A human-machine division of labor and collaborative method for longitudinal following considering multiple leading vehicles and communication delay

Abstract

The present invention discloses a longitudinal following human-machine division of labor and collaborative method considering multiple preceding vehicles and communication delay, comprising the following steps: S1, limiting the configuration, following tasks and scenarios of the intelligent vehicle; S2, constructing a longitudinal dynamics model of the vehicle; S3, constructing a longitudinal following control model considering multiple preceding vehicles and communication delay; S4, constructing a human-machine division of labor and collaborative model according to the allocation of following tasks; S5, performing stability analysis on the vehicle speed following control system; S6, designing a guaranteed cost controller; S7, designing a robust H-infinity controller. The present invention provides a method for performing stability analysis on the vehicle speed following control system using the Lyapunov-Krasovskii functional, and designs a guaranteed cost controller and a γ ‑ Suboptimal H ∞ The robust controller allows the main vehicle to not only perform pre-control based on the information of the vehicle farther ahead, but also perform further refined control based on the information of the vehicle nearby, thereby optimizing the vehicle's longitudinal following performance, reducing the driver's operating load and improving driving comfort.

Description

Technical Field

[0001] The present invention belongs to the technical field of human-machine co-driving of intelligent vehicles, and specifically relates to a longitudinal following human-machine division of labor and collaboration method taking into account multiple leading vehicles and communication delays. Background Art

[0002] Due to the complexity of traffic issues, the reliability of perception equipment, and price costs, truly full-condition autonomous driving will be difficult to achieve in the short term. There will be a long period of time in the future when people and systems will jointly participate in vehicle control and the human-machine co-driving stage will take place.

[0003] With the development of sensor technology and communication technology, vehicles can use radar to obtain unobstructed obstacle targets within a certain range around the vehicle, and can also use vehicle-to-vehicle communication technology to obtain targets that are not detected by on-board sensors. Multiple sources of information are used together for vehicle motion control. For vehicle systems with large inertia, making full use of on-board sensors and wireless communication data can improve the vehicle's control performance and adapt to real road traffic environments.

[0004] Patent document CN111562739A discloses a human-machine hybrid intelligent collaborative following control method that keeps the driver in the loop. This method obtains the relative vehicle spacing and relative vehicle speed between the two vehicles through sensors, and then divides the longitudinal following tasks and formulates corresponding controllers to design a human-machine collaborative following control method. However, for real scenarios, relying on the information of a single leading vehicle for control will be affected by factors such as the mechanical transmission efficiency and inertia delay of the vehicle transmission system and the response time of the leading vehicle, resulting in hysteresis in the vehicle perception-decision-making-control execution process. This method fails to give full play to the advantages of wireless communication technology in terms of long perception distance and not being blocked by obstacles. Summary of the invention

[0005] In view of this, the present invention provides a longitudinal following human-machine division of labor and collaborative method considering multiple front vehicles and communication delay. The human-machine division of labor and collaborative method can avoid the safety problem of people not being in the loop. In addition, the status information of multiple vehicles in front is obtained through wireless communication technology, and the influence of communication delay is considered, so that the design of the controller is more in line with the real road traffic environment, which can further reduce the operating load of the driver and improve the safety and comfort of the longitudinal following of the vehicle. The method includes the following steps:

[0006] 1) Limit the configuration, following tasks and scenarios of smart cars

[0007] 2) Construct the longitudinal dynamics model of the vehicle

[0008] 3) Construct a longitudinal following control model that takes into account multiple leading vehicles and communication delays

[0009] 4) Build a human-machine division of labor and collaboration model based on the allocation of vehicle-following tasks

[0010] 5) Stability analysis of vehicle speed following control system

[0011] 6) Design a guaranteed cost controller

[0012] 7) Design of robust H-infinity controller

[0013] Furthermore, in step 1), the smart car is equipped with an electronically controlled actuator, a vehicle-mounted radar, and a V2X communication device, so that the main vehicle and multiple leading vehicles are in the same lane. The main vehicle can obtain the relative position Δx, speed Δv, and acceleration Δa between the main vehicle and the adjacent leading vehicle through the vehicle-mounted radar, and the main vehicle can obtain the speed v of the farther leading vehicle through the V2X communication device. p2 and acceleration a p2 .

