuous packet loss counts, message sequence number changes, control command update flags, controller acknowledgment signals, abnormal CAN bus occupancy rates, or communication cycle timeouts.

[0267] If no control command is received from the controller within the set communication period, or the number of consecutive packet losses exceeds the set threshold, or the heartbeat message times out, or the control command update flag does not change, it is determined that there is a denial-of-service attack or communication interruption in the controller-actuator channel; otherwise, it is determined that the controller-actuator channel is in normal communication state.

[0268] When the controller-actuator channel is communicating normally, the steering actuator uses the currently received predictive control sequence and updates the buffer; when a denial-of-service attack occurs in the controller-actuator channel, the steering actuator retains the most recently successfully received predictive control sequence and calls the corresponding predictive control quantity according to the number of interrupt steps.

[0269] 54) Perform control sequence buffer compensation based on the communication state obtained in step 53):

[0270] When the controller-actuator channel is in normal communication mode, the steering-by-wire actuator adopts the currently received predictive control sequence and updates the current predictive control sequence to the latest cached sequence;

[0271] When the controller-actuator channel is interrupted by a denial-of-service attack, the steering actuator with drive by drive retains the most recently successfully received predictive control sequence and calls the corresponding buffered predictive control quantity according to the number of communication interruption steps.

[0272] 55) Settings This indicates the time of the most recent successful reception of the predictive control sequence. Let represent the number of steps after the controller-actuator channel interruption. Then, the buffered predictive control quantity invoked by the steering-by-wire actuator in the communication interruption state is expressed as:

[0273] ;

[0274] In the formula, Indicates the steering-by-wire actuator in the first Each sampling time point is based on the control quantity actually used according to the most recently successfully received predictive control sequence. Indicates the time of the most recent successful communication. The number of communication interruption steps in the generated and cached predictive control sequence The control quantity.

[0275] 6) Select the corresponding control quantity according to the control sequence buffer compensation mechanism in step 5), and apply it to the actual steering motor to realize the angle tracking control of the steer-by-wire system of heavy-duty vehicles, specifically including:

[0276] 61) Select the control quantity that actually acts on the steering actuator at the current moment according to the control sequence buffer compensation mechanism in step 5); when the controller-actuator channel is communicating normally, use the first control quantity of the current predicted control sequence; when the controller-actuator channel is under denial-of-service attack, use the corresponding control quantity in the predicted control sequence buffered by the steering actuator.

[0277] 62) Based on the control quantity selection results in step 61), construct the control quantity selection rule that actually acts on the steering motor, expressed as:

[0278] ;

[0279] In the formula, This represents the set of times a denial-of-service attack occurred. Represents the set of normal communication times. This represents the first control variable obtained through rolling optimization at the current moment. This represents the corresponding control quantity in the predictive control sequence buffered by the steering-by-wire actuator;

[0280] 63) The control quantity output by the actual control law is applied to the actual steering motor, so that the steering motor drives the steer-by-wire actuator to complete the corner tracking; thus, when the sensor-controller channel sampling output is lost, the system observation state is provided by the safety state observer; when the controller-actuator channel control command is interrupted, the execution control quantity is provided by the predictive control sequence buffer compensation mechanism, so as to realize the corner tracking control of the steer-by-wire system of heavy-duty vehicles under denial-of-service attack.

[0281] This invention has many specific applications. The above description is only a preferred embodiment of this invention. It should be noted that for those skilled in the art, several improvements can be made without departing from the principle of this invention, and these improvements should also be considered within the scope of protection of this invention.

Claims

1. A method for information security control of a steer-by-wire system for heavy-duty vehicles, characterized in that, The steps are as follows: 1) Establish a networked control architecture and dynamic model for the steer-by-wire system of heavy-duty vehicles; 2) Establish a dual-channel communication interruption model under network attacks on the steer-by-wire system of heavy-duty vehicles, and establish a sensor-controller channel sampling output model; the establishment of the dual-channel communication interruption model under network attacks on the steer-by-wire system of heavy-duty vehicles in step 2) specifically includes: The impact of network attacks on the dual-channel signal transmission process is analyzed, and a model is developed to address the communication interruption problem caused by denial-of-service attacks, resulting in a dual-channel communication interruption model. The dual-channel communication interruption model includes: a communication interruption state function, a communication interruption frequency boundary, and a communication interruption duration boundary; Define communication interruption status function ,in This indicates that the communication link is in an interrupted state due to a denial-of-service attack. This indicates that the communication link is in a normal communication state. The function represents the normal communication state, and ;when At that time, the communication link was normal; when At that time, the communication link was interrupted; Establish the communication interruption frequency boundary in the dual-channel communication interruption model: for any time interval , Indicates the start time of this time interval. This represents the end time of the time interval, and satisfies... ;set up This indicates the number of communication interruptions that occurred within that time interval; Establish the communication interruption duration boundary in the dual-channel communication interruption model: Let Indicates time interval The set of internal communication interruption durations constrains the cumulative duration of dual-channel communication interruptions within a given time interval; 3) Obtain the system's observation status under network attacks; 4) Generate predictive control sequences; 5) Send the predictive control sequence obtained in step 4) to the steer-by-wire actuator, and design a control sequence buffer compensation mechanism for the steer-by-wire actuator; 6) Select the corresponding control quantity according to the control sequence buffer compensation mechanism in step 5) and apply it to the actual steering motor to realize the angle tracking control of the steer-by-wire system of heavy-duty vehicles.

