CN116279475B - Cooperative control method for consistency of operation speeds of network-connected automatic driving vehicle queues - Google Patents

Abstract

The invention discloses a cooperative control method for consistency of queue running speeds of an online automatic driving vehicle, which utilizes an intelligent driver model of a comprehensive online vehicle driving strategy to establish an online running environment of the automatic driving vehicle; each network automatic driving vehicle in the network running environment obtains the motion state information of the vehicle and sends the motion state information to a control center; the control center calculates the optimal acceleration and deceleration control strategy of each networked automatic driving vehicle according to the received motion state information, and returns the optimal acceleration and deceleration control strategy to each networked automatic driving vehicle; and each network automatic driving vehicle regulates and controls the speed according to the optimal acceleration and deceleration control strategy, so that the consistency cooperative control of the running speeds of the network automatic driving vehicles in a queue is realized. The invention takes CAV vehicles as main bodies, aims to reduce the speed difference of adjacent vehicles, and provides a method for promoting the consistency of CAV queue speeds, thereby realizing the high-efficiency and stable operation of high-density CAV in the same lane.

Description

Coordinated control method for speed consistency of networked autonomous driving vehicle queues

Technical field

The invention relates to the field of intelligent transportation vehicle-vehicle collaborative automatic control, and in particular to a network-connected automatic driving vehicle queue running speed consistency collaborative control method.

Background technique

As the number of motor vehicles in our country continues to grow, traffic congestion has become increasingly severe. Under high-density traffic flow, stop-and-go traffic is more likely to cause concussive propagation in the transportation system, further aggravating traffic congestion and poor access. phenomenon, so it is increasingly important to ensure efficient and smooth traffic on urban roads.

Connected autonomous vehicles gradually carry out intelligent decision-making and collaborative control through vehicle-vehicle collaboration, vehicle-road collaboration, instant messaging, network interoperability and other technical means. The proposal and development of the above technologies provide a higher level of collaborative driving for connected autonomous vehicles. , platoon operation and other operating modes provide strong technical support. By realizing speed optimization between adjacent vehicles, adjacent vehicles are promoted to run in queues and tend to speed consistency control, which is helpful for reducing the speed difference of adjacent vehicles and reducing traffic flow. The amplitude of speed fluctuations and slowing down the speed oscillation of the traffic system are of practical significance. Some existing related technologies can eliminate or slow down traffic wave oscillations through dynamic guidance of the speed of autonomous vehicles. Existing vehicle collaboration methods cannot alleviate this problem through the most direct queuing speed.

Contents of the invention

Purpose of the invention: In view of the above problems, the purpose of the present invention is to provide a method for coordinated control of the speed consistency of networked autonomous driving vehicle queues to reduce traffic shock and mitigate the frequent stop-and-go phenomenon in real traffic systems.

Technical solution: A method for collaborative control of network-connected autonomous driving vehicle queue running speed consistency according to the present invention, including:

Use an intelligent driver model that integrates connected vehicle driving strategies to establish a connected operating environment for autonomous vehicles;

In the networked operating environment, each networked autonomous vehicle obtains its own motion status information and sends the motion status information to the control center;

The control center calculates the optimal acceleration and deceleration control strategy for each connected autonomous vehicle based on the received motion status information, and returns the optimal acceleration and deceleration control strategy to each connected autonomous vehicle;

Each connected autonomous driving vehicle regulates its own speed according to the optimal acceleration and deceleration control strategy to achieve consistent coordinated control of the speed of the connected autonomous driving vehicle queue.

Further, the motion status information includes running speed, acceleration and deceleration, relative position, and the distance between two adjacent connected autonomous driving vehicles.

Furthermore, the expression of the intelligent driver model of the comprehensive networked vehicle driving strategy is:

In the formula, v _i represents the running speed of the connected autonomous vehicle CAV _i , Li represents the distance between the connected autonomous vehicle i and the preceding vehicle CAV _i-1 , _{and Δv i represents} the distance between CAV _i and the preceding vehicle CAV _i-1. The speed difference between , i represents the i-th CAV, represents the first derivative of running speed v _i at time t, a _max represents the maximum acceleration of CAV,/> represents the expected speed of CAV _i , δ represents the free acceleration index, L ^* represents the expected distance between two adjacent CAVs, /> represents the minimum safe distance between two adjacent connected autonomous vehicles CAV, t _i represents the decision-making time of CAV _i , -a _min represents the maximum deceleration of CAV, a _i represents the acceleration of CAV _i , -a _i represents CAV _i deceleration.

