Remote man-machine cooperative driving control method for intelligent vehicle

Abstract

The invention discloses an intelligent vehicle remote man-machine cooperative driving control method which comprises the following steps: 1, modeling a remote man-machine cooperative driving system: determining relevant parameters of a vehicle and a traffic scene, and constructing a remote driver control model and an automatic driving control model; 2, calculating the stability range of the model: respectively calculating the stability regions of the remote driver control model and the automatic driving control model according to the remote man-machine cooperative driving system model and parameters; and 3, designing a collaborative driving controller: designing a switching function according to corresponding control quantities of the remote driver control model and the automatic driving control model to complete smooth switching of the driving right between the automatic driving control and the remote driver control. The collaborative driving controller is designed through a variable structure control method, smooth switching of a man-machine collaborative driving control system under the condition of remote man-machine collaborative driving is achieved, and the safety and stability of intelligent vehicle remote man-machine collaborative driving are improved.

Description

Technical Field

[0001] The present invention belongs to the field of automation and computer technology, and in particular to the field of intelligent vehicle remote human-machine collaborative driving technology, and specifically to a control method for intelligent vehicle remote human-machine collaborative driving. Background Art

[0002] Remote human-machine collaborative driving of intelligent vehicles is one of the core functions of autonomous driving technology. Human-machine collaborative driving can serve as a transitional stage to fully autonomous driving, not only laying the foundation for future autonomous driving, but also improving traffic safety and driving experience in the short term. In remote human-machine collaborative driving, the switching of vehicle control between the remote driver and the autonomous driving controller is the main problem facing this technology.

[0003] Chinese patent application: A vehicle steering control method under remote human-machine collaboration (application number: CN202110718353.5) discloses a vehicle steering control method under remote human-machine collaboration, which is characterized in that the comprehensive safety field potential value and the short-term expected trajectory are first calculated, and the steering control authority between the human and the machine is calculated by establishing a remote human-machine collaborative controller to achieve the lateral state control of the vehicle. This method does not take into account the realization of human-machine control switching.

[0004] Chinese patent application: Method for Transferring Driving Rights in Alternating Human-Machine Co-driving (Application No.: CN201810253686.3) discloses a method for transferring driving rights in alternating human-machine co-driving, which is characterized by extracting key factors of the driver, environment and vehicle, and using situation estimation to achieve the transfer of driving rights. This method does not consider the delay problem of remote driving, and does not consider the impact of the switching of human-machine co-driving on the stability of the system.

[0005] Chinese patent application: A variable structure control method for human-machine collaborative driving of intelligent vehicles (application number: CN201910079765.1) discloses a variable structure control method for human-machine collaborative driving of intelligent vehicles. It is characterized in that for human-machine collaborative driving, a sliding mode variable structure control method is used to automatically switch vehicle control rights. This method does not consider the switching problem of remote driving, nor the delay problem of remote communication. Summary of the invention

[0006] The present invention aims to solve the above problems of the prior art. A remote human-machine collaborative driving control method for an intelligent vehicle is proposed. The technical solution of the present invention is as follows:

[0007] A remote human-machine collaborative driving control method for an intelligent vehicle comprises the following steps:

[0008] Remote cooperative driving system modeling: Taking the vehicle as the following vehicle, determine the parameters of the following vehicle and traffic scene, derive the traffic flow model, and construct a remote human-machine cooperative driving system model consisting of a remote driver control model and an automatic driving controller control model;

[0009] Calculate the model stability range: According to the stability critical curve of the traffic flow model in the remote cooperative driving system, calculate the stable areas of the remote driver control model and the autonomous driving control model respectively;

[0010] Design a collaborative driving controller: Based on the control quantities corresponding to the remote driver control model and the autonomous driving control model, design a sliding mode switching function so that when the system state reaches the set switching function value, the system smoothly switches from the autonomous driving control structure to the remote driver control structure to achieve smooth switching of human-machine collaborative control.

[0011] Furthermore, the traffic flow model is a multi-vehicle full speed difference model, and the model is as follows:

[0012]

[0013] Among them, σ is the sensitivity coefficient, λ is the reaction coefficient of speed difference, and x n (t), v n (t) and Respectively represent the position, speed and acceleration of the nth vehicle, where m represents the position and speed difference information of m preceding vehicles. n (t) = x n+1 (t)-x n (t) represents the headway between the leading vehicle and the following vehicle at time t, Δv n (t) = v n+1 (t)-v n (t) represents the speed difference between the leading vehicle and the following vehicle at time t, β l is the weight describing the impact of the headway and speed difference of the lth vehicle in front on the following vehicle, V is the optimized speed function, and is the vehicle distance d n (t), which is monotonically increasing and has an upper bound.

