A heterogeneous vehicle cooperative control system and method based on a dynamic event triggering mechanism

Abstract

The application relates to a heterogeneous vehicle cooperative control system and method based on a dynamic event triggering mechanism and belongs to the technical field of vehicle cooperative control. The method comprises the following steps: establishing a nonlinear heterogeneous vehicle system dynamics model; calculating a vehicle-to-vehicle tracking distance error, fixed-time funnel control, error conversion and constructing a sliding mode surface; constructing a fixed-time-based auxiliary system to compensate for the nonlinear error of an actuator and the control input error caused by an event trigger; constructing an adaptive law for external and internal disturbances; designing a fixed-time formation controller; designing a dynamic event triggering control mechanism; judging whether an event triggering condition is met, and executing the method flow according to the triggering condition; and judging whether an ending condition is met, and executing the method flow according to the ending condition. The system and method can guarantee the cooperative control performance of heterogeneous vehicles in complex scenes, significantly reduce the communication and calculation costs, and enhance the adaptability and flexibility of the heterogeneous vehicle system.

Description

Technical Field

[0001] This invention belongs to the field of vehicle cooperative control technology, and particularly relates to a heterogeneous vehicle cooperative control system and method based on a dynamic event triggering mechanism. Background Technology

[0002] Mobile travel is one of the most important activities in human production and life. It fulfills the basic human need for spatial and temporal mobility, greatly promoting both spiritual and material exchange. With the rapid development of intelligent transportation systems, the intelligence and automation levels of vehicles are constantly improving. Heterogeneous vehicle cooperative control, as an emerging technology, aims to achieve a safer, more efficient, and environmentally friendly transportation operation mode through the efficient collaboration between multiple vehicles of different types. This cooperative control can not only improve traffic flow efficiency but also significantly reduce energy consumption and environmental pollution, which is of great significance for building future intelligent transportation systems. In recent years, with the continuous advancement of autonomous driving technology, vehicle networking technology, and artificial intelligence technology, the research and application of heterogeneous vehicle cooperative control has gradually become a research hotspot in the field of intelligent transportation.

[0003] In the research of cooperative control of heterogeneous vehicles, vehicle platooning control is an important research direction. Platooning control achieves a tight and stable driving formation by coordinating the speed and spacing of multiple vehicles, thereby improving road utilization and driving safety. However, the differences in physical characteristics, dynamic models, and control requirements among heterogeneous vehicles present numerous challenges to cooperative control. For example, different types of vehicles have different inertia, air resistance, and mechanical drag coefficients, making traditional fixed-parameter control methods difficult to adapt to complex traffic environments and dynamically changing task requirements. Furthermore, vehicles may be affected by various uncertainties during operation, such as road conditions, weather conditions, and traffic flow changes, which further increase the complexity of cooperative control of heterogeneous vehicle systems.

[0004] While existing heterogeneous vehicle cooperative control methods have addressed some issues to a certain extent, several shortcomings remain. Firstly, regarding the adaptability of control strategies, existing methods, such as PID control and model predictive control (MPC), are mostly based on fixed controller parameters or preset models. When vehicle parameters change due to variations in load, fuel consumption, component aging, or even actuator failure, these methods lack effective real-time adaptive adjustment mechanisms. Therefore, they struggle to maintain optimal control performance under complex and changing real-world conditions, often leading to decreased control accuracy or system instability. Furthermore, traditional preset performance control formulas can only roughly constrain the error range and cannot guarantee convergence within a fixed time, causing the system response speed to be significantly affected by the initial state. In dynamically changing traffic environments, this can lead to excessively frequent or untimely control signal updates, increasing communication burden and computational resource consumption. Existing literature (Finite-time sliding mode pre-set performance queue control for connected vehicles [J]. Control Theory & Applications, 2023; Fixed-time global pre-set performance vehicle queue control considering actuator nonlinearity [J]. Acta Automatica Sinica, 2024.) designs finite-time and fixed-time sliding mode controllers; however, they do not consider event-triggered strategies. Secondly, existing methods lack the ability to adaptively adjust to changes in vehicle dynamics model parameters, making it difficult to maintain optimal control performance under different operating conditions. Recent literature (Event-triggered adaptive NN tracking control for nonlinear systems with asymmetric time-varying output constraints and application to an AUVs[J]. IEEE Transactions on Vehicular Technology, 2024; Robust output regulation of linear uncertain systems by dynamic event-triggered output feedback control[J]. IEEE Transactions on Cybernetics, 2023.) further confirms that most existing event-triggered mechanisms are based on periodic, fixed time intervals or simple threshold judgments for data updates and control signal adjustments. This strategy may lead to overly frequent or untimely control signal updates in dynamically changing traffic environments, thereby increasing communication burden and computational resource consumption. Furthermore, the research objects in existing technologies are mostly homogeneous intelligent agents.To address this issue, existing technologies employ adaptive methods and fuzzy control to adjust the thresholds for dynamic event triggering. However, when dealing with various complex traffic scenarios, complex constraints, and multi-objective optimization, these methods typically rely on explicit mathematical models or specific parameter variation patterns, lacking self-learning and updating capabilities. Consequently, their robustness, adaptability, and flexibility in dealing with complex environments and constraints are limited.

[0005] It is evident that existing technologies often lack sufficient flexibility and adaptability when dealing with complex traffic scenarios and task requirements, failing to effectively balance the contradiction between control performance and resource utilization efficiency. Therefore, developing a heterogeneous vehicle cooperative control system and method based on a dynamic event-triggered mechanism is of significant theoretical and practical importance for improving the performance of heterogeneous vehicle cooperative control, reducing communication and computing costs, and enhancing the system's adaptability and robustness. Summary of the Invention

[0006] This invention addresses the shortcomings of existing technologies by proposing a heterogeneous vehicle cooperative control system and method based on a dynamic event triggering mechanism. This significantly improves the control performance and resource utilization efficiency of heterogeneous vehicle platooning, while effectively handling uncertainties, internal and external disturbances, and actuator saturation in complex traffic scenarios. It provides technical support and theoretical basis for the technological advancement of intelligent vehicles and the development of intelligent transportation.

[0007] To address the shortcomings of existing technologies, the present invention proposes the following technical solution:

[0008] A heterogeneous vehicle cooperative control system based on a dynamic event triggering mechanism comprises four parts: a heterogeneous vehicle platooning system perception module, a heterogeneous vehicle platooning control system, a heterogeneous vehicle drive and transmission system, and a vehicle network communication module.

[0009] The perception module of the heterogeneous vehicle platooning system with fixed-time convergence disturbance estimation and auxiliary system includes: a speed sensor, a position sensor (GPS signal), a slope sensor, a vehicle mass estimation module, a fixed-time convergence disturbance estimation module caused by unknown disturbances and model uncertainties, and a fixed-time convergence auxiliary system for compensating actuator saturation errors and control command deviations in event triggering mechanisms, which are connected in sequence to the heterogeneous vehicle platooning control system.

[0010] The heterogeneous vehicle platooning control system, which features funnel control, sliding mode control, and dynamic event triggering mechanisms within a unified framework of fixed-time convergence, connects to the power control units of electric or fuel-powered vehicles and issues control commands to them. These control commands are then transformed into driving torques for the vehicles and transmitted to the drive wheels through the drive and transmission systems of the heterogeneous vehicles. Simultaneously, the heterogeneous vehicle platooning control system connects to a vehicle-to-everything (V2X) communication module and establishes network communication with adjacent vehicles.

[0011] The electric and fuel-powered drive and transmission systems of the heterogeneous vehicles include those configured within the electric drive system or fuel-powered drive system and its transmission system. Electric vehicles include a power battery, a drive motor controller, a drive motor / transmission, and half-shafts / drive wheels connected in sequence; fuel-powered vehicles include an engine controller, an engine / transmission, and half-shafts / drive wheels connected in sequence.

[0012] The vehicle network communication module is installed on the upper side of the heterogeneous vehicle platooning control system and includes a wireless transmitting module and a wireless receiving module, which are electrically connected to the heterogeneous vehicle platooning control system.

[0013] The aforementioned vehicle mass estimation module is used to estimate vehicle mass. This module is designed to address the frequent changes in vehicle mass during operation, accurately reflecting these changes. This enables more precise environmental perception in heterogeneous vehicle platooning systems, resulting in smaller errors and better performance in heterogeneous vehicle platooning control.

[0014] The drive systems of the heterogeneous vehicles include electric drive systems, fuel drive systems, and hybrid drive systems with different characteristics.

[0015] In the heterogeneous vehicle platooning system, each vehicle communicates with the network through a wireless transmitter and a wireless receiver module located at the front of the vehicle, and sequentially obtains information streams from adjacent vehicles, including the position, speed, and acceleration information of adjacent vehicles.

[0016] Furthermore, the auxiliary system for compensating actuator saturation error and control command deviation in the event-triggered mechanism with fixed-time convergence can achieve faster compensation for actuator saturation error and control command deviation in the event-triggered mechanism. This allows the heterogeneous vehicle platooning control system to provide compensation information in a shorter time, improves the real-time performance of heterogeneous vehicle platooning tracking error control, and is unaffected by initial conditions, resulting in faster convergence speed.