[0014] Furthermore, in step 2), considering that the real vehicle dynamics are affected by a variety of nonlinear factors and are not suitable for modeling and simulation, a simplified vehicle dynamics model is constructed. The relationship between the acceleration and the expected acceleration in the simplified model is:

[0015]

[0016] Where a is the actual acceleration; u is the expected acceleration; μ is the engine time constant; and s is the complex frequency.

[0017] In the state space model of vehicle dynamics, the desired acceleration u is taken as the control input of the system, and the third-order longitudinal dynamic model of the vehicle is as follows:

[0018]

[0019] Among them, x(t) is the position of the vehicle at time t; v(t) is the speed of the vehicle at time t; a(t) is the acceleration of the vehicle at time t.

[0020] Define the state variable X(t) = [x(t), v(t), a(t)] T , the third-order state space model of the vehicle is as follows:

[0021]

[0022] in

[0023]

[0024] Furthermore, in step 3), the following position error between the main vehicle and the adjacent preceding vehicle is defined as:

[0025] δ(t)=xp1 (t)-x ego (t)-d

[0026] Among them, x p1 (t), x ego (t), d are the position of the nearest preceding vehicle at time t, the position of the main vehicle and the expected following distance. The following speed error between the main vehicle and the nearest preceding vehicle is further defined as:

[0027]

[0028] Among them, v p1 (t), v ego (t) are the speed of the vehicle in front and the speed of the main vehicle at time t. The following acceleration error between the main vehicle and the vehicle in front is further defined as:

[0029]

[0030] Among them, a p1 (t), a ego (t) are the speed of the preceding vehicle and the acceleration of the host vehicle at time t, respectively.

[0031] Defining state variables and measuring outputs

[0032]

[0033]

[0034] The longitudinal following control model considering multiple front vehicle information can be obtained as follows:

[0035]

[0036] in,

[0037]

[0038] The output equation of the system is:

[0039] y(t)=C s x(t)+C c x(t)

[0040] Among them C s +C c =C, C is the output matrix of the system

[0041] Design the output feedback controller as:

[0042] u(t)=K s y s (t)+K c y c (t)

[0043] Where K s =[k p k v1 k a1 0 0],K c =[0 0 0 k v2 k a2 ]

[0044] Considering the communication delay and external disturbance, the longitudinal following control model considering multiple leading vehicle information and communication delay can be obtained as follows:

[0045]

[0046] Furthermore, for step 4), the allocation of the following task is considered. The driver is responsible for the inter-vehicle distance control task. The driver's performance is as follows: when the following distance is greater than the expected following distance, the driver steps on the accelerator. Otherwise, the driver steps on the brake. The control output is:

[0047] u human (t) = K human δ(t)

[0048] Among them, K human The gain for the driver to control the distance between vehicles is not a fixed value.

[0049] The driving automation system is responsible for the vehicle speed following task. Since it does not need to consider the distance tracking error between the main vehicle and the adjacent vehicle ahead, the control model in step 3) can be temporarily reduced to obtain a closed-loop vehicle speed following control system:

[0050]

[0051] in,

[0052]

[0053]

[0054] At this time, the feedback controller of the vehicle speed following control system is:

[0055] u auto (t) = K * y * (t)

[0056] Furthermore, for step 5), the following is a condition for the asymptotic stability of the vehicle speed following control system. If there are symmetric positive matrices P and N, and matrix K * , so that the following matrix inequality holds

[0057]

[0058] Then the vehicle speed following control system is asymptotically stable, where

[0059]

[0060] Θ 12 =PB *

[0061] Θ 21 =B *T P

[0062] Θ 22 =-N

[0063] Furthermore, for step 6), define the performance indicators:

[0064]

[0065] in is a given symmetric positive definite matrix.

[0066] For a closed-loop vehicle speed following control system and a given performance index, if there exist symmetric positive definite matrices X and G, and matrices Y and Z such that the following linear matrix inequality

[0067]

[0068] Established, where * indicates the symmetry of the matrix, the guaranteed cost controller of the vehicle speed following control system can be obtained. In addition, the gain of the guaranteed cost controller can be calculated as follows:

[0069]

[0070]

[0071] Furthermore, for step 7), the following performance indicators are defined:

[0072]

[0073] Among them, for a given γ>0, all ω(t)∈l2[0,∞) should satisfy J<0.

