Queue predictive cruise and lane change control system and method based on cloud control platform

Abstract

The present application relates to lane changing decision technical field, especially relates to a kind of queue predictive cruise and lane changing control system and method based on cloud control platform, wherein, system includes: the cloud control platform being equipped with queue predictive cruise and lane changing decision module, for processing the real-time state of target queue vehicle, parameter information and dynamic traffic environment information on road, to solve the optimal longitudinal acceleration and optimal lateral lane changing decision of target queue vehicle;Vehicle-mounted platform being equipped with trajectory planning algorithm module and queue distributed controller module, for based on optimal longitudinal acceleration and optimal lateral lane changing decision, utilize trajectory planning algorithm module to carry out driving trajectory planning to lead car in target queue vehicle, obtain trajectory planning route, utilize queue distributed controller module to carry out tracking control to target queue vehicle.The system provides wide-area dynamic traffic information for target vehicle, so that vehicle more predictively adjusts driving strategy.

Description

Technical Field

[0001] This invention relates to the field of lane change decision-making technology, and in particular to a queue predictive cruise and lane change control system and method based on a cloud control platform. Background Technology

[0002] Vehicle platooning technology allows vehicles to adjust their longitudinal movement based on the status of adjacent vehicles, achieving uniform speeds and desired spacing within the platoon. Because all vehicles in the platoon are in similar states, this improves traffic efficiency, fuel economy, and driving safety.

[0003] Predictive cruise control is a method that uses predictive information (such as forward information about road gradient, the phase and timing of upcoming traffic lights, and the motion and changing trends of surrounding vehicles) to construct an optimal control problem and determine the optimal driving strategy. For predictive cruise control in highway scenarios, it primarily utilizes forward information about the road gradient ahead to plan the vehicle's longitudinal speed, thereby improving fuel economy.

[0004] Lane changing is a fundamental driving behavior. Compared to cruise control, lane changing is more complex and demands higher levels of decision-making, planning, and vehicle control. Current lane-changing decision-making methods can be categorized into four types: rule-based methods, utility function-based methods, machine learning-based methods, and game theory-based methods. These methods primarily suffer from the following drawbacks:

[0005] (1) Existing platoon lane-changing technology mainly focuses on longitudinal platoon movement, while platoon lane-changing technology is relatively lacking. Lane-changing capability is very important in actual platoon movement because there are scenarios such as overtaking and entering / exiting ramps. If the platoon does not have lane-changing capability, it will force the platoon to disband, which will not only require additional personnel to take over, but also disrupt the continuity of platoon operation.

[0006] (2) Existing predictive cruise control technologies only focus on optimizing the longitudinal speed of the target vehicle in the current lane, ignoring lane-changing behavior. This reduces the searchable dimensions of the optimization problem, narrows the range of longitudinal speed choices, and compromises the optimality of the optimization problem. Furthermore, most current predictive cruise control technologies are based on distributed control using vehicle-to-infrastructure (V2I) technology, requiring the deployment of maps and computing units on the vehicle. First, deploying maps and algorithms on the vehicle is costly and difficult to handle high-load computing tasks. Second, V2I technology has limitations in communication and sensing distance, making it difficult to provide vehicles with dynamic traffic information over longer line-of-sight distances.

[0007] (3) Existing lane-changing decision-making techniques only focus on the short-term gains of the target vehicle after changing lanes, without considering that the value judgment of lane-changing decisions is also affected by future traffic conditions. For example, even if the target vehicle can gain a faster speed or more driving space in the short term after changing lanes, it may be blocked by new slower vehicles in the target lane after a period of time. If the lane-changing decision-making model can take into account the future driving conditions of the vehicle, it will help to obtain longer-term benefits. Summary of the Invention

[0008] This invention provides a cloud-based predictive cruise and lane-changing control system and method for queuing, which addresses the limitations of existing predictive cruise control technologies based on vehicle-road cooperative technologies, such as limited communication distance and perception range, small decision-making and planning scope and low predictability, and difficulty in achieving real-time solution of control strategies.