[0017] This invention also provides a heterogeneous vehicle cooperative control method based on a dynamic event triggering mechanism, comprising the following steps:

[0018] S1. Establish a dynamic model for the nonlinear heterogeneous vehicle system;

[0019] S2. Calculation of vehicle-to-vehicle tracking distance error, fixed-time funnel control, error conversion and construction of sliding surface;

[0020] S3. Construct a fixed-time-based auxiliary system to compensate for actuator nonlinearity errors and control input errors caused by event triggering;

[0021] S4. Construct adaptive laws for external and internal disturbances;

[0022] S5. Design a fixed-time formation controller;

[0023] S6. Design a dynamic event triggering control mechanism;

[0024] S7. Determine whether the event triggering condition is met. If it is met, return to step S1; otherwise, proceed to step S8.

[0025] S8. Determine if the termination condition is met. If it is met, end the control flow; otherwise, return to step S1.

[0026] The nonlinear heterogeneous vehicle system dynamics model in step S1 considers a heterogeneous vehicle system consisting of one lead vehicle and N follower vehicles, totaling N+1 vehicles. For the i-th vehicle in the heterogeneous vehicle system, i ∈ {1,...,N}, the heterogeneous vehicle dynamics model is established as follows:

[0027] (1a)

[0028] (1b)

[0029] (1c)

[0030] (1d)

[0031] The kinematic model of vehicle i can be extended to a third-order system: position The derivative of velocity The derivative of velocity is acceleration. The derivative of acceleration is determined by the event-triggered control command. System dynamic items (A function containing velocity and acceleration) and total disturbance A joint decision, in which This is the rotational mass correction factor. For quality, It is a time constant. For air density, The frontal area of ​​vehicle i It is the air drag coefficient of vehicle i. It is the rolling resistance of vehicle i. Let i be the rolling resistance coefficient of vehicle i. It is gravitational acceleration. The slope representing vehicle i This is the gravity gradient resistance of vehicle i, and the road resistance coefficient of vehicle i is... ;

[0032] (2)

[0033] (3)

[0034] Wherein, the saturation control input is defined. When the original control quantity Exceeding the upper and lower limits or When the saturation input is at its boundary value, it takes that value; otherwise, it equals the original control input. Control input estimation error. For saturated input Its event trigger control command The difference is used to describe the signal deviation caused by the event triggering mechanism;

[0035] Furthermore, Divided into the following two parts:

[0036] (4)

[0037] in It is a known term. It is an unknown term. (1c) can be written in the following form:

[0038] (5)

[0039] In the formula, the system dynamic term It can be decomposed into known nominal terms. and unknown disturbance terms The formula for the derivative of acceleration is further simplified to: given the dynamic term and including saturation error. and disturbance error The control input error and the total disturbance including model uncertainty and external disturbances The combination, in which To control the scaling factor;

[0040] Step S2, which involves calculating the vehicle-to-vehicle tracking distance error, controlling the fixed-time funnel, and converting errors to construct the sliding surface, comprises the following four steps:

[0041] S21: Calculation of vehicle-to-vehicle tracking distance error

[0042] Using a constant-time spacing strategy, the error in the spacing between adjacent heterogeneous vehicles is as follows:

[0043] (6)

[0044] Among them, tracking error Position of the vehicle in front Current vehicle position The difference, minus the current vehicle length Time of departure from the locomotive With current vehicle speed The product and minimum safe distance , used to describe the spacing control deviation of vehicle queues;

[0045] S22: Fixed-time funnel control

[0046] Design a lower bound for a fixed-time funnel function. and the Upper Realm ,in For attenuation term, , These are boundary coefficients. This is a saturation compensation term, which is adaptively adjusted based on the actuator's saturation state. When... When, the function dynamically adjusts according to the initial error sign; when At the same time, maintain constant boundaries to ensure tracking error Constrained within the funnel area, the funnel control function is shown in the following equation:

[0047] (7)

[0048] (8)

[0049] In the formula , , , , , , , It is the initial tracking error and satisfies ;

[0050] S23: Tracking Error Conversion

[0051] To avoid the performance boundary limitations of the aforementioned funnel control, the tracking error can be transformed into an unconstrained form, as shown below:

[0052] (9)

[0053] In the formula, The value range is (0,1). This can be expressed as the product of the transformation function and the sum of the funnel boundaries minus the lower bound, obtained through logarithmic transformation. Its convergence and Equivalent, facilitating controller design, tracking error will occur within a given time. Converging to the specified range ;

[0054] Furthermore, the tracking error can be written as:

[0055] (10)

[0056] To proceed with the design of the heterogeneous vehicle cooperative controller, the derivative of the above equation yields...

[0057] but and Write it in the following form:

[0058] (11)

[0059] (12)

[0060] Among them, conversion error The first derivative is derived from the proportionality constant. It is related to the funnel boundary and error, and the rate of change of error. and compensation terms containing funnel boundary derivative information The second derivative further includes the second derivative of the error and the boundary perturbation term. and its derivative, and The derivative terms provide a dynamic basis for the design of sliding mode controllers;

[0061] S24: Construct the sliding surface

[0062] The sliding surface that converges within a fixed time is constructed as follows:

[0063] (13a)

[0064] (13b)

[0065] In the formula, , , , , and , , , , , It is a small constant. , ;

[0066] To facilitate subsequent stability verification, the sliding surface is constructed as follows:

[0067] (14)

[0068] To prove the stability of the queue, a coupled sliding surface is defined. For the first n-1 vehicles, its value is the sliding surface of the current vehicle. sliding surface of the rear vehicle The coupling, where the coefficients The coupling strength is defined; the last car is only related to its own sliding surface. Coupling design ensures that errors do not propagate within the queue. and The convergence is consistent.

[0069] Furthermore, in step S3, a fixed-time auxiliary system compensation actuator nonlinearity error and control input error caused by event triggering are constructed, and its design is as follows:

[0070] (15)

[0071] In the formula , and All are positive design parameters.

[0072] Furthermore, in step S4, the adaptive laws for external and internal disturbances are constructed as follows:

[0073] (16)

[0074] Furthermore, in step S5, firstly, Given as follows:

[0075] (17)

[0076] At the same time,

[0077] (18)

[0078] The design of a fixed-time formation controller for a heterogeneous vehicle system is as follows:

[0079] (19)

[0080] In the formula , , , , , , , , All are positive constants. The design concept is the control law for the i-th vehicle. Designed to include coupling sliding surfaces The power term (coefficient) , power , ), nonlinear damping term ( For system dynamic parameters, (small constant), disturbance compensation term ( To estimate the disturbance estimation error, the hyperbolic tangent function is used to suppress jitter, and the auxiliary system state is also considered. The combination of. Overall through. (Control scaling factor) (Coupling strength) (Time interval) and Scaling is performed using a conversion scaling factor to ensure the reasonableness of the control input.

[0081] Furthermore, the dynamic event triggering control mechanism designed in step S6 is as follows:

[0082] S61: First, design external auxiliary variables. :

[0083] (20)

[0084] further,

[0085] (twenty one)

[0086] Among them, the dynamic event triggering mechanism introduces auxiliary variables. Its derivative includes a decay term (coefficient) ) and adjustment terms; among the adjustment terms, As weight, This is the proportionality coefficient. , For items related to sliding surfaces, To control the squared error. It is defined as a combination of two power terms ( , (where is a coefficient) is used to dynamically adjust the trigger threshold.

[0087] S62: The dynamic event triggering mechanism is designed as follows:

[0088] (twenty two)

[0089] (twenty three)

[0090] In the formula, It is the actual control signal acting on heterogeneous vehicle i. The range is (0,1). , ; This is the deviation between the expected control signal and the actual control signal caused by the event-triggered mechanism. , , is the controller's update time, that is, when (23) is triggered, the time will be marked as ,and It will be applied to heterogeneous vehicle i, during the time interval At this time, the control signal will be held at a constant value, where, , , and Design parameters; dynamic event triggering time Defined as the maximum time when the triggering condition is met, where For threshold coefficient, The square of the current control error. For the threshold reference related to the sliding surface, when the expression on the left is less than or equal to the auxiliary variable If the control is not updated, then no control update is triggered; otherwise, the control value is triggered and updated.

[0091] Corresponding to the dynamic event triggering mechanism described above, when designing external auxiliary variables... When it is a constant or its derivative is 0, use This leads to the following static event triggering mechanism:

[0092] (twenty four)

[0093] (25)

[0094] The static event triggering conditions are similar to those of the dynamic mechanism, but the threshold coefficient... It is a constant and does not dynamically adjust with the system state. When the square of the control error... With benchmark item The difference multiplied by less than or equal to If the current control value is maintained, then an update is triggered.

[0095] In step S7, determining whether the event triggering condition is met is (23) when a dynamic event triggering mechanism is used and (25) when a static event triggering mechanism is used. If the condition is met, the process returns to step S1 and is executed again. If the condition is not met, the process proceeds to step S8.

