Intelligent truck platoon control method

Abstract

A method for controlling an intelligent truck platoon includes: obtaining a desired path based on collected multi-source signals and calculating a desired vehicle distance using a time-to-headroom strategy; parameterizing and reconstructing the desired path based on the multi-source signals to obtain a continuous reference trajectory; establishing a high-order kinematic model for each vehicle in the platoon; designing a lateral sliding mode controller that is resistant to path fluctuations based on the high-order kinematic model and the continuous reference trajectory of each vehicle, and calculating a desired front wheel angle; designing a second-order hyperbolic integral sliding mode controller with lateral adhesion suppression based on the high-order kinematic model, the continuous reference trajectory, and the desired vehicle distance of each vehicle, and calculating a desired acceleration; and outputting a steering torque and a driving torque or braking pressure of the wheels based on the calculated desired front wheel angle and desired acceleration to achieve coordinated movement of the intelligent truck platoon. The present invention enables precise and stable coordinated movement of intelligent truck platoons.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent driving, and in particular to an intelligent truck formation control method. Background Art

[0002] In the commercial vehicle sector, truck platooning has attracted widespread attention. Intelligently controlling multiple trucks in pairs can improve traffic efficiency. However, due to the inevitable curvature changes, steep curves, and continuous turns along transport routes, as well as the inherent underactuated nature of trucks, there is a risk of platoon instability caused by the accumulation of tracking errors. This can easily lead to platooning disruptions, traffic congestion, and even serious accidents such as side slips and tailspins, overturning, and collisions.

[0003] Therefore, how to provide a highly resilient truck platoon control method to reduce the risk of queue instability caused by the continuous accumulation of tracking errors in the case of sudden changes in the curvature of the transport route and under-actuation of the transport vehicles, and to achieve precise and stable coordinated movement of intelligent truck platoons, is an urgent problem that technical personnel in this field need to solve. Summary of the Invention

[0004] In view of this, the present invention provides a method for controlling an intelligent truck formation, so as to realize accurate and stable coordinated movement of an intelligent truck formation under the conditions of sudden changes in curvature of a transport route and under-actuation of transport vehicles.

[0005] An intelligent truck platoon control method, comprising:

[0006] Step S1, obtaining a desired path based on the collected multi-source signals, and calculating the desired vehicle distance using a time-to-headness strategy;

[0007] Step S2, based on the multi-source signals, parameterize and reconstruct the desired path to obtain a continuous reference trajectory;

[0008] Step S3, establishing a high-order kinematic model of each vehicle in the truck formation based on the continuous reference trajectory;

[0009] Step S4: Designing a lateral sliding mode controller that is resistant to path fluctuations based on the high-order kinematic model and continuous reference trajectory of each vehicle, and calculating the desired front wheel steering angle;

[0010] Step S5: Based on the high-order kinematic model of each vehicle, the continuous reference trajectory, and the desired vehicle distance, a second-order hyperbolic integral sliding mode controller with lateral adhesion suppression is designed, and the desired acceleration is calculated;

[0011] Step S6: output the steering torque of the steering wheel and the driving torque or braking pressure of the wheels according to the calculated expected front wheel angle and expected acceleration, so as to realize the coordinated movement of the intelligent truck platoon.

[0012] The intelligent truck platoon control method provided by the present invention has the following beneficial effects:

[0013] (1) The present invention uses a kinematic decoupling mechanism in the Frenet coordinate system, which only requires basic geometric motion parameters to achieve queue cooperative control. This method effectively reduces computational complexity, reduces the impact of multivariate interference in complex scenarios, and significantly improves resilience in dynamic disturbance scenarios.

[0014] (2) The curvature dynamic compensation architecture designed by the present invention integrates feedforward path curvature compensation and sliding mode feedback control to synchronously suppress lateral errors and longitudinal velocity oscillations. The present invention can fundamentally reduce the adhesion effect caused by the lateral and longitudinal dynamic coupling in traditional control, effectively reduce the risk of platoon instability caused by the continuous accumulation of tracking errors, and realize the precise and stable coordinated movement of intelligent truck formations.