[0074] For a closed-loop vehicle speed following control system and a given constant γ>0, if there exist symmetric positive definite matrices X and G, and matrices Y and Z such that the following linear matrix inequality

[0075]

[0076] holds, where * represents the symmetry of the matrix, then the γ-suboptimal H of the vehicle speed following control system can be obtained ∞Robust controller. In addition, the gain of the H-infinity controller can be obtained as follows:

[0077]

[0078]

[0079] Beneficial effects:

[0080] The longitudinal following human-machine division of labor and cooperation method considering multiple preceding vehicles and communication delay provided by the present invention has few scholars exploring the issues such as the influence of multiple preceding vehicle information and communication delay on longitudinal following human-machine cooperation. The stability analysis of the vehicle speed following control system is carried out by using the Lyapunov-Krasovskii functional, and on this basis, a guaranteed cost controller and a γ-suboptimal H are designed. ∞ The robust controller allows the main vehicle to not only perform pre-control based on the information of the vehicle farther ahead, but also perform further refined control based on the information of the vehicle nearby, thereby optimizing the vehicle's longitudinal following performance, reducing the driver's operating load and improving driving comfort. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 It is a schematic diagram of the overall process of the present invention;

[0082] Figure 2 Schematic diagram of longitudinal following taking into account information of multiple preceding vehicles. DETAILED DESCRIPTION

[0083] The following is an explanation of the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention.

[0084] Example 1

[0085] like Figure 1 As shown, a longitudinal following human-machine division of labor and collaboration method considering multiple leading vehicles and communication delays includes the following steps:

[0086] Step 1) Define smart car configuration, following tasks and scenarios

[0087] Equip the smart car with an electronically controlled actuator, on-board radar, and V2X communication device, so that the main vehicle and multiple vehicles ahead are in the same lane. The main vehicle can obtain the relative position Δx, speed Δv, and acceleration Δa between the main vehicle and the adjacent vehicle ahead through the on-board radar, and the main vehicle can obtain the speed v of the farther vehicle ahead through the V2X communication device. p2 and acceleration ap2 .

[0088] Step 2) Construct the longitudinal dynamics model of the vehicle. This includes the following sub-steps:

[0089] Step 2.1: Considering that real vehicle dynamics are affected by many nonlinear factors and are not suitable for modeling and simulation, a simplified vehicle dynamics model is constructed. The relationship between the acceleration in the simplified model and the expected acceleration is:

[0090]

[0091] Where a is the actual acceleration; u is the expected acceleration; μ is the engine time constant;

[0092] Step 2.2: In the state-space model of vehicle dynamics, the desired acceleration u is taken as the control input of the system, and the third-order longitudinal dynamic model of the vehicle is as follows:

[0093]

[0094] Among them, x(t) is the position of the vehicle at time t; v(t) is the speed of the vehicle at time t; a(t) is the acceleration of the vehicle at time t.

[0095] Step 2.3: Define state variables X(t) = [x(t), v(t), a(t)] T , the third-order state space model of the vehicle is as follows:

[0096]

[0097] in

[0098]

[0099] Step 3) Construct a longitudinal following control model that takes into account multiple leading vehicles and communication delays. This includes the following sub-steps:

[0100] Step 3.1: Define the following position error between the main vehicle and the adjacent preceding vehicle as:

[0101] δ(t)=x p1 (t)-x ego (t)-d

[0102] Among them, x p1 (t), x ego (t) and d are the position of the leading vehicle, the position of the main vehicle and the expected following distance at time t, respectively.

[0103] Step 3.2: Define the following speed error between the main vehicle and the adjacent preceding vehicle as:

[0104]

[0105] Among them, v p1 (t), v ego (t) are the speed of the preceding vehicle and the main vehicle at time t, respectively.

[0106] Step 3.3: Define the following acceleration error between the main vehicle and the adjacent preceding vehicle as:

[0107]

[0108] Among them, a p1 (t), a ego (t) are the speed of the preceding vehicle and the acceleration of the host vehicle at time t, respectively.