[0009] A first aspect of this invention provides a platoon predictive cruise and lane-changing control system based on a cloud control platform, comprising: a cloud control platform, wherein the cloud control platform is equipped with a platoon predictive cruise and lane-changing decision module, used to acquire real-time status and parameter information of vehicles in a target platoon and dynamic traffic environment information, and to process the real-time status, parameter information and dynamic traffic environment information using the platoon predictive cruise and lane-changing decision module to solve for the optimal longitudinal acceleration and optimal lateral lane-changing decision of the vehicles in the target platoon, and send the decision to the vehicles in the target platoon, wherein the vehicles in the target platoon include a lead vehicle and multiple tracking vehicles; and an on-board platform, wherein the on-board platform is mounted on the vehicles in the target platoon, the on-board platform of the lead vehicle is equipped with a trajectory planning algorithm module, and each on-board platform of the vehicles in the target platoon is equipped with a platoon distributed controller module, used to plan the driving trajectory of the lead vehicle based on the optimal longitudinal acceleration and the optimal lateral lane-changing decision using the trajectory planning algorithm module to obtain a trajectory planning route, and to track and control the vehicles in the target platoon using the platoon distributed controller module, and to upload the real-time status of the vehicles in the target platoon to the cloud control platform, forming a vehicle-cloud rolling closed-loop control.

[0010] Optionally, the cloud control platform obtains the real-time status and vehicle parameter information of the target queue vehicles through the edge cloud.

[0011] Optionally, it also includes:

[0012] Roadside infrastructure, which is connected to the cloud control platform, is used to collect the dynamic traffic environment information and send the dynamic traffic environment information to the cloud control platform;

[0013] The vehicle-mounted intelligent remote information processing terminal T-BOX is mounted on the vehicles in the target convoy and is used to enable the vehicle-mounted platform to communicate and process information with the cloud control platform.

[0014] Optionally, the queue predictive cruise and lane-changing decision module includes: a dynamic sensing domain unit, a prediction unit, a cost function and constraint unit, and an optimization solution unit, wherein,

[0015] The dynamic perception domain unit is used to filter multiple environmental vehicles around the target queue of vehicles based on the dynamic traffic environment information and the real-time status, and to obtain the environmental vehicles that affect the driving of the target queue of vehicles within a preset prediction time domain.

[0016] The prediction unit is used to establish the state transition equations of the target queue vehicles and the environmental vehicles in the discrete space, so as to estimate the future states of the target queue vehicles and the environmental vehicles.

[0017] The cost function and constraint unit are used to establish a queue predictive cruise and lane-changing decision cost function based on multiple preset optimization objectives and the future state, and to constrain the target queue vehicles.

[0018] The optimization solution unit is used to classify the cost function of the platoon predictive cruise and lane change decision into multiple nonlinear programming subproblems to solve for the optimal longitudinal acceleration and optimal lateral lane change decision, and then send them to the target platoon vehicles.

[0019] Optionally, each vehicle in the target convoy is equipped with a model predictive controller and a linear quadratic regulator. The target convoy vehicles transmit their own position, speed, and acceleration to the following vehicles and use the model predictive controller to calculate longitudinal control commands. The lead vehicle sends the trajectory planning route to each following vehicle and uses the linear quadratic regulator to calculate lateral control commands.

[0020] Optionally, the trajectory planning algorithm module includes: a cloud instruction parsing unit, a lane change safety judgment unit, a safety degradation strategy unit, and a trajectory generation unit, wherein,

[0021] The cloud-based instruction parsing unit is used to parse the optimal longitudinal acceleration and the optimal lateral lane change decision to obtain the recommended acceleration and recommended lateral lane change decision that the lead vehicle should drive.

[0022] The lane change safety judgment unit is used to analyze the drivable safe area of ​​the lead vehicle through the recommended acceleration and the recommended lateral lane change decision, so as to determine whether the lane change meets the preset safety conditions.