[0015] (3) In response to multiple sources of interference such as communication delay, sensor noise and sudden change in path curvature, the present invention proposes a comprehensive lateral anti-curvature interference and longitudinal anti-lateral adhesion interference and multiple interference collaborative suppression strategy. This design takes into account both active defense against sudden interference such as program anomalies and real-time correction of calculation errors, thereby improving the control effectiveness in complex scenarios while ensuring safety.

[0016] (4) The present invention adopts a lightweight geometric control framework and can be quickly transplanted to various autonomous driving platforms, such as passenger cars, commercial vehicles, special operation vehicles, etc., which greatly reduces development costs and provides highly versatile solutions for scenarios such as logistics formations and park docking. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A schematic flow chart of an intelligent truck platoon control method according to an embodiment of the present invention;

[0018] Figure 2 This is a schematic diagram comparing lateral errors of three lateral control methods in an exemplary complex road real vehicle test;

[0019] Figure 3 Schematic diagram of the overall lateral error of the vehicle formation controlled by the method of the present invention;

[0020] Figure 4 This is a schematic diagram of the overall lateral error of the vehicle formation controlled by the traditional sliding mode controller;

[0021] Figure 5 This is a schematic diagram of the overall lateral error of the vehicle formation controlled by the pure tracking controller;

[0022] Figure 6 This is a schematic diagram comparing the longitudinal errors of three longitudinal control methods in an exemplary complex road real vehicle test;

[0023] Figure 7 Schematic diagram of the overall longitudinal error of the vehicle formation controlled by the method of the present invention;

[0024] Figure 8 is a schematic diagram of the overall longitudinal error of the vehicle formation controlled by the exponential sliding mode controller;

[0025] Figure 9 It is a schematic diagram of the overall longitudinal error of the vehicle formation controlled by the proportional sliding mode controller. DETAILED DESCRIPTION

[0026] The following describes embodiments of the present invention in detail, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the embodiments of the present invention, and should not be construed as limiting the present invention.

[0027] See also Figure 1 , an embodiment of the present invention provides an intelligent truck formation control method, comprising steps S1-S6:

[0028] Step S1: obtaining a desired path based on the collected multi-source signals and calculating the desired vehicle distance using a time-to-headness strategy.

[0029] Among them, there are a total of The vehicle speed of the nth truck in the Frenet coordinate system is obtained by using multi-sensor fusion to collect multi-source signals. , vehicle acceleration , actual front wheel angle , heading angle , expected path S, longitudinal distance , vehicle center of mass position N, projection point M, tangent direction at projection point M ;

[0030] Calculate the expected distance between the nth truck using the time-distance strategy , the expression is:

[0031]

[0032] in, For time interval, For the static safety distance, The first in the truck formation The vehicle speed of the truck is The truck is the preceding vehicle of the nth truck.

[0033] Step S2: Based on the multi-source signals, the desired path is parameterized and reconstructed to obtain a continuous reference trajectory.

[0034] Among them, the B-spline interpolation method optimizes the local curvature through the control point set to obtain a continuous reference trajectory and avoid longitudinal velocity oscillation caused by geometric mutations in queue control.