[0096] Step S8, determining whether the termination condition is met, means that the heterogeneous vehicle cooperative control system receives a termination command or achieves the control objective; the termination command refers to a shutdown signal; the control objective refers to simultaneously satisfying the following:

[0097] (1) At a given time To achieve fixed-time bicycle stability, that is, when hour, or ;

[0098] (2) Achieve string stability with fixed-time convergence, that is, within a given time. Subsequently, vehicle spacing tracking error It will not grow any further, that is, , yes The Laplace function, ;

[0099] (3) Achieve fixed-time funnel control:

[0100]

[0101] In other words To converge to , , It is the boundary controlled by the funnel;

[0102] (4) Implement dynamic event triggering control and exclude Zeno's behavior;

[0103] If the above termination conditions are met, the control flow ends; otherwise, return to step S1 and continue executing S1-S8.

[0104] Furthermore, the parameter adaptation in step S6 is , , and The design parameters of the event triggering mechanism are based on the vehicle's , , , , , and The state information is used to adaptively determine the outcome, but it requires... , , , And simultaneously satisfy This allows for dynamic event triggering control and excludes Zeno's behavior;

[0105] The aforementioned The minimum trigger interval threshold is calculated by recording the time difference between two consecutive dynamic event triggers. Based on the stability analysis results of vehicle cooperative control, the safety value can be pre-set through Lyapunov function derivation or empirical safety value determined by simulation experiment.

[0106] The beneficial effects of this invention are as follows:

[0107] (1) By redesigning the combined formula of performance constraints and event triggering, the inherent defects of the traditional preset performance formula in the logic of performance constraints and event triggering are solved.

[0108] Existing technologies for collaborative control of heterogeneous vehicle systems, such as PID control and MPC control, lack the ability to adaptively adjust to changes in vehicle parameters or malfunctions. This makes it difficult to guarantee the steady-state and transient performance of heterogeneous vehicle systems and to maintain optimal control performance under different operating conditions. Furthermore, traditional preset performance control formulas can only roughly constrain the error range and cannot guarantee convergence within a fixed time, resulting in the system response speed being significantly affected by the initial state. Simultaneously, traditional event-triggered mechanisms often use fixed thresholds, which are out of touch with the actual performance requirements of the system. This leads to either performance exceeding limits due to overly lenient thresholds or resource waste due to overly strict thresholds.

[0109] This invention introduces a funnel performance formula based on a fixed-time convergence framework, relating it to the convergence time. The directly associated dynamic boundary parameters constrain the tracking error within a funnel shape formed by two performance functions, ensuring the tracking error converges within a fixed time. This allows the formula to rigidly constrain the error to converge to the target range within a specified time, overcoming the uncertainty of convergence time in traditional preset performance formulas. This improves the transient performance of heterogeneous vehicle systems to overcome scenarios involving vehicle parameter variations and nonlinear disturbances. Then, a fixed-time convergence adaptive law and sliding mode control are employed to enhance the stability of the heterogeneous vehicle system, further addressing the issues of poor robustness and adaptability under nonlinear disturbances and complex scenarios. Simultaneously, the dynamic event triggering mechanism formula quantifies and correlates the trigger threshold with real-time error and the rate of error change, and through dynamic adjustment, achieves dynamic adaptation between triggering conditions and performance requirements. This avoids the inherent contradictions of traditional static threshold formulas, providing a precise quantitative basis for balancing system performance and resource consumption.

[0110] (2) By using a coordinated control method for performance and resources, the problem of the disconnect between the two in traditional technologies is overcome.

[0111] Most existing control methods rely on periodic control and data updates and control signal adjustments based on fixed time intervals or simple threshold judgments. In dynamically changing traffic environments, this strategy can lead to overly frequent or untimely control signal updates, increasing communication burden and computational resource consumption. Furthermore, existing event-triggered mechanisms in heterogeneous vehicle systems are often static, or even dynamic triggering mechanisms cannot guarantee convergence within a specified time, thus failing to meet the real-time requirements of heterogeneous vehicle systems. In existing technologies, performance control and resource optimization are often implemented independently: when using preset performance control, accuracy is maintained solely through high-frequency computation without considering resource consumption; when using event-triggered mechanisms, computation is reduced solely through simple thresholds without being linked to performance constraints, leading to the contradiction that high accuracy inevitably comes with high resource consumption, or low resource consumption inevitably sacrifices performance. To address this issue, existing technologies employ adaptive methods, fuzzy control, or directly adjust the thresholds for dynamic event triggering. However, when dealing with various complex traffic scenarios, complex constraints, and multi-objective optimization, these methods typically rely on explicit mathematical models or specific parameter variation patterns, lacking self-learning and updating capabilities, resulting in limited robustness, adaptability, and flexibility in handling complex environments and constraints.

[0112] This invention constructs a collaborative control logic for performance constraints and resource allocation: a preset performance formula first defines an insurmountable error convergence boundary, with a convergence time not exceeding [a certain threshold]. The dynamic triggering mechanism formula, within this boundary, adaptively adjusts the timing of control law updates based on real-time error status. High-frequency triggering ensures performance when the error approaches the boundary, while low-frequency triggering conserves resources when the error stabilizes. This collaborative method creates a closed loop between performance control and resource optimization, resolving the disconnect between the two in traditional technologies and minimizing resource consumption while maintaining performance within limits. This allows for both constraint of dynamic errors in multivariable systems to prevent oscillations and effective reduction of steady-state triggering frequency. Further improving robustness and adaptability, in heterogeneous vehicle collaborative control, it's crucial to consider not only avoiding Zeno behavior but also tracking error, communication costs, and system stability. To overcome these issues, trigger thresholds can be dynamically and in real-time adjusted to balance multiple objectives, ensuring both control performance and resource utilization efficiency of the heterogeneous vehicle system.

[0113] (3) Thus, the synergistic optimization of transient performance and computational resource consumption is achieved, breaking through the inherent limitations of existing technologies.

[0114] Existing technologies cannot simultaneously meet the dual requirements of high-precision transient response and low computational resource consumption within a fixed time: when using preset performance control alone, although the accuracy can be improved, the amount of computation is large, resulting in a high processor load; when using event triggering mechanism alone, although resources can be saved, the transient performance is unstable, such as large overshoot and slow convergence, which can easily lead to insufficient real-time performance or control failure risks in scenarios such as autonomous driving and robot control.

[0115] This invention achieves a breakthrough by improving the formula construction and innovating the collaborative control method: compared with traditional preset performance control, while ensuring a fixed time... Under the premise that the internal error converges to the target range and the overshoot is reduced to the target range, the computational resource consumption is effectively reduced. Compared with the traditional event-triggered mechanism, the dynamic event-triggered mechanism of control signals u1, u2, u3, u4, and u5 accounts for 12.48%, 36.75%, 65.63%, 62.94%, and 91.24% of the static event-triggered mechanism, respectively, and the total number of events of the dynamic triggering mechanism is reduced by about 57.16% compared with the static triggering mechanism. While reducing the average number of triggering events of the total control signals by nearly 60%, it can ensure that its convergence time is not affected by the initial state. This effect overcomes the core contradiction of existing technologies and significantly improves the practicality and reliability of the control method in real-time and resource-constrained scenarios.

[0116] Therefore, existing cooperative control methods for heterogeneous vehicle systems often lack sufficient robustness and adaptability when dealing with changes in vehicle parameters and nonlinear disturbances, complex traffic scenarios, and task requirements, failing to effectively balance control accuracy and resource utilization efficiency. This invention provides a cooperative control system and method for heterogeneous vehicles based on a dynamic event-triggered mechanism. It utilizes a funnel control method within a fixed-time theoretical framework to ensure the transient performance of the heterogeneous vehicle system, combines it with a fixed-time sliding mode control method to improve the steady-state performance, and finally employs a fixed-time convergent dynamic event-triggered mechanism to reduce resource usage. The system comprises four parts: a heterogeneous vehicle platooning system perception module, a heterogeneous vehicle platooning control system, heterogeneous vehicle drive and transmission systems, and a vehicle-to-everything (V2X) communication module. The method first establishes a dynamic model of a nonlinear heterogeneous vehicle system; then, it calculates the vehicle-to-vehicle tracking distance error, performs fixed-time funnel control, error transformation, and constructs a sliding mode surface; it constructs a fixed-time auxiliary system to compensate for actuator nonlinearity errors and control input errors caused by event triggering; it constructs adaptive laws for external and internal disturbances; it designs a fixed-time formation controller; it designs a dynamic event-triggered control mechanism; it determines whether the event triggering condition is met; if so, it returns to step S1 and re-executes; otherwise, it proceeds to step S8; finally, it determines whether the termination condition is met; if so, the control flow ends; otherwise, it returns to step S1 to continue executing the control flow. This system and method ensure the performance of heterogeneous vehicle cooperative control while significantly reducing communication and computation costs, effectively enhancing the system's adaptability and robustness. Attached Figure Description

[0117] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0118] Figure 1 Heterogeneous vehicle architecture for fuel-powered vehicles;

[0119] Figure 2 Heterogeneous vehicle electric vehicle architecture;

[0120] Figure 3 Control scheme for heterogeneous vehicle platooning system;

[0121] Figure 4 Flowchart of the control method for a heterogeneous vehicle platooning system;

[0122] Figure 5 The control effects of the fixed-time formation controller proposed in this invention are: (a) position; (b) speed; (c) acceleration; (d) tracking error of vehicle i (i = 1, 2, 3, 4, 5).