[0023] The safety degradation strategy unit is used to obtain the information of the vehicle in front of the lead vehicle when the preset safety conditions are not met, and output the safety acceleration command of the lead vehicle with the goal of maintaining a safe headway.

[0024] The trajectory generation unit is used to plan the driving trajectory of the navigator based on the uniform acceleration model and the fifth-order polynomial curve, the recommended acceleration, the recommended lateral lane change decision, the lane change safety judgment result, and the safety acceleration command, so as to obtain the planned trajectory route.

[0025] Optionally, the queue distributed controller module includes: a queue vertical controller unit and a queue horizontal controller unit, wherein,

[0026] The queue longitudinal controller unit is used to acquire the position, speed and acceleration of the vehicle in front and the state of the following vehicles in the target queue, so as to calculate the expected acceleration of each following vehicle. Each following vehicle maintains the distance and tracks the speed of the vehicle in front according to its corresponding expected acceleration.

[0027] The queue lateral controller unit is used to calculate the front wheel angle of each following vehicle and perform path tracking based on the trajectory planning route and the state of each following vehicle.

[0028] A second aspect of the present invention provides a queuing predictive cruise and lane-changing control method based on a cloud control platform, comprising the following steps:

[0029] The real-time status and vehicle parameter information of the target queue vehicles are obtained through the edge cloud in the cloud control platform, and dynamic traffic environment information is collected through roadside infrastructure.

[0030] The queuing predictive cruise and lane-changing decision module in the cloud control platform processes the real-time status, parameter information, and dynamic traffic environment information to solve for the optimal longitudinal acceleration and optimal lateral lane-changing decision of the target queuing vehicles, and sends the decision to the target queuing vehicles.

[0031] The optimal longitudinal acceleration and the optimal lateral lane-changing decision are received by the onboard platform of the vehicles in the target platoon.

[0032] The trajectory planning algorithm module in the vehicle platform is used to plan the driving trajectory of the target convoy of vehicles to obtain the trajectory planning route.

[0033] The vehicle-mounted platform uses a queue distributed controller module to track and control the vehicles in the target queue, and uploads the real-time status of the vehicles in the target queue to the cloud control platform, forming a vehicle-cloud rolling closed-loop control.

[0034] A third aspect of the present invention provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the cloud-based predictive cruise and lane-changing control method as described in the above embodiments.

[0035] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described queue predictive cruise and lane-changing control method based on a cloud control platform.

[0036] The cloud-based predictive cruise and lane-changing control system and method based on the present invention have the following beneficial effects:

[0037] Compared to vehicle-road cooperative technology or single-vehicle intelligent technology, cloud control platforms can overcome the limitations of communication distance and perception range, improve the scope of decision-making and planning, enable vehicles to adjust driving strategies more predictively, and reduce the vehicle's computing burden, achieving efficient utilization of computing resources.

[0038] Compared to existing predictive cruise control, which can expand the search dimensions, the cloud control platform considers both the platoon's cruise and lane-changing strategies, improving the optimality of the optimization results. At the same time, by utilizing predictive information, the lane-changing strategy can take into account the long-term expected benefits of the platoon's operation, which helps the platoon obtain more long-term returns.

[0039] The trajectory planning algorithm and distributed queue controller, which work in conjunction with cloud-based algorithms, supplement the execution process of lane changing and cruising in the queue. The trajectory planning algorithm considers the safe driving space of the queue, determines the safe distance model by analyzing the vehicle braking process, and designs a safety degradation strategy to ensure the safety of the queue driving. The distributed queue controller designs queue longitudinal controller and queue lateral controller module controllers according to the longitudinal and lateral control objectives to ensure the tracking performance of the queue vehicles.

[0040] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0041] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0042] Figure 1 This is a schematic diagram of a cloud-based predictive cruise and lane-changing control system for queuing, provided by an embodiment of the present invention.