Furthermore, when a distributed collaborative control strategy is adopted, each two adjacent CAVs are taken as a unit, and the preceding vehicle in each control unit is used as the control center of the control unit; when a global collaborative control strategy is adopted, the head of the entire queue The car serves as a control center.

Furthermore, when a distributed collaborative control strategy is adopted, for a queue consisting of n CAVs, model predictive control MPC is used to coordinate and control the operation of vehicles in the queue. At each sampling time point t _k , the control center controls the optimal acceleration and deceleration. The strategy is sent to adjacent connected autonomous driving vehicles for collaborative control.

Furthermore, when a global collaborative control strategy is adopted, for a queue consisting of n CAVs, model predictive control MPC is used to coordinate and control the operation of vehicles in the queue. At each sampling time point t _k , the control center controls the optimal acceleration and deceleration. The strategy is sent to each connected autonomous vehicle for collaborative control.

Further, the distributed collaborative control strategy includes:

Starting from the head vehicle of the connected driving vehicle queue, the two adjacent connected autonomous driving vehicles CAV _n-1 and CAV _n are controlled in sequence. The control center is CAV _n-1 , and the average speed of the current control unit is calculated. and the common expected speed v _d,n , the expressions are respectively:

In the formula, n-1 is the number of the n-1th CAV, v _i-in represents the initial speed of CAV _i , t _k represents the sampling time point, Δt represents the prediction period, a _best is the control input, indicating CAV _n-1 and the optimal acceleration and deceleration of CAV _n , M represents the number of prediction cycles, CAV ₁ -CAV _n-1 has achieved speed consistency coordinated control within M prediction cycles, m represents the prediction cycle number;

During the prediction period [t _k+ MΔt,t _k +(M+1)Δt], when When , the running speed of CAV _n is optimized as/> when When , the running speed of CAV _n is optimized to v _d,n .

Further, the global collaborative control strategy includes:

Taking a queue of n CAVs as the control unit, the average speed of the current control unit is calculated through the control center and the common expected speed v _d,n , the expressions are respectively:

In the formula, a _best is the control input, which represents the optimal acceleration and deceleration of all CAVs;

During the prediction period [t _k ,t _k +Δt], when When , the running speed of CAV _n is optimized as/> When/> When , the running speed of CAV _n is optimized to v _d,n .

Beneficial effects: Compared with the existing technology, the significant advantages of the present invention are: the present invention fully considers the possibility of stop-and-go and other traffic wave oscillations during the operation of high-density CAVs in the same lane in a connected autonomous driving environment. Phenomenon, it is proposed to use CAV vehicles as the main body, with the purpose of narrowing the speed gap between adjacent vehicles, and proposes a method to promote the speed consistency of CAV queues, achieving high-efficiency and stable operation of high-density CAVs in the same lane, and effectively It provides technical reference for the development of intelligent transportation, connected autonomous driving, vehicle-vehicle collaboration, and queue control.

Description of drawings

Figure 1 is a flow chart of the coordinated control method for speed consistency of networked autonomous driving vehicle queues;

Figure 2 is a flow chart of distributed collaborative control strategy;

Figure 3 is a schematic diagram of the CAV queue speed consistency collaborative control strategy in a distributed network-connected autonomous driving environment;

Figure 4 is a global collaborative control strategy flow chart;

Figure 5 is a schematic diagram of the CAV queue speed consistency collaborative control method in a global connected autonomous driving environment;

Figure 6 is a diagram showing the effect of CAV queue speed consistency collaborative control in a distributed network-connected autonomous driving environment;

Figure 7 is a diagram showing the effect of CAV queue speed consistency collaborative control in a global connected autonomous driving environment.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present application more clear, the present application will be further described in detail below with reference to the drawings and embodiments.

Figure 1 is a flow chart of the network-connected autonomous driving vehicle queue running speed consistency collaborative control method described in this embodiment, which includes the following steps:

(1) Use an intelligent driver model that integrates connected vehicle driving strategies to establish a connected operating environment for autonomous vehicles;

The above steps are based on CVDS-IDM to establish a connected operating environment for autonomous vehicles and realize vehicle-to-vehicle interconnection during the operation of autonomous vehicles. The vehicles in the connected operating environment are all connected autonomous vehicles. Each vehicle is in Information sharing, collaborative decision-making and intelligent interconnection between vehicles can be achieved on the premise of realizing autonomous operation decision-making in a connected operating environment;

(2) In the connected operating environment, each connected autonomous vehicle obtains its own motion status information and sends the motion status information to the control center;

The above-mentioned control center refers to the vehicle whose control unit has control functions;

(3) The control center calculates the optimal acceleration and deceleration control strategy for each connected autonomous vehicle based on the received motion status information, and returns the optimal acceleration and deceleration control strategy to each connected autonomous vehicle;

The above-mentioned optimal acceleration and deceleration control strategy refers to the running speed optimization strategy calculated based on motion status information;

(4) Each connected autonomous driving vehicle regulates its own speed according to the optimal acceleration and deceleration control strategy to achieve consistent coordinated control of the speed of the connected autonomous driving vehicle queue.