[0014] Furthermore, the autonomous driving control model is as follows:

[0015]

[0016] Where x = [a s ,v s ,v t ,d] T is the state vector of the vehicle, a s is the actual acceleration of the following car, v sis the following vehicle speed, v t is the speed of the front vehicle, d is the distance between the following vehicle and the front vehicle, is the ideal acceleration of the vehicle, which is also the output of the controller, A M , B M , C are the state transfer matrix, control matrix and observation matrix respectively; the system observation y M (t) is the distance between the following vehicle and the preceding vehicle.

[0017] Furthermore, the remote driver control model is as follows: τ1 is defined as the communication delay between the remote driver control model and the cooperative driving controller, and τ2 is defined as the communication delay between the remote driver control model and the cooperative driving controller. The remote driver model is as follows:

[0018]

[0019] Among them, x R =[a s ,v s ,v t ,d] T is the state vector of the vehicle, a s is the actual acceleration of the following car, v s is the following vehicle speed, v t is the speed of the front vehicle, d is the distance between the following vehicle and the front vehicle, A R , B R , C R They are the state transfer matrix, control matrix and observation matrix respectively, and the system observation quantity y R (t) is the distance between the following vehicle and the preceding vehicle.

[0020] Furthermore, the calculation of the model stability range specifically includes: respectively calculating the stability range of the traffic flow model, the automatic driving control model and the remote driver control model;

[0021] Given that the initial state of the traffic flow is stable, based on the relationship between the constant parameters in the traffic flow model and the vehicle model variables, the critical curve that allows the initial steady-state flow in the traffic flow model to remain stable after being disturbed is obtained;

[0022]

[0023] Among them, V(d) is the optimized speed function, d is the headway of the vehicle flow, σ is the sensitivity coefficient, λ is the reaction coefficient of the speed difference, and e1 is the stability constant;

[0024] Given that the initial states of the autonomous driving control model and the remote driver control model are the same, the stability ranges of the state variables in the remote driver control model and the autonomous driving control model are obtained through the stability curve of the traffic flow model.

[0025] The stable range of the state quantity of the vehicle automatic driving control model is:

[0026]

[0027] Among them, x M is the vehicle status under autonomous driving, is the stable boundary value of the autonomous driving output, B M is the vehicle control state parameter matrix, D M is the initial state parameter matrix of the machine driving;

[0028] make It is the stable boundary value of the vehicle state under autonomous driving;

[0029] The stable range of the state quantity of the remote driver control model is:

[0030]

[0031] Among them, x R is the vehicle status when the remote driver is driving, is the stable boundary value of the remote driver control output, B R is the remote driver control state parameter matrix, is the communication delay of remote driving, D R is the remote driver's initial state parameter matrix,

[0032] make It is the stable boundary value of the vehicle state under the remote driver.

[0033] Furthermore, the use of the variable structure control method to design a cooperative driving controller includes the following steps:

[0034] Through the remote driver control model and the automatic driving control model, the relationship between the model state quantity and the output control quantity is calculated. According to the output control quantity of the remote driver and the automatic driving system, the control quantity of the variable structure controller before and after the switch is constructed:

[0035]

[0036] According to the control quantity u(t), the vehicle state quantity x = [a s ,v s ,v t ,d] T The following car acceleration value a s As the switching criterion, according to the stable range of the automatic driving control model Get the switching function:

[0037]

[0038] Where I, F, G, and H are state parameter matrices, I = [0 0 0 1], F = [1 0 0 0], G = [0 1 0 0], and H = [0 0 1 0].

[0039] The advantages and beneficial effects of the present invention are as follows:

[0040] The present invention proposes a remote human-machine collaborative driving control method for an intelligent vehicle. Aiming at the control switching problem between the driver and the automatic driving controller and the remote communication delay problem in remote collaborative driving, a collaborative driving control method based on the stable range is designed. The innovation of this method lies in ensuring the stability and safety of the vehicle during the switching process through a clear switching mechanism and communication delay compensation strategy. Specifically, the claims cover technical points such as automatic judgment of driving right switching, dynamic adjustment of stable range and communication delay compensation, thereby effectively improving the safety and reliability of remote human-machine collaborative driving. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a schematic diagram of a model of a remote human-machine collaborative driving control method for an intelligent vehicle provided by the preferred embodiment of the present invention;

[0042] Figure 2 It is a design flow chart of the collaborative driving controller of the present invention.