[0123] Figure 6Simulation results of the trigger time intervals of each control input in the static ETC of heterogeneous vehicle control proposed in this invention: (a) u1; (b) u2; (c) u3; (d) u4; (e) u5;

[0124] Figure 7 Simulation results of the trigger time intervals of each control input in the dynamic ETC of heterogeneous vehicle control proposed in this invention: (a) u1; (b) u2; (c) u3; (d) u4; (e) u5. Detailed Implementation

[0125] The following are specific embodiments of the present invention, and the technical solution of the present invention will be further described in conjunction with the accompanying drawings.

[0126] like Figure 1 and Figure 2 As shown, a heterogeneous vehicle cooperative control system based on a dynamic event triggering mechanism includes four parts: a heterogeneous vehicle platooning system perception module, a heterogeneous vehicle platooning control system, a heterogeneous vehicle drive and transmission system, and a vehicle network communication module.

[0127] The perception module of the heterogeneous vehicle platooning system with fixed-time convergence disturbance estimation and auxiliary system includes: a speed sensor, a position sensor (GPS signal), a slope sensor, a vehicle mass estimation module, a fixed-time convergence disturbance estimation module caused by unknown disturbances and model uncertainties, and a fixed-time convergence auxiliary system for compensating actuator saturation errors and control command deviations in event triggering mechanisms, which are connected in sequence to the heterogeneous vehicle platooning control system.

[0128] The heterogeneous vehicle platooning control system, which features funnel control, sliding mode control, and dynamic event triggering mechanisms within a unified framework of fixed-time convergence, connects to the power control units of electric or fuel-powered vehicles and issues control commands to them. These control commands are then transformed into driving torques for the vehicles and transmitted to the drive wheels through the drive and transmission systems of the heterogeneous vehicles. Simultaneously, the heterogeneous vehicle platooning control system connects to a vehicle-to-everything (V2X) communication module and establishes network communication with adjacent vehicles.

[0129] like Figure 1 and Figure 2 As shown, the electric and fuel-powered drive systems and transmission systems of the heterogeneous vehicles include those disposed within the electric drive system or fuel-powered drive system and its transmission system. Electric vehicles include a power battery, a drive motor controller, a drive motor / transmission, and half-shafts / drive wheels connected in sequence; fuel-powered vehicles include an engine controller, an engine / transmission, and half-shafts / drive wheels connected in sequence.

[0130] The vehicle network communication module is installed on the upper side of the heterogeneous vehicle platooning control system and includes a wireless transmitting module and a wireless receiving module, which are electrically connected to the heterogeneous vehicle platooning control system.

[0131] The vehicle mass estimation module is designed to address the frequent changes in vehicle mass during driving. It can accurately reflect these changes, thereby enabling more precise environmental perception in heterogeneous vehicle platooning systems, resulting in smaller errors and better performance in heterogeneous vehicle platooning control.

[0132] The drive systems of the heterogeneous vehicles include electric drive systems, fuel drive systems, and hybrid drive systems with different characteristics.

[0133] In the heterogeneous vehicle platooning system, each vehicle communicates with the network through a wireless transmitter and a wireless receiver module located at the front of the vehicle, and sequentially obtains information streams from adjacent vehicles, including the position, speed, and acceleration information of adjacent vehicles.

[0134] Furthermore, the auxiliary system for compensating actuator saturation error and control command deviation in the event-triggered mechanism with fixed-time convergence can achieve faster compensation for actuator saturation error and control command deviation in the event-triggered mechanism. This allows the heterogeneous vehicle platooning control system to provide compensation information in a shorter time, improves the real-time performance of heterogeneous vehicle platooning tracking error control, and is unaffected by initial conditions, resulting in faster convergence speed.

[0135] Control schemes for heterogeneous vehicle platooning systems, such as Figure 3 As shown, the present invention also provides a heterogeneous vehicle cooperative control method based on a dynamic event triggering mechanism, such as... Figure 4 As shown, it includes the following steps:

[0136] S1. Establish a dynamic model for the nonlinear heterogeneous vehicle system;

[0137] S2. Calculation of vehicle-to-vehicle tracking distance error, fixed-time funnel control, error conversion and construction of sliding surface;

[0138] S3. Construct a fixed-time-based auxiliary system to compensate for actuator nonlinearity errors and control input errors caused by event triggering;

[0139] S4. Construct adaptive laws for external and internal disturbances;

[0140] S5. Design a fixed-time formation controller;

[0141] S6. Design a dynamic event triggering control mechanism;

[0142] S7. Determine whether the event triggering condition is met. If it is met, return to step S1; otherwise, proceed to step S8.

[0143] S8. Determine if the termination condition is met. If it is met, end the control flow; otherwise, return to step S1.

[0144] The nonlinear heterogeneous vehicle system dynamics model in step S1 considers a heterogeneous vehicle system consisting of one lead vehicle and N follower vehicles, totaling N+1 vehicles. For the i-th vehicle in the heterogeneous vehicle system, i ∈ {1,...,N}, the heterogeneous vehicle dynamics model is established as follows:

[0145] (1a)

[0146] (1b)

[0147] (1c)

[0148] (1d)

[0149] The kinematic model of vehicle i can be extended to a third-order system: position The derivative of velocity The derivative of velocity is acceleration. The derivative of acceleration is determined by the event-triggered control command. System dynamic items (A function containing velocity and acceleration) and total disturbance A joint decision, in which This is the rotational mass correction factor. For quality, It is a time constant. For air density, The frontal area of ​​vehicle i It is the air drag coefficient of vehicle i. It is the rolling resistance of vehicle i. Let i be the rolling resistance coefficient of vehicle i. It is gravitational acceleration. The slope representing vehicle i This is the gravity gradient resistance of vehicle i, and the road resistance coefficient of vehicle i is... ;

[0150] (2)

[0151] (3)

[0152] Wherein, the saturation control input is defined. When the original control quantity Exceeding the upper and lower limits or When the saturation input is at its boundary value, it takes that value; otherwise, it equals the original control input. Control input estimation error. For saturated input Its event trigger control command The difference is used to describe the signal deviation caused by the event triggering mechanism;

[0153] Furthermore, Divided into the following two parts:

[0154] (4)

[0155] in It is a known term. It is an unknown term. (1c) can be written in the following form:

[0156] (5)

[0157] In the formula, the system dynamic term It can be decomposed into known nominal terms. and unknown disturbance terms The acceleration derivative formula is further simplified as follows: Given the dynamic term and control input error (including saturation error) and disturbance error ) and total disturbance A combination of (model uncertainty and external disturbances), where To control the scaling factor;

[0158] Step S2, which involves calculating the vehicle-to-vehicle tracking distance error, controlling the fixed-time funnel, and converting errors to construct the sliding surface, comprises the following four steps:

[0159] S21: Calculation of vehicle-to-vehicle tracking distance error

[0160] Using a constant-time spacing strategy, the error in the spacing between adjacent heterogeneous vehicles is as follows:

[0161] (6)

[0162] Among them, tracking error Position of the vehicle in front Current vehicle position The difference, minus the current vehicle length Time of departure from the locomotive With current vehicle speed The product and minimum safe distance , used to describe the spacing control deviation of vehicle queues;

[0163] S22: Fixed-time funnel control

[0164] Design a fixed-time funnel function (Lower Boundary) and (Upper boundary), in which For attenuation term, , These are boundary coefficients. This is a saturation compensation term (adaptively adjusted based on the actuator's saturation state). When When, the function dynamically adjusts according to the initial error sign; when At the same time, maintain constant boundaries to ensure tracking error Constrained within the funnel area, the funnel control function is shown in the following equation:

[0165] (7)

[0166] (8)

[0167] In the formula , , , , , , , It is the initial tracking error and satisfies ;

[0168] S23: Tracking Error Conversion

[0169] To avoid the performance boundary limitations of the aforementioned funnel control, the tracking error can be transformed into an unconstrained form, as shown below:

[0170] (9)

[0171] In the formula, The value range is (0,1). This can be expressed as the product of the transformation function and the sum of the funnel boundaries minus the lower bound, obtained through logarithmic transformation. Its convergence and Equivalent, facilitating controller design, tracking error will occur within a given time. Converging to the specified range ;

[0172] Furthermore, the tracking error can be written as:

[0173] (10)

[0174] To proceed with the design of the heterogeneous vehicle cooperative controller, the derivative of the above equation yields...