[0035] In this embodiment, step S2 specifically includes:

[0036] Based on multi-source signals, the B-spline interpolation method is used to parameterize and reconstruct the expected path to obtain a continuous reference trajectory. Order B-spline interpolation satisfies the following formula:

[0037] ;

[0038] in, is a set of continuous reference trajectory points, is the parameter variable of the continuous reference curve, and , is the abscissa of the continuous reference trajectory, is the ordinate of the continuous reference trajectory, represents transpose, is the power parameter matrix, is the basis function coefficient matrix, is the control point set;

[0039] ;

[0040] ;

[0041] ;

[0042] in, for The element in row 1 and column 1 of for The element in row 1 and column 2, for The first row, Elements of the column, for The element in the second row and first column of for The element in the second row and second column of for In the second row, Elements of the column, for Middle The element in row and column 1, for Middle The element in row 1 and column 2, for Middle Row, No. Elements of the column, for The element in row 1 and column 1 of for The element in row 1 and column 2, for The element in the second row and first column of for The element in the second row and second column of for Middle The element in row and column 1, for Middle The element in row 1 and column 2, is the index of the path point in the local area path segment of the projection point M;

[0043] Based on the control point set , and obtain a continuous reference trajectory :

[0044]

[0045] in, is the time differential;

[0046] Then based on the continuous reference trajectory , calculate the curvature at the projection point M respectively , the curvature derivative at the projection point M , the distance between the vehicle's center of mass position N and the projection point M , the desired heading angle at the projection point M , the heading angle error of the nth truck at the projection point M , the expression is:

[0047]

[0048]

[0049]

[0050]

[0051] in, yes The first derivative with respect to time, yes The second derivative with respect to time, yes The first derivative with respect to time, yes The second derivative with respect to time, is the horizontal coordinate of the center of mass of the nth truck, is the vertical coordinate of the center of mass of the nth truck.

[0052] Step S3: establishing a high-order kinematic model of each vehicle in the truck formation based on the continuous reference trajectory.

[0053] Wherein, step S3 specifically includes:

[0054] Based on the continuous reference trajectory, a feedforward local control architecture is adopted as the control strategy. Based on the one-way communication link between adjacent vehicles, the following vehicle only receives the motion state information of the leading vehicle. A high-order formation kinematic model is established. Under the feedforward local control architecture, the kinematic model of each truck is modeled. In the Frenet coordinate system, the high-order kinematic model of the nth truck satisfies the following equation:

[0055]

[0056]

[0057]

[0058]

[0059]

[0060]

[0061] in, yes The first derivative with respect to time, , is the longitudinal distance at point i, yes The second derivative with respect to time, yes The first derivative with respect to time, yes right The derivative of yes The first derivative with respect to time, is the first intermediate variable of the nth truck, is the second intermediate variable of the nth truck, is the acceleration of the nth truck, is the length of the truck.

[0062] In step S4, based on the high-order kinematic model and continuous reference trajectory of each vehicle, a lateral sliding mode controller that is resistant to path fluctuations is designed, and the desired front wheel steering angle is calculated.

[0063] In step S4, the sliding mode surface of the lateral sliding mode controller that is resistant to path fluctuations is designed. for:

[0064]

[0065] in, is the heading gain, is the lateral gain, is the curvature gain;

[0066] Define the lateral sliding mode control law to resist path fluctuations for:

[0067]

[0068] in, is an adjustable parameter greater than 0;

[0069] Then, the high-order kinematic model is combined with the lateral sliding mode control law to obtain the desired front wheel steering angle of the nth truck. , the expression is:

[0070]

[0071]

[0072] in, is the curvature feedforward compensation term.

[0073] In step S5, based on the high-order kinematic model of each vehicle, the continuous reference trajectory, and the desired vehicle distance, a second-order hyperbolic integral sliding mode controller with lateral adhesion suppression is designed, and the desired acceleration is calculated.

[0074] In step S5, the longitudinal error between the n-th truck and the vehicle preceding the n-th truck is first defined. for:

[0075]

[0076] in, The first in the truck formation The longitudinal distance between two trucks;

[0077] Then the sliding surface of the second-order hyperbolic integral sliding mode controller with lateral adhesion suppression is designed. for:

[0078]

[0079] in, is the error gain, yes is the first derivative with time;

[0080] Then set up a second-order hyperbolic integral sliding mode control law with lateral adhesion suppression , the expression is:

[0081]

[0082] in, and is the superhelical sliding mode gain, is a sign function, is the bounded adhesion compensation term (related to the vehicle's motion outside the longitudinal direction).