In one embodiment, the above-mentioned motion status information includes running speed, acceleration and deceleration, relative position, and distance between two adjacent connected autonomous driving vehicles.

In one embodiment, the expression of the above-mentioned intelligent driver model integrating the connected vehicle driving strategy is:

In the formula, v _i represents the running speed of the connected autonomous vehicle CAV _i , Li represents the distance between the connected autonomous vehicle i and the preceding vehicle CAV _i-1 , _{and Δv i represents} the distance between CAV _i and the preceding vehicle CAV _i-1. The speed difference between , i represents the i-th CAV, represents the first derivative of running speed v _i at time t, a _max represents the maximum acceleration of CAV,/> represents the expected speed of CAV _i , δ represents the free acceleration index, L ^* represents the expected distance between two adjacent CAVs, /> represents the minimum safe distance between two adjacent connected autonomous vehicles CAV, t _i represents the decision-making time of CAV _i , -a _min represents the maximum deceleration of CAV, a _i represents the acceleration of CAV _i , -a _i represents CAV _i deceleration.

In one embodiment, when a distributed collaborative control strategy is adopted, each two adjacent CAVs are taken as a unit, and the preceding vehicle in each control unit is used as the control center of the control unit; when a global collaborative control strategy is adopted, The leading car of the entire queue serves as the control center.

As shown in Figures 2 and 5, when a distributed collaborative control strategy is adopted, the above control centers change sequentially during the queue operation. For example, when CAV ₁ and CAV ₂ are used as control units for consistency control, the control center at this time The center is CAV _1. When CAV ₁ and CAV ₂ have the same speed and CAV ₂ and CAV ₃ are used as control units for speed consistency control, the control center is CAV ₂ at this time. When using a global collaborative control strategy, the control center is always the leader of the queue.

In one embodiment, when a distributed collaborative control strategy is adopted, for a queue consisting of n CAVs, model predictive control MPC is used to coordinate and control the operation of vehicles in the queue. At each sampling time point t _k , the control center will The optimal acceleration and deceleration control strategy is sent to adjacent networked autonomous vehicles for collaborative control.

In one embodiment, when a global collaborative control strategy is adopted, for a queue consisting of n CAVs, model predictive control MPC is used to coordinate and control the operation of vehicles in the queue. At each sampling time point t _k , the control center will The optimal acceleration and deceleration control strategy is sent to each connected autonomous vehicle for collaborative control.

In one embodiment, the flow chart is shown in Figure 3. The above-mentioned distributed collaborative control strategy includes:

Starting from the head vehicle of the connected driving vehicle queue, the two adjacent connected autonomous driving vehicles CAV _n-1 and CAV _n are controlled in sequence. The control center is CAV _n-1 , and the average speed of the current control unit is calculated. and the common expected speed v _d,n , the expressions are respectively:

In the formula, n-1 is the number of the n-1th CAV, v _i-in represents the initial speed of CAV _i , t _k represents the sampling time point, Δt represents the prediction period, a _best is the control input, indicating CAV _n-1 and the optimal acceleration and deceleration of CAV _n , M represents the number of prediction cycles, CAV ₁ -CAV _n-1 has achieved speed consistency coordinated control within M prediction cycles, m represents the prediction cycle number;

During the prediction period [t _k+ MΔt,t _k +(M+1)Δt], when When , the running speed of CAV _n is optimized as/> when When , the running speed of CAV _n is optimized to v _d,n .

Specifically, a queue consisting of three connected autonomous vehicles is taken as an example to illustrate the specific implementation process of the above distributed collaborative control strategy, which includes the following steps:

(a) Taking the connected vehicle CAV ₁ at the front of the queue and the adjacent connected vehicle CAV ₂ as the speed consistency control unit, the control center is CAV ₁ at this time, and the average speed of the current control unit is calculated through the control center and the common expected speed v _d,n , the expressions are respectively:

In the formula, refers to the average value of v ₁ and v ₂ , v _d,2 refers to the common expected speed of CAV ₁ and CAV ₂ , v _i-in refers to the initial speed of CAV _i , a _best is the control input, refers to CAV ₁ and Optimal acceleration and deceleration of CAV ₂ .