[0175] but and Write it in the following form:

[0176] (11)

[0177] (12)

[0178] Among them, conversion error The first derivative is derived from the proportionality constant. (Related to funnel boundary and error), error change rate and compensation items (Including funnel boundary derivative information) determines; the second derivative further includes the error second derivative and boundary perturbation terms. and its derivative, and The derivative terms provide a dynamic basis for the design of sliding mode controllers;

[0179] S24: Construct the sliding surface

[0180] The sliding surface that converges within a fixed time is constructed as follows:

[0181] (13a)

[0182] (13b)

[0183] In the formula, , , , , and , , , , , It is a small constant. , ;

[0184] To facilitate subsequent stability verification, the sliding surface is constructed as follows:

[0185] (14)

[0186] To prove the stability of the queue, a coupled sliding surface is defined. For the first n-1 vehicles, its value is the sliding surface of the current vehicle. sliding surface of the rear vehicle Coupling (coefficient) (For coupling strength); the last car is only related to its own sliding surface. Coupling design ensures that errors do not propagate within the queue, which is evident. and The convergence is consistent.

[0187] Furthermore, in step S3, a fixed-time auxiliary system compensation actuator nonlinearity error and control input error caused by event triggering are constructed, and its design is as follows:

[0188] (15)

[0189] In the formula , and All are positive design parameters.

[0190] Furthermore, in step S4, the adaptive laws for external and internal disturbances are constructed as follows:

[0191] (16)

[0192] Furthermore, in step S5, firstly, Given as follows:

[0193] (17)

[0194] At the same time,

[0195] (18)

[0196] The design of a fixed-time formation controller for a heterogeneous vehicle system is as follows:

[0197] (19)

[0198] In the formula , , , , , , , , All are positive constants. The design concept is the control law for the i-th vehicle. Designed to include coupling sliding surfaces The power term (coefficient) , power , ), nonlinear damping term ( For system dynamic parameters, (small constant), disturbance compensation term ( To estimate the disturbance estimation error, the hyperbolic tangent function is used to suppress jitter, and the auxiliary system state is also considered. The combination of. Overall through. (Control scaling factor) (Coupling strength) (Time interval) and Scaling is performed using a conversion scaling factor to ensure the reasonableness of the control input.

[0199] Furthermore, the dynamic event triggering control mechanism designed in step S6 is as follows:

[0200] S61: First, design external auxiliary variables. :

[0201] (20)

[0202] further,

[0203] (twenty one)

[0204] Among them, the dynamic event triggering mechanism introduces auxiliary variables. Its derivative includes a decay term (coefficient) ) and adjustment items ( As weight, This is the proportionality coefficient. , For items related to sliding surfaces, To control the squared error. It is defined as a combination of two power terms ( , (where is a coefficient) is used to dynamically adjust the trigger threshold.

[0205] S62: The dynamic event triggering mechanism is designed as follows:

[0206] (twenty two)

[0207] (twenty three)

[0208] In the formula, It is the actual control signal acting on heterogeneous vehicle i. The range is (0,1). , ; This is the deviation between the expected control signal and the actual control signal caused by the event-triggered mechanism. , , is the controller's update time, that is, when (23) is triggered, the time will be marked as ,and It will be applied to heterogeneous vehicle i, during the time interval At this time, the control signal will be held at a constant value, where, , , and Design parameters; dynamic event triggering time Defined as the maximum time when the triggering condition is met, where For threshold coefficient, The square of the current control error. For the threshold reference related to the sliding surface, when the expression on the left is less than or equal to the auxiliary variable If the control is not updated, then no control update is triggered; otherwise, the control value is triggered and updated.

[0209] Corresponding to the dynamic event triggering mechanism described above, when designing external auxiliary variables... When it is a constant or its derivative is 0, use This leads to the following static event triggering mechanism:

[0210] (twenty four)

[0211] (25)

[0212] The static event triggering conditions are similar to those of the dynamic mechanism, but the threshold coefficient... It is a constant and does not dynamically adjust with the system state. When the square of the control error... With benchmark item The difference multiplied by less than or equal to If the current control value is maintained, then an update is triggered.

[0213] In step S7, determining whether the event triggering condition is met is (23) when a dynamic event triggering mechanism is used and (25) when a static event triggering mechanism is used. If the condition is met, the process returns to step S1 and is executed again. If the condition is not met, the process proceeds to step S8.

[0214] Step S8, determining whether the termination condition is met, means that the heterogeneous vehicle cooperative control system receives a termination command or achieves the control objective; the termination command refers to a shutdown signal; the control objective refers to simultaneously satisfying the following:

[0215] (1) At a given time To achieve fixed-time bicycle stability, that is, when hour, or ;

[0216] (2) Achieve string stability with fixed-time convergence, that is, within a given time. Subsequently, vehicle spacing tracking error It will not grow any further, that is, , yes The Laplace function, ;

[0217] (3) Achieve fixed-time funnel control:

[0218]

[0219] In other words To converge to , , It is the boundary controlled by the funnel;

[0220] (4) Implement dynamic event triggering control and exclude Zeno's behavior;

[0221] If the above termination conditions are met, the control flow ends; otherwise, return to step S1 and continue executing steps S1-S8.

[0222] Furthermore, the parameter adaptation in step S6 is , , and The design parameters of the event triggering mechanism are based on the vehicle's , , , , , and The state information is used to adaptively determine the outcome, but it requires... , , , And simultaneously satisfy This allows for dynamic event triggering control and excludes Zeno's behavior;

[0223] The following demonstrates that the method of the heterogeneous vehicle cooperative control system based on the dynamic event triggering mechanism proposed in this invention converges in a fixed time.

[0224] Step 1: Boundedness

[0225] First, construct the following Lyapunov function. :

[0226] (26)

[0227] In the formula, the Lyapunov function is constructed to analyze the system stability. , including the square term of the coupled sliding surface Perturbation estimation error squared term Auxiliary system state square term and dynamically triggered auxiliary variables The coefficients of each term are all 1 / 2 to simplify the derivation. Its time derivative... Composed of the derivatives of each component, it reflects the trend of energy change in the system and is used for subsequent stability proofs. Differentiation yields:

[0228] (27)

[0229] From formulas (4), (6), (12) and (13b), we can obtain:

[0230] (28)

[0231] Coupled sliding surface The derivative formula describes its dynamic characteristics, where Scaling factor ( For coupling strength, For time header spacing, (for conversion ratio) For saturation control input items ( To control the scaling factor). For the total disturbance, This is a dynamic compensation term for the system, which reflects the overall change of the sliding surface with control input and disturbance.

[0232] Substituting (19) into (28), we get:

[0233] (29)

[0234] The product of its derivative and the term is used to analyze the energy decay characteristics: the first two terms are power-law decay terms. For coefficients, The third term is a nonlinear damping term, used to attenuate the power term and ensure rapid system convergence. For system parameters, The first term is a small constant that suppresses oscillations; the last two terms are the cross term of the disturbance and the sliding surface and the dynamic compensation term, respectively, which together constitute the energy balance equation.

[0235] According to Fang Suo:

[0236]

[0237]

[0238]

[0239] (30)

[0240] The coupled sliding surface is obtained through calculation and derivation. Perturbation estimation error and control input error The comprehensive inequalities are shown in the modified formula. The modified formula reflects the cross-constraint relationship between the coupled sliding surface and the control quantity and disturbance. The first two terms are the scaled error product constraints, and the last two terms are the sign and amplitude attenuation relationship of the power term, providing a basic inequality for stability analysis.

[0241] Based on system dynamic compensation terms and coupling sliding surface Derive the inequality constraints for the nonlinear damping term. and As can be seen from the definition:

[0242] (31)

[0243] This formula constructs a denominator containing a small constant. The term, constraining its magnitude to Within the range, avoid controlling input singularities while ensuring energy decay characteristics.

[0244] Using the auxiliary system (corresponding to the rewritten form of the auxiliary system) and the adaptive law (corresponding to the rewritten disturbance estimation law), derive the disturbance estimation error. derivative of the estimator The constraint relationship. From (16) and (19), we can obtain:

[0245] (32)

[0246] Formula (32) includes a hyperbolic tangent term of the coupled sliding surface to suppress jitter and a power-law decay term to ensure convergence at a fixed time, reflecting the compensation effect of the adaptive mechanism on disturbance.

[0247] and

[0248] (33)

[0249] Auxiliary system status The product of its derivative and the term characterizes the compensation mechanism for saturation error: the first two terms are error cross terms, where, This is the scaling factor. To control for error, the middle four terms are square-decay terms and power-decay terms. For coefficients, The attenuation power is represented by the last two terms, which are error compensation terms. Overall, this ensures that the auxiliary system can effectively offset the adverse effects of saturation.

[0250] Further scaling using inequalities:

[0251] (34)

[0252] By transforming the basic inequality, the cross-product term is decomposed into the sum of two square terms through the completing the square method, thus constructing the underlying inequality for the system energy analysis. This lays a solid foundation for the derivation of the energy function and the analysis of decay characteristics in the subsequent stability proof.

[0253] (35)

[0254] Focusing on the second derivative of the control input Disturbance terms With auxiliary system status The cross-dynamic characteristics are analyzed using an inequality scaling strategy, decomposing them into a sum of several square terms. This operation constrains the system's response under dynamic adjustments to control inputs and disturbances, providing crucial information for subsequent analysis of the auxiliary system's compensation mechanism and support stability.