[0043] Figure 2This is a vehicle-cloud layered architecture for a cloud-based queuing predictive cruise and lane-change control system according to an embodiment of the present invention.

[0044] Figure 3 This is a schematic diagram of a target platoon of vehicles driving according to an embodiment of the present invention;

[0045] Figure 4 This is a schematic diagram of the forward sensing area provided according to an embodiment of the present invention;

[0046] Figure 5 A schematic diagram of the rear sensing area provided according to an embodiment of the present invention;

[0047] Figure 6 This is a flowchart illustrating the trajectory planning algorithm module provided in an embodiment of the present invention.

[0048] Figure 7 This is a schematic diagram illustrating the safe distance when vehicles in a target platoon change lanes, according to an embodiment of the present invention.

[0049] Figure 8 A schematic diagram of the framework of a queue longitudinal controller unit provided according to an embodiment of the present invention;

[0050] Figure 9 A schematic diagram of the framework of a queue lateral controller unit provided according to an embodiment of the present invention;

[0051] Figure 10 This is a flowchart of a cloud-based predictive cruise and lane-changing control method for queuing provided in an embodiment of the present invention.

[0052] Figure 11 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of the present invention. Detailed Implementation

[0053] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0054] Figure 1 This is a schematic diagram of the queuing predictive cruise and lane-changing control system based on a cloud control platform according to an embodiment of the present invention.

[0055] like Figure 1 As shown, the cloud-based predictive cruise and lane-changing control system 100 includes a cloud control platform 1 and an on-board platform 2.

[0056] The cloud control platform 1 is equipped with a queue predictive cruise and lane change decision module, which is used to acquire real-time status and parameter information of vehicles in the target queue as well as dynamic traffic environment information. The queue predictive cruise and lane change decision module processes the real-time status, parameter information and dynamic traffic environment information to solve the optimal longitudinal acceleration and optimal lateral lane change decision of vehicles in the target queue, and sends them to the vehicles in the target queue. The vehicles in the target queue include a lead vehicle and multiple follower vehicles.

[0057] The vehicle platform 2 is mounted on the target platoon of vehicles. The lead vehicle's vehicle platform is equipped with a trajectory planning algorithm module, and each vehicle platform in the target platoon of vehicles is equipped with a platoon distributed controller module. Based on the optimal longitudinal acceleration and optimal lateral lane change decision, the trajectory planning algorithm module is used to plan the driving trajectory of the lead vehicle to obtain the trajectory planning route. The platoon distributed controller module is used to track and control the target platoon of vehicles and upload the real-time status of the target platoon of vehicles to the cloud control platform to form a vehicle-cloud rolling closed-loop control.

[0058] In some embodiments, the present invention further includes:

[0059] Roadside infrastructure 300 is connected to cloud control platform 1 to collect dynamic traffic environment information and send the dynamic traffic environment information to cloud control platform 1;

[0060] The vehicle-mounted intelligent remote information processing terminal T-BOX4 is mounted on the vehicles in the target convoy and is used to enable the vehicle-mounted platform 2 to communicate and process information with the cloud control platform 1.

[0061] Furthermore, in one embodiment of the present invention, the queue predictive cruise and lane-changing decision module 11 includes: a dynamic sensing domain unit 111, a prediction unit 112, a cost function and constraint unit 113, and an optimization solution unit 114, wherein,

[0062] The dynamic perception domain unit 111 is used to filter multiple environmental vehicles around the target queue of vehicles based on dynamic traffic environment information and real-time status, and to obtain the environmental vehicles that affect the driving of the target queue of vehicles within a preset prediction time domain.

[0063] Prediction unit 112 is used to establish the state transition equations of the target queue vehicles and the environment vehicles in the discrete space, so as to estimate the future states of the target queue vehicles and the environment vehicles.

[0064] The cost function and constraint unit 113 is used to establish a queue predictive cruise and lane change decision cost function based on multiple preset optimization objectives and future states, and to constrain the vehicles in the target queue.