if , which means that before the sampling time t _k , the common expected speed of CAV ₁ and CAV ₂ cannot provide further speed consistency optimization space for each vehicle. The expected speed and the acceleration (deceleration) of each vehicle are coordinated based on the average speed. Control, during [t _k ,t _k +Δt], the running speed of CAV ₂ will be along/> optimize;

if It means that after the sampling time t _k , the common expected speed of CAV ₁ and CAV ₂ can provide further speed consistency optimization space for each vehicle, and collaboratively control the running speed and acceleration (deceleration) of each vehicle according to the expected speed. , during [t _k ,t _k +Δt], the operating speed of CAV ₂ will be optimized along v _d,2 .

When CAV ₁ and CAV ₂ reach speed consistency after m prediction periods, at time t _k +mΔt, the speed of CAV ₂ and CAV ₃ will start to be sampled, and speed consistency optimization will be implemented for the first 3 CAVs.

(b), CAV ₂ and CAV ₃ speed consistency control, at this time the control center is CAV ₂ , calculate the average speed of the current control unit and the common expected speed v _d,n , the expressions are respectively:

In the formula, refers to the average value of v ₂ and v ₃ , v _d,3 refers to the common expected speed of CAV ₂ and CAV ₃ , v _i-in refers to the initial speed of CAV _i , a _best is the control input, refers to CAV ₂ and Optimal acceleration and deceleration of CAV ₃

if During [t _k+ mΔt,t _k +(m+1)Δt], the running speed of CAV ₃ will be along the optimization;

if During [t _k+ mΔt,t _k +(m+1)Δt], the operating speed of CAV ₃ will be optimized along v _d,3 .

Through the above steps, a queue composed of four connected vehicles CAV will achieve speed consistency control under the distributed collaborative control strategy.

In one embodiment, the flow chart is shown in Figure 4. The above-mentioned global collaborative control strategy includes:

Taking a queue of n CAVs as the control unit, the average speed of the current control unit is calculated through the control center and the common expected speed v _d,n , the expressions are respectively:

In the formula, a _best is the control input, which represents the optimal acceleration and deceleration of all CAVs;

During the prediction period [t _k ,t _k +Δt], when When , the running speed of CAV _n is optimized as/> When/> When , the running speed of CAV _n is optimized to v _d,n . After repeated prediction cycles, the queue composed of n CAVs will achieve speed consistency under the action of the global cooperative control strategy.

In order to further clearly explain the effect of the coordinated control method of the running speed of the networked autonomous vehicle queue in this embodiment, the following specific examples are used for simulation display. Let the maximum expected speed is 100km/h, the maximum acceleration (deceleration) speed ±a _i is 3m/s ² , the expected distance L ^* between two adjacent CAVs is 30m, the decision time t _i of CAV _i is set to a constant 0.6s, and free acceleration The index δ is set to 0.8, the minimum safe distance between two adjacent CAVs/> Set to 20m. The scene during the simulation process is set to a ring highway section with obvious stop-and-go phenomena. Only a single lane is selected, the saturation is 0.8, the lane width is 3.75m, and the lane length is not affected by certain time conditions (400s). limit. The simulation process will use CVDS-IDM as the basis for operation and control, and use distributed and global speed consistency collaborative control to simulate the CAV queue operation control process in the connected autonomous driving environment.

Figure 6 shows the CAV queue speed consistency collaborative control effect in a distributed network-connected autonomous driving environment, and Figure 7 shows the CAV queue speed consistency collaborative control effect in a global network-connected autonomous driving environment. As shown in Figure 6-7, both collaborative control strategies achieve speed-consistent collaborative control of the networked autonomous driving vehicle queue within a certain period of time. During the optimization process, the speed change trend in the simulation process using the two collaborative control strategies is roughly the same. In the initial process, the oscillation is larger, the distributed oscillation time range is 0-100s, the global oscillation time range is 0-13s, and the middle The process oscillation is small, distributed for 100-350s, and global for 13-79s. Finally, all CAVs achieve speed consistency control in the connected autonomous driving environment. In addition, there is an obvious time difference in the optimization process of the two control strategies. Under the same experimental conditions, the distributed optimization time is longer than the global time. During the distributed simulation process, the CAV queue speed reached consistency after about 370s, while using the global method only took about 80s to achieve consistent control of the CAV queue speed. The results show that both methods can achieve CAV queue speed consistency control under networked autonomous driving conditions. Under the same operating conditions, the global speed consistency collaborative control method can achieve faster speed consistency control than the distributed speed consistency collaborative control method. CAV queue speed consistency cooperative control.