[0255] Similarly, we can obtain:

[0256] (36)

[0257] Perturbation estimation error in adaptive law The power-order cross terms are subjected to scaling operations. This transforms higher-order power-order product terms into linear combinations of single-power-order terms, significantly simplifying the processing complexity of higher-order nonlinear terms in energy function analysis. This makes it easier to quantify and derive the compensation effect of the adaptive mechanism on disturbances through energy inequalities, providing support for constructing a rigorous energy decay relationship.

[0258] (37)

[0259] Another set of perturbation estimation errors in the adaptive law The power-order cross terms are subjected to scaling. This process ensures the effective constraint of higher-order nonlinear terms within the energy analysis framework. Combined with Equation 46, it further improves the adaptive mechanism's compensation analysis of system disturbances, ensuring the rigor and completeness of the energy function decay characteristic derivation.

[0260] Further results were obtained:

[0261] (38)

[0262] Based on the bounded property of the hyperbolic tangent function, i.e., when the independent variable is positive, For the coupling sliding surface Total disturbance The dynamic terms of the intersection with the hyperbolic tangent function are subjected to scaling. This operation simplifies the complex nonlinear hyperbolic tangent term in energy analysis into a linear intersection term, reducing the complexity of energy inequality derivation and providing concise and crucial constraints for subsequently constructing a rigorous energy decay relationship and demonstrating system stability.

[0263] From (23), we can know Furthermore, because , , and Substituting (28)-(38) into (27) yields:

[0264] (39)

[0265] Further scaling down:

[0266] (40)

[0267] energy function The derivative is used to perform more precise scaling operations, and inequality techniques are used to couple the sliding surfaces. Perturbation estimation error The sum of multiple system state terms is transformed into a single power term. This is then combined with the control gain after the substitution. The adjustments further constrain the system's energy decay process. This formula, through standardized symbol substitution, enhances the rigor of the stability proof and seamlessly integrates with the overall derivation logic.

[0268] We can obtain:

[0269] (41)

[0270] In the formula:

[0271]

[0272]

[0273]

[0274]

[0275] Furthermore, to The proof will be divided into two cases:

[0276] First scenario: If ,

[0277] (42)

[0278] In the formula, and According to the fixed-time stability theory, It is stable over a fixed period of time.

[0279] The second scenario: If

[0280] Step 1 scaling:

[0281] (43)

[0282] The second step is scaling (defining new coefficients):

[0283] (44)

[0284] In the formula, and Obviously, It is also stable at a fixed time.

[0285] Combining the two situations above, when Case 1 ( ) and Case 2 ( When merging, the derivative of the energy function satisfies:

[0286] (45)

[0287] The comprehensive coefficient is defined as follows: , .

[0288] According to the fixed-time stability theory Stable over a fixed period of time, that is... At the specified time Internally, it converges to There exists a positive constant. satisfy The following relationship holds:

[0289] (46)

[0290] Satisfy the following expression:

[0291] (47)

[0292] therefore, , , All will converge to the following range:

[0293] (48)

[0294] Step 2: convergence

[0295] By selecting appropriate parameters It will converge to a smaller region of zero, as in the range of (48). Considering and They are equivalent; therefore, the sliding surface (13b) will be rewritten as follows:

[0296] (49)

[0297] The following Lyapunov candidate functions are selected:

[0298] (50)

[0299] right Differentiation yields:

[0300] (51)

[0301] Based on formula (12) and Relationship, The convergence of will be proven through the following two cases:

[0302] Scenario 1: If or , We can obtain:

[0303] (52)

[0304] Substituting (52) into (51), we get:

[0305] (53)

[0306] 1) When ,

[0307] (54)

[0308] In the formula, , .

[0309] 2) When ,

[0310] (55)

[0311] In the formula, , .

[0312] Therefore, according to the fixed-time stability theory, it can be known that It is stable over a fixed period of time.

[0313] Scenario 2: If , We can obtain:

[0314] (56)

[0315] because There is more A faster convergence speed means a shorter convergence time in this stage.

[0316] Based on the analysis of scenarios 1 and 2, It converges in a fixed time; furthermore, its convergence time is... The following relationship must be satisfied:

[0317] (57)

[0318] visible, It converges in a fixed time, and the convergence time is... satisfy Furthermore, considering and They are equivalent, that is to say, It is fixed-time convergence, that is, the heterogeneous vehicle formation control system of the heterogeneous vehicle system, the disturbance estimation module caused by unknown disturbances and model uncertainties, and the auxiliary system and event triggering mechanism for compensating actuator saturation error and control command deviation in the event triggering mechanism are all fixed-time convergence.

[0319] Therefore, it can be concluded that the entire control time of a heterogeneous vehicle system converges within a certain time range, i.e. .

[0320] Proof of Zeno's behavior exclusion

[0321] Considering heterogeneous vehicle systems (1), event triggering mechanisms (22), (23), (24), (25), if The choice satisfies Therefore, the deviation between the output trajectory and the expected trajectory of the heterogeneous vehicle system will converge, and the states of the heterogeneous vehicle system will all converge. This has already been derived from the stability proof of the heterogeneous vehicle system above. The following is a proof of excluding Zeno's behavior in the event-triggered mechanism, divided into three steps:

[0322] Step 1: Assumptions of the analysis derivative

[0323] We prove by contradiction that the proposed control scheme does not exhibit Zeno behavior. Therefore, assuming Zeno behavior does exist, there must exist a finite point in time. , making

[0324] (58)

[0325] In other words, within a limited time interval Within this, there are an infinite number of triggering events. For any... In the interval Inside, It is a constant. When We have It is worth noting that, within the interval Inside, It starts from zero, and may reach zero again many times afterward. Let... for At the last time point that zero is reached, we have Furthermore, in the interval Inside, This is true. Therefore, we can calculate... In the interval The derivative within the interval is shown below:

[0326] (59)

[0327] Step Two: Differential of the upper bound

[0328] It must be a continuous function. Because all variables... Since the interval is bounded, according to equation (45), we have It must be bounded. Therefore, there exists a positive constant. , making In the interval established within.

[0329] From equation (20-23), we can obtain Therefore, we have

[0330] (60)

[0331] because As shown in (60), we can deduce that the only necessary condition for (23) to hold is:

[0332] (61)

[0333] Step 3: Adjacent trigger interval Derivation

[0334] because Therefore, the next triggering time Must meet:

[0335] (62)

[0336] because We can obtain:

[0337] (63)

[0338] Therefore, the time interval between any two consecutive events is small and subject to a positive constant. The limitations. Assume Indicates greater than The smallest number. If Then we can get This contradicts (58). This indicates a finite time interval. An array cannot contain an infinite number of events. Therefore, Zeno's behavior does not exist.

[0339] Simulation verification

[0340] Consider a heterogeneous vehicle system consisting of one leader and five followers. MATLAB / Simulink software is used to verify the effectiveness and superiority of the control scheme proposed in this invention. The initial position values ​​of the heterogeneous vehicle system are as follows: m, velocity , More relevant parameter selections are shown in Tables 1 and 2. The leader's acceleration value is selected as follows:

[0341]

[0342] Saturated input Choose from the following:

[0343]

[0344] To verify the effectiveness and superiority of the heterogeneous vehicle cooperative control scheme proposed in this invention under actual working conditions, this embodiment selected five different types of heterogeneous vehicles, including a fuel-powered off-road vehicle, an electric van, a fuel-powered passenger car, an electric truck, and an electric delivery vehicle, and formed a convoy for simulation verification. These vehicles have significant differences in driving energy, mass, time constant, and vehicle length, which can fully reflect the characteristics of heterogeneous vehicles. The main parameters of the heterogeneous vehicle system used for verification in this invention are shown in Table 1:

[0345] Table 1. Key parameters of heterogeneous vehicle systems

[0346]

[0347] Based on the aforementioned vehicle physical parameters in Table 1, this invention further designs and tunes relevant parameters at the control system level. These parameters directly affect the system's convergence speed, tracking accuracy, triggering frequency, and stability. The selection of relevant parameters for the heterogeneous vehicle control system designed in this invention is shown in Table 2:

[0348] Table 2. Main parameters of heterogeneous vehicle control systems

[0349]

[0350] (1) Effect of fixed-time convergence sliding mode control

[0351] Figure 5 The simulation results of this invention are presented. Figure 5 (a)-(c) show the position, velocity, and acceleration of each vehicle state, respectively. This indicates that all vehicles can maintain formation following within the specified time, there are no collisions between adjacent vehicles, and all subsequent vehicles can eventually follow vehicle 0 with a bounded error in velocity and acceleration. Figure 5 (d) gives the tracking error To ensure the performance required at a fixed time, Figure 5 Each tracking error in (d) All at a given time Converging inward to the target region It can be seen that the tracking error for each vehicle... During the control process, it always remains within the specified performance range, achieving both steady-state and transient control objectives.