[0109] Step 3.4: Define state variables and measure outputs:

[0110]

[0111]

[0112] in,

[0113]

[0114] Step 3.5: Define the output equation of the system:

[0115] y(t)=C s x(t)+C c x(t)

[0116] Among them C s +C c =C, C is the output matrix of the system

[0117] Step 3.6: Design the output feedback controller:

[0118] u(t)=K s y s (t)+K c y c (t)

[0119] in,

[0120] K s =[k p k v1 k a1 0 0],K c =[0 0 0 k v2 k a2 ]

[0121] Step 3.7: Considering the communication delay and external disturbance, the longitudinal following control model considering multiple leading vehicle information and communication delay can be obtained as:

[0122]

[0123] Step 4) According to the allocation of the following task, a human-machine division of labor and collaboration model is constructed. It includes the following sub-steps:

[0124] Step 4.1: Consider the allocation of the following task. The driver is responsible for the inter-vehicle distance control task. The driver's performance is as follows: when he realizes that the following distance is greater than the expected following distance, he steps on the accelerator. Otherwise, he steps on the brake. The control output is:

[0125] u human (t) = K human δ(t)

[0126] Among them, K human The gain for the driver to control the distance between vehicles is not a fixed value.

[0127] Step 4.2: The driving automation system is responsible for the vehicle speed following task. Since it does not need to consider the distance tracking error between the main vehicle and the adjacent vehicle ahead, the control model in step 3.7) can be temporarily reduced to obtain a closed-loop vehicle speed following control system:

[0128]

[0129] in,

[0130]

[0131]

[0132] At this time, the feedback controller of the vehicle speed following control system is:

[0133] u auto (t) = K * y * (t)

[0134] Step 5) Perform stability analysis on the vehicle speed following control system.

[0135] If there exist symmetric positive definite matrices P and N, and a matrix K * , so that the following matrix inequality holds

[0136]

[0137] Then the vehicle speed following control system is asymptotically stable, where

[0138]

[0139] Θ 12 =PB *

[0140] Θ 21 =B *T P

[0141] Θ 22 =-N

[0142] Step 6) Design a guaranteed cost controller. This includes the following sub-steps:

[0143] Step 6.1: Define performance metrics:

[0144]

[0145] in is a given symmetric positive definite matrix.

[0146] Step 6.2: For a closed-loop vehicle speed following control system and a given performance indicator, if there exist symmetric positive definite matrices X and G, and matrices Y and Z such that the following linear matrix inequality

[0147]

[0148] Established, where * indicates the symmetry of the matrix, the guaranteed cost controller of the vehicle speed following control system can be obtained. In addition, the gain of the guaranteed cost controller can be calculated as follows:

[0149]

[0150]

[0151] Step 7) Design a robust H-infinity controller. This includes the following sub-steps:

[0152] Step 7.1: Define the following performance indicators:

[0153]

[0154] Among them, for a given γ>0, all ω(t)∈l2[0,∞) should satisfy J<0.

[0155] Step 7.2: For a closed-loop vehicle speed following control system and a given constant γ>0, if there exist symmetric positive definite matrices X and G, and matrices Y and Z such that the following linear matrix inequality

[0156]

[0157] holds, where * represents the symmetry of the matrix, then the γ-suboptimal H of the vehicle speed following control system can be obtained ∞Robust controller. In addition, the gain of the H-infinity controller can be obtained as follows:

[0158]

[0159]

[0160] The above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solutions of the present invention, which should be included in the scope of the claims of the present invention.