[0083] Combining the high-order kinematic model with the second-order hyperbolic integral sliding mode control law with lateral adhesion suppression, we obtain the Expected acceleration of a truck , the expression is:

[0084]

[0085] in, It is The first intermediate variable of a truck, It is The second intermediate variable of a truck, yes The second derivative with respect to time.

[0086] Step S6: output the steering torque of the steering wheel and the driving torque or braking pressure of the wheels according to the calculated expected front wheel angle and expected acceleration, so as to realize the coordinated movement of the intelligent truck platoon.

[0087] In step S6, the steering torque of the output steering wheel is generated by the proportional-differential controller. :

[0088]

[0089] in, is the proportional gain, is the differential gain, yes The differential of

[0090] In step S6, the total longitudinal demand force is calculated using the following formula: :

[0091]

[0092] in, is the mass of the truck, is the rolling resistance coefficient, is the acceleration due to gravity, is the road inclination, is the air density, is the drag coefficient, is the frontal area of ​​the vehicle;

[0093] When the required force is positive, the driving torque of the output wheel :

[0094]

[0095] in, is the wheel radius;

[0096] When the required force is negative, the output is the brake pressure of the wheel :

[0097]

[0098] in, is the braking efficiency factor.

[0099] The following simulation test is conducted in an exemplary complex road real-car test scenario. The path curvature is turbulent between 28m and 38m. Five vehicles are arranged longitudinally on both sides of the preset road at intervals of 2m. The first two vehicles are on the right side, and the last three vehicles are on the left side. The test results are as follows: Figures 2 to 9 shown.

[0100] from Figures 2 to 5 It can be seen that the lateral error of the vehicle formation controlled by the method of the present invention enters a steady state at 0.56s, the traditional sliding mode controller enters a steady state at 0.81s, and the pure tracking controller enters a steady state at 2.43s; after entering the turbulent road section, the lateral steady-state error peak value of the vehicle formation controlled by the method of the present invention is 0.03m, while the lateral steady-state error peak values ​​of the vehicle formation controlled by the traditional sliding mode controller and the pure tracking controller are both around 0.09m. Figure 3 、 Figure 4 、 Figure 5 By comparison, it can be seen that the overall error convergence speed of the vehicle formation controlled by the method of the present invention is faster than that of the other two methods, and the fluctuation is smaller and more stable; compared with other methods, this method has the excellent performance of fast convergence and resistance to path fluctuations.

[0101] from Figures 6 to 9It can be seen that the longitudinal error of the vehicle formation controlled by the method of the present invention enters a steady state within 2.2s, while the traditional exponential approaching sliding mode controller and proportional sliding mode controller enter a steady state after 7.53s; after entering the curvature fluctuation section, the steady-state peak value of the longitudinal distance of the vehicle formation controlled by the method of the present invention is 2.005m, while the steady-state peak values ​​of the longitudinal distance of the vehicle formation controlled by the traditional exponential approaching sliding mode controller and proportional sliding mode controller are both around 2.061m. Figure 7 、 Figure 8 、 Figure 9 By comparison, it can be seen that the overall error convergence speed of the vehicle formation controlled by the method of the present invention is faster than that of the other two methods, and the fluctuation is smaller and more stable; compared with other methods, this method has the excellent performance of fast convergence and resistance to lateral adhesion.

[0102] In summary, the intelligent truck platoon control method according to the above embodiment has the following beneficial effects:

[0103] (1) The present invention uses a kinematic decoupling mechanism in the Frenet coordinate system, which only requires basic geometric motion parameters to achieve queue cooperative control. This method effectively reduces computational complexity, reduces the impact of multivariate interference in complex scenarios, and significantly improves resilience in dynamic disturbance scenarios.