[0043] Figure 3 It is a simulation example diagram of the present invention. DETAILED DESCRIPTION

[0044] The following will describe the technical solutions in the embodiments of the present invention in detail in conjunction with the accompanying drawings in the embodiments of the present invention. The described embodiments are only part of the embodiments of the present invention.

[0045] The technical solution of the present invention to solve the above technical problems is:

[0046] A remote human-machine cooperative driving control method for intelligent vehicles is firstly constructed, and a remote cooperative driving system model is established. A traffic flow micro-model and a multi-front vehicle full speed difference model are established. The stable regions of the remote driver control model and the automatic driving control model are calculated respectively. According to the control quantity, a corresponding cooperative driving controller is designed to complete the switching of driving rights between automatic driving and remote driving.

[0047] The specific implementation of the present invention is described below with reference to the accompanying drawings and specific examples.

[0048] like Figure 1 The figure shows a model schematic diagram of a remote human-machine collaborative driving control method for an intelligent vehicle of the present invention, comprising:

[0049] 1 Remote cooperative driving system modeling: Taking the vehicle as the following vehicle, determine the parameters of the following vehicle and traffic scene, and construct a remote human-machine cooperative driving system model consisting of a remote driver control model and an automatic driving control model. The model parameters include the communication delay τ1 from the remote driver to the cooperative driving controller, the communication delay τ2 from the cooperative driving controller to the remote driver, the vehicle speed v s , front vehicle speed v t , the distance between the two cars d, the actual acceleration of the car a s , Actual acceleration of the vehicle ahead a t And the vehicle control parameters output by the cooperative driving controller, including the safety distance d0 and the ideal acceleration of the vehicle The human-machine collaborative driving problem is transformed into a switching problem between the remote driver and the autonomous driving controller under a traffic flow model.

[0050] 1.1 Traffic Flow Model

[0051] Construct a traffic flow model with multiple front vehicles and full speed difference:

[0052]

[0053] Among them, σ is the sensitivity coefficient, λ is the reaction coefficient of speed difference, and x n (t), v n (t) and denote the position, velocity and acceleration of the nth vehicle, Δx n (t) = x n+1 (t)-x n (t) represents the headway between the leading vehicle and the following vehicle at time t. l is the weight describing the impact of the headway and speed difference of the lth vehicle in front on the following vehicle. The optimization function v(·) depends on the headway of the preceding vehicle and is the vehicle spacing d n (t), which is monotonically increasing and has an upper bound.

[0054] Among them, β l is a monotonically decreasing function of l, that is, β l <β l-1 , because the impact of the front vehicle on this vehicle decreases as the distance between the front vehicle and this vehicle increases.

[0055] 1.2 Autonomous Driving Control Model

[0056] The state space representation of the vehicle kinematic model is:

[0057]

[0058] Where x = [a s ,vs ,v t ,d] T is the state vector of the vehicle, a s is the actual acceleration of the following car, v s is the following vehicle speed, v t is the speed of the front vehicle, and d is the distance between the following vehicle and the front vehicle. is the ideal acceleration of the vehicle, which is also the output of the controller, A M , B M , C are the state transfer matrix, control matrix and observation matrix respectively.

[0059] in C = [0 0 0 1], ω is the process time, θ is the actuator time constant, and k is the steady-state gain. System observation y M (t) is the distance between vehicles.

[0060] 1.3 Remote driver control model

[0061] The remote driver model is explained using a mechanical simulation model as an example. τ1 is defined as the communication delay between the remote driver control model and the cooperative driving controller, and τ2 is defined as the communication delay between the cooperative driving controller and the remote driver control model. The mechanical simulation model is shown below:

[0062]

[0063] in, C R =[0 0 0 1], ω is the process time, θ is the actuator time constant, and k is the steady-state gain. System observation y R (t) is the distance between vehicles. R , B R , C R They are the state transfer matrix, control matrix and observation matrix respectively.

[0064] 2 Stable range calculation: According to the remote cooperative driving system model and parameters, the stable areas of the remote driver control model and the autonomous driving control model are calculated respectively.

[0065] Given the initial state as a steady state, the critical curve of traffic flow can be obtained.

[0066]

[0067] Given that the remote driver control model has the same initial state x as the autonomous driving control model, the control quantity for the autonomous driving control model is:

[0068]

[0069] Where I = [0 0 0 1], F = [1 0 0 0], G = [0 1 0 0], H = [0 0 1 0], the state quantity of the automatic driving control model is obtained:

[0070]

[0071] For the remote driver control model, the control quantity is:

[0072]

[0073] Obtain the state quantity of the remote driver control model

[0074]

[0075] 3 Design of cooperative driving controller

[0076] In the process of remote human-machine collaborative driving, the remote driver alone cannot switch control in time. A collaborative driving controller is selected to realize the switching between automatic driving control and remote driver control. The switching function is designed based on the known control quantity to make the error converge to zero, and the control right is switched. The acceleration is used as the controller input and the distance between vehicles is used as the controller output.