Claims (7)

1. A coordinated control method for speed consistency of networked autonomous driving vehicle queues, which is characterized by: Use an intelligent driver model that integrates connected vehicle driving strategies to establish a connected operating environment for autonomous vehicles; In the networked operating environment, each networked autonomous vehicle obtains its own motion status information and sends the motion status information to the control center; The control center calculates the optimal acceleration and deceleration control strategy for each connected autonomous vehicle based on the received motion status information, and returns the optimal acceleration and deceleration control strategy to each connected autonomous vehicle; Each connected autonomous driving vehicle regulates its own speed according to the optimal acceleration and deceleration control strategy to achieve consistent coordinated control of the speed of the connected autonomous driving vehicle queue; The expression of the intelligent driver model of the comprehensive connected vehicle driving strategy is: In the formula, v _i represents the running speed of the connected autonomous vehicle CAV _i , Li represents the distance between CAV _i and the preceding vehicle CAV _i-1 , _{and Δv i represents} the speed difference between CAV _i and the preceding vehicle CAV _i-1. , i represents the i-th CAV, represents the first derivative of running speed v _i at time t, a _max represents the maximum acceleration of CAV,/> represents the expected speed of CAV _i , δ represents the free acceleration index, L ^* represents the expected distance between two adjacent CAVs, /> represents the minimum safe distance between two adjacent connected autonomous vehicles CAV, t _i represents the decision-making time of CAV _i , -a _min represents the maximum deceleration of CAV, a _i represents the acceleration of CAV _i , -a _i represents CAV _i deceleration. 2. The method for collaborative control of the running speed consistency of the network-connected autonomous vehicle queue according to claim 1, characterized in that the motion state information includes operating speed, acceleration and deceleration, relative position and two adjacent network-connected autonomous vehicles. the distance between. 3. The network-connected autonomous driving vehicle queue running speed consistency collaborative control method according to claim 1, characterized in that when a distributed collaborative control strategy is adopted, every two adjacent CAVs are taken as a unit, and each control unit The lead vehicle in the queue will be used as the control center of the control unit; when the global collaborative control strategy is adopted, the leading vehicle in the entire queue will be used as the control center. 4. The network-connected autonomous driving vehicle queue running speed consistency collaborative control method according to claim 3, characterized in that when a distributed collaborative control strategy is adopted, for a queue composed of n CAVs, model predictive control MPC is used The vehicle operation in the queue is coordinated and controlled. At each sampling time point t _k , the control center sends the optimal acceleration and deceleration control strategy to adjacent connected autonomous driving vehicles for collaborative control. 5. The network-connected autonomous driving vehicle queue running speed consistency collaborative control method according to claim 3, characterized in that when a global collaborative control strategy is adopted, for a queue composed of n CAVs, model predictive control MPC is used The vehicle operation in the queue is coordinated and controlled. At each sampling time point t _k , the control center sends the optimal acceleration and deceleration control strategy to each connected autonomous driving vehicle for collaborative control. 6. The network-connected autonomous driving vehicle queue running speed consistency collaborative control method according to claim 4, characterized in that the distributed collaborative control strategy includes: Starting from the head vehicle of the connected driving vehicle queue, the two adjacent connected autonomous driving vehicles CAV _n-1 and CAV _n are controlled in sequence. The control center is CAV _n-1 , and the average speed of the current control unit is calculated. and the common expected speed v _d,n , the expressions are respectively: In the formula, n-1 represents the number of the n-1th CAV, v _i-in represents the initial speed of CAV _i , t _k represents the sampling time point, Δt represents the prediction period, a _best is the control input, representing CAV _n-1 and the optimal acceleration and deceleration of CAV _n , M represents the number of prediction cycles, CAV ₁ -CAV _n-1 has achieved speed consistency coordinated control within M prediction cycles, m represents the prediction cycle number; During the prediction period [t _k+ MΔt,t _k +(M+1)Δt], when When , the running speed of CAV _n is optimized as/> when When , the running speed of CAV _n is optimized to v _d,n . 7. The network-connected autonomous driving vehicle queue running speed consistency collaborative control method according to claim 5, characterized in that the global collaborative control strategy includes: Taking a queue of n CAVs as the control unit, the average speed of the current control unit is calculated through the control center and the common expected speed v _d,n , the expressions are respectively: In the formula, a _best is the control input, which represents the optimal acceleration and deceleration of all CAVs; During the prediction period [t _k ,t _k +Δt], when When , the running speed of CAV _n is optimized as/> When/> When , the running speed of CAV _n is optimized to v _d,n .