[0065] The optimization solution unit 114 is used to classify the cost function of platoon predictive cruise and lane change decision into multiple nonlinear programming subproblems to solve for the optimal longitudinal acceleration and optimal lateral lane change decision, and send them to the target platoon vehicles.

[0066] It should be noted that, as a real-time cloud-controlled application, queuing predictive cruise and lane change decision-making need to be deployed in a cloud-controlled application platform and provide access interfaces to intelligent connected vehicle queuings of various models.

[0067] Specifically, such as Figure 1 and 2 As shown, the cloud control platform 1 deploys a queue predictive cruise and lane change decision module 11, which includes four units: dynamic perception domain 111, prediction model 112, cost function and constraints 113, and optimization solution 114.

[0068] The cloud control platform 1 can obtain real-time status and vehicle parameter information of the target queue vehicles in advance through the edge cloud, such as vehicle status (position, speed, acceleration and gear) and vehicle model. This allows the cloud control platform 1 to store vehicle parameters such as fuel consumption model and engine model, which can be queried by vehicle model. It can also receive dynamic traffic environment information (including vehicle position, speed and other data in the surrounding environment of the queue) collected by the roadside infrastructure 3. The acquisition of the above information requires the support of relevant resource platforms, such as maps and location services.

[0069] Then, the dynamic perception domain unit 111 in the queue predictive cruise and lane change decision module 11 filters multiple environmental vehicles around the target queue based on dynamic traffic environment information and the real-time status of the vehicles in the queue, and finds environmental vehicles that may affect the queue's movement in the prediction time domain. It should be noted that considering too many environmental vehicles will increase the amount of computation and waste resources, while considering too few environmental vehicles will lead to unreliable calculation results and loss of use value. Therefore, the dynamic perception domain unit 111 filters multiple environmental vehicles around the queue to achieve a balance between computational resources and the reliability of calculation results.

[0070] Based on the screening results of the dynamic sensing domain unit 111, the prediction unit 112 models the environmental vehicles in the discrete space, establishes the state transition equation of the vehicles, and uses the historical information of the target queue vehicles and the future control input to estimate the future state of the target queue vehicles and the environmental vehicles, providing prior information for the establishment of subsequent optimization problems.

[0071] The cost function and constraint unit 113 take into account multiple optimization objectives of the system (including safety, efficiency and economy, etc.) and integrate them into a comprehensive performance index. The penalty obtained by the target queue vehicles for taking different behaviors is reflected in the value of the cost function. At the same time, considering factors such as traffic rules, vehicle actuator limitations and driving comfort, a queue predictive cruise and lane change decision cost function is established to constrain the system.

[0072] The optimization solution unit 114 is designed with a hierarchical solution mechanism, which divides the mixed integer nonlinear programming problem, namely the queue predictive cruise and lane change decision cost function, into multiple nonlinear programming subproblems, and solves the optimal longitudinal acceleration sequence and lateral lane change decision sequence. The decision sequence calculated by this unit will be sent to the queue on-board platform 2 to guide the vehicle to execute longitudinal acceleration and lateral lane change decisions.

[0073] Furthermore, in one embodiment of the present invention, each vehicle in the target convoy is equipped with a model predictive controller and a linear quadratic regulator. The target convoy vehicles transmit their own position, speed, and acceleration to the following vehicles and use the model predictive controller to calculate longitudinal control commands. The lead vehicle sends the trajectory planning route to each following vehicle and uses the linear quadratic regulator to calculate lateral control commands.

[0074] Furthermore, in one embodiment of the present invention, the trajectory planning algorithm module includes: a cloud instruction parsing unit, a lane change safety judgment unit, a safety degradation strategy unit, and a trajectory generation unit, wherein,

[0075] The cloud-based instruction parsing unit is used to analyze the optimal longitudinal acceleration and optimal lateral lane change decision to obtain the recommended acceleration and recommended lateral lane change decision that the lead vehicle should take.