[0077] like Figure 2 The flowchart of the present invention for designing a cooperative driving controller is shown. The design of the cooperative driving controller includes the following steps:

[0078] According to the corresponding control quantities of the remote driver control model and the automatic driving control model, the switching function is designed, and the parameters and sliding surface are set. When the state of the system reaches the set switching function value, the system smoothly switches from the automatic driving control structure to the remote driver control structure to achieve smooth switching of human-machine collaborative control.

[0079] 3.1 Determine the control quantity: According to the relationship between the state quantity and the output quantity in the autonomous driving control model, the output quantity u in the autonomous driving state can be obtained through the corresponding calculation method. + for:

[0080]

[0081] According to the relationship between the state quantity and the output quantity in the remote driver control model, the output quantity u under the remote driver driving state can be obtained by the corresponding calculation method: - for:

[0082]

[0083] The output of the autonomous driving system corresponds to the control quantity before and after the switch, so the comprehensive autonomous driving output u +and the remote driver driving output u - , the control quantity of the cooperative driving controller is

[0084]

[0085] 3.2 Designing the switching function

[0086] The state quantity x in the vehicle kinematic model is s ,v s ,v t ,d] T The acceleration value a of the vehicle in s As the switching criterion, the sliding mode of the system can achieve asymptotic stability and have good dynamic performance. According to the stable region, the corresponding switching function is designed.

[0087]

[0088] Figure 3 This is an example diagram of a simulation of a collaborative driving controller designed by the present invention. Figure 3 (a) The solid line in the figure represents the control quantity output by the remote driver, the dashed line represents the control quantity output by the automatic driving, and the dotted line represents the final output of the cooperative driving controller. Figure 3 (b) The solid line in the middle represents the acceleration of the following vehicle, and the dotted line represents the acceleration of the leading vehicle; Figure 3 (c) The solid line in the middle indicates the distance between the following vehicle and the vehicle ahead; Figure 3 (d) The solid line in the middle represents the speed of the following vehicle, and the dotted line represents the speed of the leading vehicle.

[0089] Figure 3 The simulation scenario is: the speed of the vehicle is 21.1m / s, the speed of the front vehicle is 16.6m / s, the distance between the following vehicle and the front vehicle is 60m, and the initial acceleration of the front vehicle is 0. It can be seen that before 1.1s, the system judged that the state was stable, and the following vehicle did not make any control operation on the acceleration. But after 1.1s, it was judged to be unstable, and the automatic driving control made a deceleration response before the remote driver, and then made a driving conversion operation, and the sliding mode controller executed the control instructions made by the automatic driving after the judgment. The speed and acceleration of the following vehicle can be stably reached to the ideal value in a very short time. The second controller switching was carried out at about 2.8s, and the speed slowly decreased. Finally, the following vehicle reached a stable value at about 3.2s, which was equal to the speed and acceleration of the front vehicle. In the process of switching control rights, especially after the system judged that the state was unstable, the automatic driving system made a deceleration operation earlier than the remote driver, and the entire switching process did not cause sudden fluctuations in vehicle speed or acceleration. Through the adjustment of the cooperative driving controller, the following vehicle can quickly and smoothly reach the ideal speed and acceleration values. During the subsequent control switching process, the vehicle's speed and acceleration can still maintain smooth changes and eventually remain consistent with the state of the vehicle in front.

[0090] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions.

[0091] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0092] The above embodiments should be understood to be only used to illustrate the present invention and not to limit the protection scope of the present invention. After reading the contents of the present invention, technicians can make various changes or modifications to the present invention, and these equivalent changes and modifications also fall within the scope defined by the claims of the present invention.

Claims

1. A remote human-machine collaborative driving control method for an intelligent vehicle, characterized in that: The following steps are involved: Remote cooperative driving system modeling: Taking the vehicle as the following vehicle, determine the parameters of the following vehicle and traffic scene, derive the traffic flow model, and construct a remote human-machine cooperative driving system model consisting of a remote driver control model and an automatic driving controller control model; Calculate the model stability range: According to the stability critical curve of the traffic flow model in the remote cooperative driving system, calculate the stable areas of the remote driver control model and the autonomous driving control model respectively; Design a collaborative driving controller: Based on the control quantities corresponding to the remote driver control model and the autonomous driving control model, design a sliding mode switching function so that when the system state reaches the set switching function value, the system smoothly switches from the autonomous driving control structure to the remote driver control structure to achieve smooth switching of human-machine collaborative control.