[0352] (2) The effect of static event triggering mechanism

[0353] Figure 6 This figure illustrates the results of trigger interval changes over time obtained through a static triggering mechanism under different control inputs. The subplots, from top to bottom, represent cases u1, u2, u3, u4, and u5, respectively denoted as (a) u1; (b) u2; (c) u3; (d) u4; and (e) u5. The horizontal axis represents simulation time, and the vertical axis represents the trigger interval time. Figure 6 The number of events is given in the table.

[0354] Figure 6This indicates the dynamic characteristics of the trigger intervals, meaning that the trigger intervals for each control input exhibit different patterns, reflecting the response of the heterogeneous vehicle system to changes in different conditions. In the initial stage, the trigger intervals are short, requiring high-frequency adjustments during the system stabilization process. As time increases, the adjustment requirements decrease after the system reaches steady state. The u1 interval increases rapidly, and the control adjustment frequency decreases, consistent with the process of the heterogeneous vehicle system "adapting to initial conditions and tending towards stability." u2 to u5 show a similar trend, with the intervals all increasing over time.

[0355] However, while static event-triggered mechanisms provide a structured control update method, they have several limitations. Their main drawback is their inability to flexibly adapt to dynamic changes in system state. Preset trigger intervals cannot flexibly adapt to dynamic system changes (such as sudden disturbances or changes in operating conditions), potentially leading to performance loss or instability. Due to the specificity of the dynamic characteristics and operating conditions of each input, the extension rate and magnitude of the u1~u5 intervals will differ. Simulation results of static event-triggered mechanisms emphasize the importance of employing adaptive control strategies that can adjust the response based on the current state and performance of heterogeneous vehicle systems.

[0356] In summary, the simulation results verify the effectiveness of the proposed ETC in managing vehicle dynamics using a static event-triggered mechanism. However, the increasing trigger interval over time indicates a need for more flexible control methods to better adapt to dynamic changes in the system. Although the static triggering mechanism provides a foundation for control, further optimization can be achieved by incorporating adaptive elements into the control strategy.

[0357] (3) The effect of dynamic event triggering mechanism

[0358] Figure 7 The results of simulating trigger intervals for different control inputs using a dynamic event triggering mechanism are shown. The subplots in the figure, from top to bottom, represent cases u1, u2, u3, u4, and u5, respectively denoted as (a) u1; (b) u2; (c) u3; (d) u4; and (e) u5. The horizontal axis represents the simulation time, and the vertical axis represents the trigger interval time. Figure 7 The number of events is given in the table.

[0359] Figure 7 The system performance was demonstrated, and the advantages of dynamic event triggering mechanisms over static triggering mechanisms were highlighted. Figure 7 The trigger intervals shown exhibit significant variability, reflecting the dynamic nature of control adjustments. Unlike static triggering mechanisms, where the intervals gradually lengthen over time, dynamic event triggering mechanisms show more frequent and shorter intervals, especially noticeable in the initial stages. This variability indicates that the system has a stronger responsiveness and can adapt to changes in vehicle status more quickly.

[0360] exist Figure 7In the figure, the trigger interval of u1 is significantly shorter and more frequent compared to the static mechanism, indicating the adoption of a more proactive control strategy. This trend is consistent across all control inputs; u3 and u4 in the figure show similar frequent adjustment patterns, especially in the initial stage.

[0361] A comparison between dynamic event triggering mechanisms and static mechanisms verifies the former's significant advantages in system adaptability and control robustness:

[0362] (1) Adaptability and Response Performance

[0363] Static mechanisms, due to their preset trigger intervals, are difficult to adapt to sudden dynamic changes in the system, such as disturbances or changes in operating conditions, which can easily lead to delayed response or performance degradation. Dynamic mechanisms, on the other hand, dynamically adjust the intervals based on the real-time status, allowing for targeted optimization of control strategies and making them more robust to dynamic disturbances.

[0364] (2) Relationship between trigger interval and stability

[0365] In dynamic mechanisms, the correlation between trigger interval and system stability is more significant: in the initial stage, the system is quickly stabilized by high-frequency adjustments, and after steady state, the interval naturally extends as intervention needs decrease, avoiding over-control; although static mechanisms have a longer steady-state interval, they are prone to insufficient response due to a lack of adaptability.

[0366] (3) Advantages in the number of events and efficiency

[0367] To further illustrate the impact of dynamic and static triggering mechanisms on the system, the number of triggering events throughout the simulation process has been statistically analyzed. Simulation statistics ( Figure 6 Figure 7 shows that the number of events in the static mechanism is significantly higher than that in the dynamic mechanism: the number of static events u1, u2, u3, u4, and u5 are 769, 166, 192, 197, and 331, respectively, while the corresponding number of events in the dynamic mechanism are 96, 61, 126, 124, and 302. The proportion of the number of events in the dynamic mechanism to that in the static mechanism is 12.48%, 36.75%, 65.63%, 62.94%, and 91.24%, respectively. The total number of events in the dynamic triggering mechanism is reduced by approximately 57.16% compared to the static triggering mechanism, meaning that the number of dynamic events accounts for only 42.84% of that in the static mechanism. It is evident that the dynamic mechanism achieves the same control objective with fewer events, reducing computational overhead while improving real-time performance, making it particularly suitable for scenarios with high response requirements.

[0368] In summary, dynamic event triggering mechanisms offer significant advantages in scenarios sensitive to computational load and demanding efficient resource utilization. In the control process of heterogeneous vehicle systems, dynamic event triggering mechanisms demonstrate performance far exceeding static mechanisms: by accurately perceiving the real-time state of the system, they dynamically adjust trigger intervals, using adaptive control logic to drive the system to quickly converge to a stable state, while reasonably reducing intervention frequency during the steady-state maintenance phase. In terms of event frequency management, DETC can significantly reduce unnecessary triggering actions, effectively reducing the repeated consumption of computing resources compared to static mechanisms, fundamentally alleviating the system's computational burden. This control method, combining rapid stability with efficient computing characteristics, is highly suitable for application scenarios like autonomous driving, which have stringent requirements for response speed and environmental adaptability. It provides a more advantageous control strategy to support the reliable operation of heterogeneous autonomous vehicle systems under complex conditions, helping to achieve more intelligent and efficient vehicle collaborative control.

Claims

1. A heterogeneous vehicle cooperative control system based on a dynamic event triggering mechanism, characterized in that, The system comprises four parts: a heterogeneous vehicle platooning system perception module with a disturbance estimation and auxiliary system that has fixed-time convergence; a heterogeneous vehicle platooning control system with funnel control, sliding mode control and dynamic event triggering mechanism under a unified framework of fixed-time convergence; electric and fuel drive systems and transmission systems of heterogeneous vehicles; and a vehicle networking communication module. The perception module of the heterogeneous vehicle platooning system with a fixed-time convergence disturbance estimation and auxiliary system includes: a speed sensor, a position sensor (GPS signal), a slope sensor, a vehicle mass estimation module, a fixed-time convergence disturbance estimation module caused by unknown disturbances and model uncertainties, and a fixed-time convergence auxiliary system for compensating for actuator saturation errors and control command deviations in event triggering mechanisms, which are connected in sequence to the heterogeneous vehicle platooning control system. A heterogeneous vehicle platooning control system with funnel control, sliding mode control, and dynamic event triggering mechanisms under a unified framework of fixed-time convergence: It connects to the power control unit of electric vehicles or fuel vehicles and issues control commands to them. The control commands are then converted into torques to drive the vehicles and transmitted to the drive wheels through the drive and transmission systems of the heterogeneous vehicles. At the same time, the heterogeneous vehicle platooning control system connects to the vehicle network communication module and completes network communication with adjacent vehicles. Electric and fuel drive systems and transmission systems of heterogeneous vehicles: including electric drive systems or fuel drive systems and their transmission systems in heterogeneous vehicles; electric vehicles include a power battery, a drive motor controller, a drive motor / transmission and half shafts / drive wheels connected in sequence; fuel vehicles include an engine controller, an engine / transmission and half shafts / drive wheels connected in sequence. Vehicle-to-everything (V2X) communication module: Located on the upper side of the heterogeneous vehicle platooning control system, it includes a wireless transmitting module and a wireless receiving module and is electrically connected to the heterogeneous vehicle platooning control system. The aforementioned vehicle mass estimation module is used to estimate vehicle mass; The drive systems of the heterogeneous vehicles include electric drive systems, fuel drive systems, and hybrid drive systems with different characteristics; In the heterogeneous vehicle platooning system, each vehicle communicates with the network through a wireless transmitter and a wireless receiver module located at the front of the vehicle, and sequentially obtains information streams from adjacent vehicles, including the position, speed, and acceleration information of adjacent vehicles.

2. The heterogeneous vehicle cooperative control system based on a dynamic event triggering mechanism as described in claim 1, characterized in that, The auxiliary system for compensating actuator saturation error and control command deviation in the event-triggered mechanism with fixed-time convergence can achieve faster compensation for actuator saturation error and control command deviation in the event-triggered mechanism. This allows the heterogeneous vehicle platooning control system to provide compensation information in a shorter time, improve the real-time performance of platooning tracking error control for heterogeneous vehicles, and is unaffected by initial conditions, resulting in faster convergence.