2. The information security control method for a heavy-duty vehicle steer-by-wire system according to claim 1, characterized in that, Step 1) specifically includes: 11) Establish a networked control architecture for the steer-by-wire system of heavy-duty vehicles. The networked control architecture includes: a steering angle sensor, an electronic control unit, a CAN communication network, and a steer-by-wire actuator. The steer-by-wire actuator includes: a permanent magnet synchronous steering motor, a reduction mechanism, a rack and pinion mechanism, and a front wheel steering mechanism. The steering angle sensor sends a steering angle signal to the electronic control unit through the sensor-controller channel in the CAN communication network. The electronic control unit sends a steering control command to the steer-by-wire actuator through the controller-actuator channel in the CAN communication network. The permanent magnet synchronous steering motor generates a steering driving torque according to the steering control command. The steering driving torque is transmitted to the rack and pinion mechanism through the reduction mechanism, and the rack and pinion mechanism drives the front wheel steering mechanism to rotate through the steering tie rod. 12) Establish a dynamic model of the steer-by-wire system for heavy-duty vehicles. The dynamic model includes: steering motor voltage equation, steering motor dynamic equation, rack and pinion dynamic equation, and front wheel dynamic equation. 13) Establish the expressions for the load torque at the shaft end of the steering motor and the load torque of the rack; 14) Define the system state, control input, and tire return torque input; 15) Discretize the continuous state-space model established in step 14) to obtain the discrete dynamic model required for controller design.

3. The information security control method for a heavy-duty vehicle steer-by-wire system according to claim 1, characterized in that, Step 2) of establishing the sensor-controller channel sampling output model specifically includes: set up The sensor-controller channel sampling output model under the influence of communication interruption caused by a denial-of-service attack at the sampling time is expressed as follows: ； In the formula, Indicates the controller at the sampling time The actual sampled output, This represents the output matrix of the front wheel steering angle measurement. Indicates the system status. Indicates the measurement noise term; when hour, The sensor-controller channel is in a communication interruption state, and the controller cannot obtain the current valid steering angle sampling output; when hour, The sensor-controller channel is in normal communication mode, and the controller obtains normal steering angle sampling output.

4. The information security control method for a heavy-duty vehicle steer-by-wire system according to claim 1, characterized in that, Step 3) specifically includes: 31) Based on the dynamic model established in step 1) and the sensor-controller channel sampling output model established in step 2), design a safety state observer to reconstruct the system state under the effects of sampling output loss, measurement noise, and steering disturbance of heavy-duty vehicles. 32) Define the state estimation error and define the sampling delay By combining the sensor-controller channel sampling output model under the influence of communication interruption caused by denial-of-service attacks, the system equation for state estimation error is obtained. 33) Solve for the observer gain matrix using the Lyapunov method; if a symmetric positive definite matrix exists... , Invertible matrix ,matrix If, along with the associated relaxation matrix and positive scalar, the linear matrix inequality holds, then the estimation error system is uniformly eventually bounded. 34) Based on the solution of the linear matrix inequality, the gain matrix of the safety state observer is obtained; 35) Substitute the security state observer gain matrix obtained in step 34) into the security state observer established in step 31) to obtain the system observation state under network attack online. ; When the sensor-controller channel is in normal communication mode, the controller outputs based on the current sampling. With output estimate The deviation between the observation states is corrected; when the sensor-controller channel is attacked by a denial-of-service attack and the sampling output is interrupted, the controller recursively updates the observation state based on the dynamic model of the heavy-duty vehicle's steer-by-wire system, historical control inputs, and the observation state at the previous moment.