[0076] The lane change safety judgment unit is used to analyze the drivable safe area of ​​the lead vehicle by recommending acceleration and recommending lateral lane change decision to determine whether the lane change meets the preset safety conditions.

[0077] The safety degradation strategy unit is used to obtain information about the vehicle ahead of the lead vehicle when the preset safety conditions are not met, and outputs a safety acceleration command for the lead vehicle with the goal of maintaining a safe headway.

[0078] The trajectory generation unit is used to plan the driving trajectory of the navigator based on the uniform acceleration model and the fifth-order polynomial curve, the recommended acceleration, the recommended lateral lane change decision, the lane change safety judgment result, and the safety acceleration command, so as to obtain the trajectory planning route.

[0079] Furthermore, in one embodiment of the present invention, the queue distributed controller module includes: a queue vertical controller unit and a queue horizontal controller unit, wherein,

[0080] The queue longitudinal controller unit is used to acquire the position, speed and acceleration of the vehicle in front and the state of the following vehicles in the target queue, so as to calculate the expected acceleration of each following vehicle. Each following vehicle maintains distance and tracks speed from the vehicle in front based on its corresponding expected acceleration.

[0081] The queue lateral controller unit is used to plan the route based on the trajectory and the state of each following vehicle to calculate the front wheel angle of each following vehicle. Each following vehicle tracks the lead vehicle according to its corresponding front wheel angle.

[0082] Specifically, such as Figure 1 and 2 As shown, the vehicle platform 2 is equipped with a trajectory planning algorithm module 21 and a queue distributed controller module 22 to supplement the execution process of cruise and lane changing. The trajectory planning algorithm module 21 includes four units: a cloud instruction parsing unit 211, a lane changing safety judgment unit 212, a safety degradation strategy unit 213, and a trajectory generation unit 214. The queue distributed controller module 22 includes two units: a queue longitudinal controller unit 221 and a queue lateral controller unit 222.

[0083] The cloud-based command parsing unit 211 is used to parse the longitudinal acceleration sequence and lateral lane change decision sequence commands sent from the cloud control platform 1 to the vehicle. The longitudinal acceleration sequence and lateral lane change decision sequence sent from the cloud control platform 1 contain time information. Based on the current system time and the timestamp in the sequence, this unit indexes and determines the current real-time recommended acceleration and lateral lane change decision of the lead vehicle.

[0084] The lane-change safety assessment unit 212 is used to determine the safe driving area of ​​the lead vehicle and assess the safety of lane-change driving in the convoy. This unit analyzes the braking process of the lead vehicle, combines the status of the preceding vehicle, the status of the preceding vehicle in the target lane, and the status of the following vehicle in the target lane, calculates the minimum safe distance for lane-change, and outputs the lane-change safety assessment result.

[0085] The safety degradation strategy unit 213 is used to provide a safe acceleration command to ensure the vehicle in the convoy follows the vehicle in front, based on the information of the vehicle in front obtained from the on-board sensors, when the lane change safety judgment unit determines that the lane change is unsafe. This unit is based on an intelligent driver model and uses the state information of the vehicle in front to output the acceleration of the vehicle to ensure safe following, with the goal of maintaining a safe headway, thus ensuring the safety performance of the target vehicle in real traffic conditions.

[0086] The trajectory generation unit 214 is used to generate the driving trajectory of the target platoon vehicles under cruise and lane-changing conditions. Based on the longitudinal acceleration sequence and lateral lane-changing decision sequence issued by the cloud control platform 1, the output results of the lane-changing safety judgment unit 212 and the safety degradation strategy operator 213, this unit establishes the driving trajectory, i.e., the trajectory planning route, based on the uniform acceleration model and the fifth-order polynomial curve.

[0087] The queuing longitudinal controller unit 221 is used to calculate the desired acceleration of vehicles in the queuing, ensuring that the vehicles in the target queuing maintain the desired geometric configuration while keeping pace with the speed of the vehicle in front. This unit establishes a longitudinal linear error model for the vehicles based on a third-order longitudinal dynamics model and feedback linearization theory. Using a distributed model predictive control algorithm, with control accuracy and control stability as control objectives, longitudinal control commands are calculated through quadratic programming.