[0104] (2) The curvature dynamic compensation architecture designed by the present invention integrates feedforward path curvature compensation and sliding mode feedback control to synchronously suppress lateral errors and longitudinal velocity oscillations. The present invention can fundamentally reduce the adhesion effect caused by the lateral and longitudinal dynamic coupling in traditional control, effectively reduce the risk of platoon instability caused by the continuous accumulation of tracking errors, and realize the precise and stable coordinated movement of intelligent truck formations.

[0105] (3) In response to multiple sources of interference such as communication delay, sensor noise and sudden change in path curvature, the present invention proposes a comprehensive lateral anti-curvature interference and longitudinal anti-lateral adhesion interference and multiple interference collaborative suppression strategy. This design takes into account both active defense against sudden interference such as program anomalies and real-time correction of calculation errors, thereby improving the control effectiveness in complex scenarios while ensuring safety.

[0106] (4) The present invention adopts a lightweight geometric control framework and can be quickly transplanted to various autonomous driving platforms, such as passenger cars, commercial vehicles, special operation vehicles, etc., which greatly reduces development costs and provides highly versatile solutions for scenarios such as logistics formations and park docking.

[0107] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the patent for this invention shall be determined by the appended claims.

Claims

1. An intelligent truck formation control method, characterized in that: include: Step S1, obtaining a desired path based on the collected multi-source signals, and calculating the desired vehicle distance using a time-to-headness strategy; Step S2, based on the multi-source signals, parameterize and reconstruct the desired path to obtain a continuous reference trajectory; Step S3, establishing a high-order kinematic model of each vehicle in the truck formation based on the continuous reference trajectory; Step S4: Designing a lateral sliding mode controller that is resistant to path fluctuations based on the high-order kinematic model and continuous reference trajectory of each vehicle, and calculating the desired front wheel steering angle; Step S5: Based on the high-order kinematic model of each vehicle, the continuous reference trajectory, and the desired vehicle distance, a second-order hyperbolic integral sliding mode controller with lateral adhesion suppression is designed, and the desired acceleration is calculated; Step S6: outputting the steering torque of the steering wheel and the driving torque or braking pressure of the wheels according to the calculated desired front wheel steering angle and desired acceleration, so as to realize the coordinated movement of the intelligent truck platoon; In step S4, the sliding mode surface of the lateral sliding mode controller that is resistant to path fluctuations is designed. for: in, is the heading gain, is the lateral gain, is the curvature gain, is the heading angle error of the n-th truck at the projection point M, is the distance between the vehicle's center of mass position N and the projection point M, is the curvature at the projection point M; Define the lateral sliding mode control law to resist path fluctuations for: in, is an adjustable parameter greater than 0; Then, the high-order kinematic model is combined with the lateral sliding mode control law to obtain the desired front wheel steering angle of the nth truck. , the expression is: in, is the curvature feedforward compensation term, is the length of the truck, is the vehicle speed of the nth truck, yes The first derivative with respect to time, is the curvature derivative at the projection point M; In step S5, the longitudinal error between the n-th truck and the vehicle preceding the n-th truck is first defined. for: in, The first in the truck formation The longitudinal distance between two trucks, is the longitudinal distance, is the expected distance between the nth truck; Then the sliding surface of the second-order hyperbolic integral sliding mode controller with lateral adhesion suppression is designed. for: in, is the error gain, yes is the first derivative with time; Then set up a second-order hyperbolic integral sliding mode control law with lateral adhesion suppression , the expression is: in, and is the superhelical sliding mode gain, is a sign function, is the bounded adhesion compensation term; Combining the high-order kinematic model with the second-order hyperbolic integral sliding mode control law with lateral adhesion suppression, we obtain the Expected acceleration of a truck , the expression is: in, It is The first intermediate variable of a truck, is the first intermediate variable of the nth truck, is the second intermediate variable of the nth truck, is the acceleration of the nth truck, It is The second intermediate variable of a truck, yes The second derivative with respect to time.