2. The intelligent vehicle remote human-machine collaborative driving control method according to claim 1 is characterized in that: The traffic flow model is a multi-front vehicle full speed difference model, and the model is as follows: Among them, σ is the sensitivity coefficient, λ is the reaction coefficient of speed difference, and x n (t), v n (t) and Respectively represent the position, speed and acceleration of the nth vehicle, where m represents the position and speed difference information of m preceding vehicles. n (t) = x n+1 (t)-x n (t) represents the headway between the leading vehicle and the following vehicle at time t, Δv n (t) = v n+1 (t)-v n (t) represents the speed difference between the leading vehicle and the following vehicle at time t, β l is the weight describing the impact of the headway and speed difference of the lth vehicle in front on the following vehicle, V is the optimized speed function, and is the vehicle distance d n (t), which is monotonically increasing and has an upper bound.

3. The intelligent vehicle remote human-machine collaborative driving control method according to claim 2 is characterized in that: The autonomous driving control model is as follows: Where x = [a s ,v s ,v t ,d] T is the state vector of the vehicle, a s is the actual acceleration of the following car, v s is the following vehicle speed, v t is the speed of the front vehicle, d is the distance between the following vehicle and the front vehicle, is the ideal acceleration of the vehicle, which is also the output of the controller, A M , B M , C are the state transfer matrix, control matrix and observation matrix respectively; the system observation y M (t) is the distance between the following vehicle and the preceding vehicle.

4. The intelligent vehicle remote human-machine collaborative driving control method according to claim 3 is characterized in that: The remote driver control model is: τ1 is defined as the communication delay between the remote driver control model and the cooperative driving controller, and τ2 is defined as the communication delay between the remote driver control model and the cooperative driving controller. The remote driver model is as follows: Among them, x R =[a s ,v s ,v t ,d] T is the state vector of the vehicle, a s is the actual acceleration of the following car, v s is the following vehicle speed, v t is the speed of the front vehicle, d is the distance between the following vehicle and the front vehicle, A R , B R , C R They are the state transfer matrix, control matrix and observation matrix respectively, and the system observation quantity y R (t) is the distance between the following vehicle and the preceding vehicle.

5. The intelligent vehicle remote human-machine collaborative driving control method according to claim 4 is characterized in that: The calculation model stability range specifically includes: respectively calculating the stability range of the traffic flow model, the automatic driving control model and the remote driver control model; Given that the initial state of the traffic flow is stable, based on the relationship between the constant parameters in the traffic flow model and the vehicle model variables, the critical curve that allows the initial steady-state flow in the traffic flow model to remain stable after being disturbed is obtained; Among them, V(d) is the optimized speed function, d is the headway of the vehicle flow, σ is the sensitivity coefficient, λ is the reaction coefficient of the speed difference, and e1 is the stability constant; Given that the initial states of the autonomous driving control model and the remote driver control model are the same, the stability ranges of the state variables in the remote driver control model and the autonomous driving control model are obtained through the stability curve of the traffic flow model. The stable range of the state quantity of the vehicle automatic driving control model is: Among them, x M is the vehicle status under autonomous driving, is the stable boundary value of the autonomous driving output, B M is the vehicle control state parameter matrix, D M is the initial state parameter matrix of the machine driving; make It is the stable boundary value of the vehicle state under autonomous driving; The stable range of the state quantity of the remote driver control model is: Among them, x R is the vehicle status when the remote driver is driving, is the stable boundary value of the remote driver control output, B R is the remote driver control state parameter matrix, is the communication delay of remote driving, D R is the remote driver's initial state parameter matrix, make It is the stable boundary value of the vehicle state under the remote driver.

6. The intelligent vehicle remote human-machine collaborative driving control method according to claim 5 is characterized in that: The variable structure control method is used to design a cooperative driving controller, including the following steps: Through the remote driver control model and the automatic driving control model, the relationship between the model state quantity and the output control quantity is calculated. According to the output control quantity of the remote driver and the automatic driving system, the control quantity of the variable structure controller before and after the switching is constructed: According to the control quantity u(t), the vehicle state quantity x = [a s ,v s ,v t ,d] T The following car acceleration value a s As the switching criterion, according to the stable range of the automatic driving control model Get the switching function: Where I, F, G, and H are state parameter matrices, I = [0 0 0 1], F = [1 0 0 0], G = [0 1 0 0], and H = [0 01 0].

Images