Claims

1. A longitudinal following human-machine division of labor and collaboration method considering multiple leading vehicles and communication delay, characterized in that: The following steps are involved: S1. Limit the configuration, following tasks and scenarios of smart cars; S2. Construct the longitudinal dynamics model of the vehicle; S21. Construct a simplified vehicle dynamics model. The relationship between the acceleration and the expected acceleration in the simplified model is as follows: Where a is the actual acceleration; u is the expected acceleration; μ is the engine time constant; s is the complex frequency; S22. In the state space model of vehicle dynamics, the desired acceleration u is used as the control input of the system, and the third-order longitudinal dynamic model of the vehicle is obtained as follows: Where p(t) is the position of the vehicle at time t; v(t) is the speed of the vehicle at time t; a(t) is the acceleration of the vehicle at time t; S23. Define state variable X(t) = [p(t), v(t), a(t)] T , the third-order state space model of the vehicle is obtained as follows: in S3, construct a longitudinal following control model considering multiple leading vehicles and communication delay; S31. Define the following position error between the main vehicle and the adjacent preceding vehicle as: δ(t)=p p1 (t)-p ego (t)-d Among them, p p1 (t), p ego (t) and d are the position of the preceding vehicle, the position of the main vehicle and the expected following distance at time t, respectively; S32. Define the following speed error between the main vehicle and the adjacent preceding vehicle as: Among them, v p1 (t), v ego (t) are the speed of the preceding vehicle and the main vehicle at time t, respectively; S33. Define the following vehicle acceleration error between the main vehicle and the adjacent preceding vehicle as: Among them, a p1 (t), a ego (t) are the speed of the preceding vehicle and the acceleration of the host vehicle at time t; S34. Define state variables and measurement outputs: The longitudinal following control model considering multiple front vehicle information is obtained as follows: in, The output equation of the system is: y(t)=C s x(t)+C c x(t) Among them, C s +C c =C, C is the output matrix of the system; S35. Design the output feedback controller as: u(t)=K s y s (t)+K c y c (t) Among them, K s =[k p k v1 k a1 0 0],K c =[0 0 0 k v2 k a2 ]; S36. Considering the communication delay and external disturbance, the longitudinal following control model considering multiple front vehicle information and communication delay is obtained as follows: S4. According to the allocation of vehicle-following tasks, a human-machine division of labor and collaboration model is constructed; S41. According to the allocation of the following task, the driver is responsible for the inter-vehicle distance control task. When the following distance is greater than the expected following distance, the driver steps on the accelerator. When the following distance is less than the expected following distance, the driver steps on the brake. The control output is: u human (t)=K human δ(t) Among them, K human The gain for the driver to control the distance between vehicles is an unfixed value; S42: When the driving automation system is responsible for the vehicle speed following task, it does not need to consider the vehicle distance tracking error between the main vehicle and the adjacent vehicle ahead. Then the control model in S36 is temporarily reduced to obtain a closed-loop vehicle speed following control system: in, At this time, the feedback controller of the vehicle speed following control system is: u auto (t)=K * y * (t) S5. Conduct stability analysis on the vehicle speed following control system; S6. Design a guaranteed cost controller; S7. Design a robust H-infinity controller.

2. The longitudinal following human-machine division of labor and collaborative method considering multiple preceding vehicles and communication delay according to claim 1 is characterized in that: The S1 comprises the following steps: S11. Equip the smart car with an electronically controlled actuator, a vehicle-mounted radar and a V2X communication device, and place the main vehicle and multiple vehicles ahead in the same lane; S12, the main vehicle obtains the relative position Δx, speed Δv and acceleration Δa between the main vehicle and the adjacent preceding vehicle through the vehicle-mounted radar; S13: The main vehicle obtains the speed v of the vehicle ahead via the V2X communication device. p2 and acceleration a p2 .

3. The longitudinal following human-machine division of labor and collaborative method considering multiple preceding vehicles and communication delay according to claim 2 is characterized in that: The S5 is specifically: If there exist symmetric positive definite matrices P and N and matrix K * , so that the following matrix inequality holds Then the vehicle speed following control system is asymptotically stable, where I 12 =PB * I 21 =B *T P I 22 =-N.

4. The longitudinal following human-machine division of labor and collaborative method considering multiple preceding vehicles and communication delay according to claim 3 is characterized in that: The S6 comprises the following steps: S61. Define the following performance indicators: in is a given symmetric positive definite matrix; S62. For a closed-loop vehicle speed following control system and a defined performance index, if there exist symmetric positive definite matrices X and G, and matrices Y and Z such that the following linear matrix inequality holds, Then the guaranteed cost controller and the gain of the guaranteed cost controller of the vehicle speed following control system can be obtained, where * represents the symmetry of the matrix, and the gain of the guaranteed cost controller is calculated as follows:

5. The longitudinal following human-machine division of labor and collaborative method considering multiple preceding vehicles and communication delay according to claim 4 is characterized in that: The S7 comprises the following steps: S71. Define the following performance indicators: Among them, for a given γ>0, all ω(t)∈l2[0,∞), should satisfy J<0; S72. For a closed-loop vehicle speed following control system and a given constant γ>0, if there exist symmetric positive definite matrices X and G, and matrices Y and Z such that the following linear matrix inequality holds, Then the γ-suboptimal H of the vehicle speed following control system is obtained ∞ The gains of the robust controller and the H-infinity controller, where * indicates the symmetry of the matrix, the gains of the H-infinity controller are as follows:

Images