2. The intelligent truck platoon control method according to claim 1, characterized in that: Step S1 specifically includes: Multi-sensor fusion is used to collect multi-source signals and obtain the vehicle speed of the nth truck in the Frenet coordinate system. , vehicle acceleration , actual front wheel angle , heading angle , expected path S, longitudinal distance , vehicle center of mass position N, projection point M, tangent direction at projection point M ; Calculate the expected distance between the nth truck using the time-distance strategy , the expression is: in, For time interval, For the static safety distance, The first in the truck formation The vehicle speed of the truck is The truck is the preceding vehicle of the nth truck.

3. The intelligent truck platoon control method according to claim 2, characterized in that: Step S2 specifically includes: Based on multi-source signals, the B-spline interpolation method is used to parameterize and reconstruct the expected path to obtain a continuous reference trajectory. Order B-spline interpolation satisfies the following formula: ； in, is a set of continuous reference trajectory points, is the parameter variable of the continuous reference curve, and , is the abscissa of the continuous reference trajectory, is the ordinate of the continuous reference trajectory, represents transpose, is the power parameter matrix, is the basis function coefficient matrix, is the control point set; ； ； ； in, for The element in row 1 and column 1 of for The element in row 1 and column 2, for The first row, Elements of the column, for The element in the second row and first column of for The element in the second row and second column of for In the second row, Elements of the column, for Middle The element in row and column 1, for Middle The element in row 1 and column 2, for Middle Row, No. Elements of the column, for The element in row 1 and column 1 of for The element in row 1 and column 2, for The element in the second row and first column of for The element in the second row and second column of for Middle The element in row and column 1, for Middle The element in row 1 and column 2, is the index of the path point in the path segment; Based on the control point set , and obtain a continuous reference trajectory : in, is the time differential; Then based on the continuous reference trajectory , calculate the curvature at the projection point M respectively , the curvature derivative at the projection point M , the distance between the vehicle's center of mass position N and the projection point M , the desired heading angle at the projection point M , the heading angle error of the nth truck at the projection point M , the expression is: in, yes The first derivative with respect to time, yes The second derivative with respect to time, yes The first derivative with respect to time, yes The second derivative with respect to time, is the horizontal coordinate of the center of mass of the nth truck, It is the vertical coordinate of the center of mass of the n-th truck.

4. The intelligent truck platoon control method according to claim 3, characterized in that: Step S3 specifically includes: Based on the continuous reference trajectory, a feedforward local control architecture is adopted as the control strategy. Based on the one-way communication link between adjacent vehicles, the following vehicle only receives the motion state information of the leading vehicle. A high-order formation kinematic model is established. Under the feedforward local control architecture, the kinematic model of each truck is modeled. In the Frenet coordinate system, the high-order kinematic model of the nth truck satisfies the following equation: in, yes The first derivative with respect to time, , is the longitudinal distance at point i, yes The second derivative with respect to time, yes The first derivative with respect to time, yes right The derivative of yes The first derivative with respect to time, is the first intermediate variable of the nth truck, is the second intermediate variable of the nth truck, is the acceleration of the nth truck, is the length of the truck.

5. The intelligent truck formation control method according to claim 4, characterized in that: In step S6, the steering torque of the output steering wheel is generated by the proportional-differential controller : in, is the proportional gain, is the differential gain, yes The differential of In step S6, the total longitudinal demand force is calculated using the following formula: : in, is the mass of the truck, is the rolling resistance coefficient, is the acceleration due to gravity, is the road inclination, is the air density, is the drag coefficient, is the frontal area of ​​the vehicle; When the required force is positive, the driving torque of the output wheel : in, is the wheel radius; When the required force is negative, the output is the brake pressure of the wheel : in, is the braking efficiency factor.

Images