3. A heterogeneous vehicle cooperative control method based on a dynamic event triggering mechanism, characterized in that, Includes the following steps: S1. Establish a dynamic model for the nonlinear heterogeneous vehicle system; S2. Calculation of vehicle-to-vehicle tracking distance error, fixed-time funnel control, error conversion and construction of sliding surface; S3. Construct a fixed-time-based auxiliary system to compensate for actuator nonlinearity errors and control input errors caused by event triggering; S4. Construct adaptive laws for external and internal disturbances; S5. Design a fixed-time formation controller; S6. Design a dynamic event triggering control mechanism; S7. Determine whether the event triggering condition is met. If it is met, return to step S1; otherwise, proceed to step S8. S8. Determine if the termination condition is met. If it is met, end the control flow; otherwise, return to step S1. The nonlinear heterogeneous vehicle system dynamics model in step S1 considers a heterogeneous vehicle system consisting of one lead vehicle and N follower vehicles, totaling N+1 vehicles. For the i-th vehicle in the heterogeneous vehicle system, i ∈ {1,...,N}, the heterogeneous vehicle dynamics model is established as follows: (1a) (1b) (1c) (1d) The kinematic model of vehicle i can be extended to a third-order system: position The derivative of velocity The derivative of velocity is acceleration. The derivative of acceleration is determined by the event-triggered control command. System dynamic items (A function containing velocity and acceleration) and total disturbance A joint decision, in which This is the rotational mass correction factor. For quality, It is a time constant. For air density, The frontal area of ​​vehicle i It is the air drag coefficient of vehicle i. It is the rolling resistance of vehicle i. Let i be the rolling resistance coefficient of vehicle i. It is gravitational acceleration. The slope representing vehicle i This is the gravity gradient resistance of vehicle i, and the road resistance coefficient of vehicle i is... ; (2) (3) Wherein, the saturation control input is defined. When the original control quantity Exceeding the upper and lower limits or When the saturation input is at its boundary value, it takes that value; otherwise, it equals the original control input. Control input estimation error. For saturated input Its event-triggered control commands The difference is used to describe the signal deviation caused by the event triggering mechanism; Furthermore, Divided into the following two parts: (4) in It is a known term. It is an unknown term. (1c) can be written in the following form: (5) In the formula, the system dynamic term It can be decomposed into known nominal terms. and unknown disturbance terms The acceleration derivative formula is further simplified as follows: Given the dynamic term and control input error (including saturation error) and disturbance error ) and total disturbance A combination of (model uncertainty and external disturbances), where To control the scaling factor; Step S2, which involves calculating the vehicle-to-vehicle tracking distance error, controlling the fixed-time funnel, and converting errors to construct the sliding surface, comprises the following four steps: S21: Calculation of vehicle-to-vehicle tracking distance error Using a constant-time spacing strategy, the error in the spacing between adjacent heterogeneous vehicles is as follows: (6) Among them, tracking error Position of the vehicle in front Current vehicle position The difference, minus the current vehicle length Time of departure from the locomotive With current vehicle speed The product and minimum safe distance , used to describe the spacing control deviation of vehicle queues; S22: Fixed-time funnel control Design a fixed-time funnel function (Lower Boundary) and (Upper boundary), in which For attenuation term, , These are boundary coefficients. This is a saturation compensation term (adaptively adjusted based on the actuator's saturation state). When When, the function dynamically adjusts according to the initial error sign; when At the same time, maintain constant boundaries to ensure tracking error Constrained within the funnel area, the funnel control function is shown in the following equation: (7) (8) In the formula , , , , , , , It is the initial tracking error and satisfies ; S23: Tracking Error Conversion To avoid the performance boundary limitations of the aforementioned funnel control, the tracking error can be transformed into an unconstrained form, as shown below: (9) In the formula, The value range is (0,1). This can be expressed as the product of the transformation function and the sum of the funnel boundaries minus the lower bound, obtained through logarithmic transformation. Its convergence and Equivalent, facilitating controller design, tracking error will occur within a given time. Converging to the specified range ; Furthermore, the tracking error can be written as: (10) To proceed with the design of the heterogeneous vehicle cooperative controller, the derivative of the above equation yields... but and Write it in the following form: (11) (12) Among them, conversion error The first derivative is derived from the proportionality constant. (Related to funnel boundary and error), error change rate and compensation items (Including funnel boundary derivative information) determines; the second derivative further includes the error second derivative and boundary perturbation terms. and its derivative, and The derivative terms provide a dynamic basis for the design of sliding mode controllers; S24: Construct the sliding surface The sliding surface that converges within a fixed time is constructed as follows: (13a) (13b) In the formula, , , , , and , , , , , It is a small constant. , ; To facilitate subsequent stability verification, the sliding surface is constructed as follows: (14) To prove the stability of the queue, a coupled sliding surface is defined. For the first n-1 vehicles, its value is the sliding surface of the current vehicle. sliding surface of the rear vehicle Coupling (coefficient) (For coupling strength); the last car is only related to its own sliding surface. Coupling design ensures that errors do not propagate within the queue, which is evident. and The convergence is consistent.

4. The heterogeneous vehicle cooperative control method based on a dynamic event triggering mechanism according to claim 3, characterized in that, In step S3, a fixed-time auxiliary system is constructed to compensate for actuator nonlinearity errors and control input errors caused by event triggering. The design is as follows: (15) In the formula , and All are positive design parameters.

5. The heterogeneous vehicle cooperative control method based on a dynamic event triggering mechanism according to claim 3, characterized in that, In step S4, the adaptive laws for external and internal disturbances are constructed as follows: (16) 6. The heterogeneous vehicle cooperative control method based on a dynamic event triggering mechanism according to claim 3, characterized in that, In step S5, firstly, Given as follows: (17) At the same time, (18) The design of a fixed-time formation controller for a heterogeneous vehicle system is as follows: (19) In the formula , , , , , , , , They are all positive constants.

7. The heterogeneous vehicle cooperative control method based on a dynamic event triggering mechanism according to claim 3, characterized in that, The dynamic event triggering control mechanism designed in step S6 is as follows: S61: First, design external auxiliary variables. : (20) further, (21) Among them, the dynamic event triggering mechanism introduces auxiliary variables. Its derivative includes a decay term (coefficient) ) and adjustment items ( As weight, This is the proportionality coefficient. , For items related to sliding surfaces, To control the squared error. It is defined as a combination of two power terms ( , (where is a coefficient) is used to dynamically adjust the trigger threshold. S62: The dynamic event triggering mechanism is designed as follows: (22) (23) In the formula, It is the actual control signal acting on the heterogeneous vehicle i. The range is (0,1). , ; This is the deviation between the expected control signal and the actual control signal caused by the event-triggered mechanism. , , is the controller's update time, that is, when (23) is triggered, the time will be marked as ,and It will be applied to heterogeneous vehicle i, during the time interval At this time, the control signal will be held at a constant value, where, , , and Design parameters; dynamic event triggering time Defined as the maximum time when the triggering condition is met, where For threshold coefficient, The square of the current control error. For the threshold reference related to the sliding surface, when the expression on the left is less than or equal to the auxiliary variable If the control is not updated, then no control update is triggered; otherwise, the control value is triggered and updated. Corresponding to the dynamic event triggering mechanism described above, when designing external auxiliary variables... When it is a constant or its derivative is 0, use This leads to the following static event triggering mechanism: (24) (25) The static event triggering conditions are similar to those of the dynamic mechanism, but the threshold coefficient... It is a constant and does not dynamically adjust with the system state. When the square of the control error... With benchmark item The difference multiplied by less than or equal to If the current control value is maintained, then an update is triggered. In step S7, determining whether the event triggering condition is met is (23) when a dynamic event triggering mechanism is used and (25) when a static event triggering mechanism is used. If the condition is met, the process returns to step S1 and is executed again. If the condition is not met, the process proceeds to step S8. Step S8, determining whether the termination condition is met, means that the heterogeneous vehicle cooperative control system receives a termination command or achieves the control objective; the termination command refers to a shutdown signal; the control objective refers to simultaneously satisfying the following: (1) At a given time To achieve fixed-time bicycle stability, that is, when hour, or ; (2) Achieve string stability with fixed-time convergence, that is, within a given time. Subsequently, vehicle spacing tracking error It will not grow any further, that is, , yes The Laplace function, ; (3) Achieve fixed-time funnel control: In other words To converge to , , It is the boundary controlled by the funnel; (4) Implement dynamic event triggering control and exclude Zeno's behavior; If the above termination conditions are met, the control flow ends; otherwise, return to step S1 and continue executing S1-S8.

8. The heterogeneous vehicle cooperative control method based on a dynamic event triggering mechanism according to claim 3, characterized in that, In step S6, parameter adaptation is... , , and The design parameters of the event triggering mechanism are based on the vehicle's , , , , and The state information is used to adaptively determine the outcome, but it requires... , , , And simultaneously satisfy This allows for dynamic event triggering control and excludes Zeno-like behavior.

Images