5. The information security control method for a heavy-duty vehicle steer-by-wire system according to claim 1, characterized in that, Step 4) specifically includes: 41) Design a robust safety control algorithm based on Tube-MPC; the robust safety control algorithm calculates the optimal control sequence in the future prediction time domain during each rolling optimization; 42) Establish the equations of the controlled system under a controller-actuator channel attack; set the actuator at time... The actual control quantity used is ; 43) Establish a nominal prediction system: Set the nominal state as The nominal output is The nominal control quantity is ; 44) Define the error between the output of the controlled system and the output of the nominal system. And construct the Tube-MPC control law; 45) Obtain the error dynamics based on the actual controlled system and the nominal system; 46) Define the closed-loop state transition matrix of the error equation. Perturbation input matrix And define lumped disturbance ; 47) Constructing a set of lumped disturbances and robust positive invariant set ;like make If it is stable, then there exists a robust positive invariant set. The following conditions must be met: ； 48) Tighten the output constraints and control input constraints based on the robust positive invariant set; set the original output constraint set as... The original set of control input constraints is Then the tightening constraint set is: ； 49) Establish the set of reachable steady states of the nominal system: Define the nominal steady state as... , Represents the nominal steady-state vector. This represents the nominal steady-state control input vector; 410) Construct a predictive output model; 411) Construct the Tube-MPC optimization objective function; 412) Considering the safety constraints of the steer-by-wire system, constraints are imposed on the nominal control voltage and its rate of change, tire return torque, and front wheel steering angle; 413) Construct a rolling optimization problem under tightening and terminal constraints; 414) The nominal optimal control sequence is obtained by solving the rolling optimization problem, and the predictive control sequence for the actuator-side buffer is further obtained.

6. The information security control method for a heavy-duty vehicle steer-by-wire system according to claim 1, characterized in that, Step 5) specifically includes: 51) Establish a control sequence cache compensation mechanism for the steering-by-wire actuator. The control sequence cache compensation mechanism is used to update the predictive control sequence cache when the controller-actuator channel is communicating normally, and to call the cached predictive control quantity when the controller-actuator channel is interrupted due to a denial-of-service attack. The control sequence cache compensation mechanism includes predictive control sequence transmission, predictive control sequence caching, communication status judgment, cache sequence update, and cached control quantity invocation. 52) Based on the predictive control sequence obtained in step 4), the predictive control sequence is sent to the steering-by-wire actuator through the controller-actuator channel, and the most recently successfully received predictive control sequence is stored in the steering-by-wire actuator; 53) Determine whether the controller-actuator channel is communicating normally based on the detection results of denial-of-service attacks. Specifically, this involves detecting denial-of-service attacks through control message reception timestamps, heartbeat message timeouts, continuous packet loss counts, message sequence number changes, control command update flags, controller acknowledgment signals, abnormal CAN bus occupancy rates, or communication cycle timeouts. If no control command is received from the controller within the set communication period, or the number of consecutive packet losses exceeds the set threshold, or the heartbeat message times out, or the control command update flag does not change, it is determined that there is a denial-of-service attack or communication interruption in the controller-actuator channel; otherwise, it is determined that the controller-actuator channel is in normal communication state. When the controller-actuator channel is communicating normally, the steering actuator uses the currently received predictive control sequence and updates the buffer; when a denial-of-service attack occurs in the controller-actuator channel, the steering actuator retains the most recently successfully received predictive control sequence and calls the corresponding predictive control quantity according to the number of interrupt steps. 54) Perform control sequence buffer compensation based on the communication state obtained in step 53): When the controller-actuator channel is in normal communication mode, the steering-by-wire actuator adopts the currently received predictive control sequence and updates the current predictive control sequence to the latest cached sequence; When the controller-actuator channel is interrupted by a denial-of-service attack, the steering actuator with drive by drive retains the most recently successfully received predictive control sequence and calls the corresponding buffered predictive control quantity according to the number of communication interruption steps. 55) Settings This indicates the time of the most recent successful reception of the predictive control sequence. Let represent the number of steps after the controller-actuator channel interruption. Then, the buffered predictive control quantity invoked by the steering-by-wire actuator in the communication interruption state is expressed as: ； In the formula, Indicates the steering-by-wire actuator in the first Each sampling time point is based on the control quantity actually used according to the most recently successfully received predictive control sequence. Indicates the time of the most recent successful communication. The number of communication interruption steps in the generated and cached predictive control sequence The control quantity.

7. The information security control method for a heavy-duty vehicle steer-by-wire system according to claim 6, characterized in that, Step 6) specifically includes: 61) Select the control quantity that actually acts on the steering actuator at the current moment according to the control sequence buffer compensation mechanism in step 5); when the controller-actuator channel is communicating normally, use the first control quantity of the current predicted control sequence; when the controller-actuator channel is under denial-of-service attack, use the corresponding control quantity in the predicted control sequence buffered by the steering actuator. 62) Based on the control quantity selection results in step 61), construct the control quantity selection rules that actually act on the steering motor; 63) The control quantity output by the actual control law is applied to the actual steering motor, so that the steering motor drives the steer-by-wire actuator to complete the corner tracking; thus, when the sensor-controller channel sampling output is lost, the system observation state is provided by the safety state observer; when the controller-actuator channel control command is interrupted, the execution control quantity is provided by the predictive control sequence buffer compensation mechanism, so as to realize the corner tracking control of the steer-by-wire system of heavy-duty vehicles under denial-of-service attack.

Images