[0088] The lateral controller unit 222 calculates the front wheel steering angles of vehicles in the queue, ensuring that the vehicles in the queue can follow the planned path. This unit establishes a lateral error model based on a two-degree-of-freedom lateral dynamics model, the geometric relationship between vehicle positions and the projection points of the reference path. A linear quadratic regulator is used to calculate the system's feedback control law, and a feedforward controller is added to eliminate the system's steady-state error.

[0089] Based on the foregoing, the specific working principle of the cloud-based predictive cruise and lane-changing control system for queuing proposed in this embodiment of the invention is as follows: Vehicles in the target queuing request predictive cruise and lane-changing control services from the cloud control platform 1. The cloud control platform 1 obtains the real-time status and vehicle parameter information of the target queuing vehicles through its internal edge cloud and communication network, and collects dynamic traffic environment information through roadside infrastructure 3. The predictive cruise and lane-changing decision module deployed on the cloud control platform 1 considers the real-time status of the target queuing vehicles, dynamic traffic environment information, and vehicle models, solves for the longitudinal acceleration and lateral lane-changing decisions of the queuing, and sends these decisions to the lead vehicle of the target queuing vehicles through the onboard intelligent remote information processing terminal T-BOX4. After receiving the decision from the cloud control platform 1, the lead vehicle uses the trajectory planning algorithm module in its onboard platform to plan the driving trajectory, obtain the planned route, and synchronize it with all following vehicles. Then, the queue distributed controller module performs tracking control and uploads the real-time vehicle status, forming a vehicle-cloud rolling closed-loop control.

[0090] The following specific embodiment further illustrates the queuing predictive cruise and lane-changing control system based on the cloud control platform proposed in this invention.

[0091] First, such as Figure 3As shown, trucks represent vehicles in the target queue, cars represent vehicles in the environment, (m,n) represent vehicle numbers, L represents the left lane, R represents the right lane, E represents vehicles in the queue, m∈{L,R,E}, n∈{1,2,...}.

[0092] The target queuing of vehicles is modeled in discrete space, and the following variables are defined: x m,n (k) represents the longitudinal position of the vehicle (m,n) at step k, v m,n (k) represents the longitudinal velocity of the vehicle (m,n) at step k, δ m,n (k) represents the lane of vehicle (m,n) at step k, δ m,n (k)∈{0,1}, 0 represents the right lane, 1 represents the left lane, X m,n (k) represents the state variables of vehicle (m,n) at step k, including longitudinal position, longitudinal velocity, and lane location, X m,n (k)={x m,n (k),v m,n (k),δ m,n (k)} T a m,n (k) represents the longitudinal acceleration of the vehicle (m,n) at step k, λ m,n (k) indicates that vehicle (m,n) changes lanes to the left at step k, λ m,n (k)∈{0,1}, 0 indicates no lane change, 1 indicates left lane change, γ m,n (k) indicates that vehicle (m,n) changes lanes to the right at step k, γ m,n (k)∈{0,1}, 0 indicates no lane change, 1 indicates right lane change, U m,n (k) represents the control variables of vehicle (m,n) at step k, including longitudinal acceleration, left lane change, and right lane change, U m,n (k)={a m,n (k),λ m,n (k),γ m,n (k)} T .

[0093] The dynamic perception domain unit 111 comprises two parts: a forward perception region and a rear perception region. For the region in front of the target convoy, environmental vehicles within this region may affect the convoy's movement and lane changes. For example... Figure 4 As shown, at time t0, there is an environmental vehicle (CPV) in front of the target convoy, with an initial distance of x between them. p At time t0+Δt, the target convoy of vehicles travels to a position x behind the environmental vehicle's CPV. thWithin this range, the environmental vehicle's CPV affects the movement of the target convoy of vehicles. Assume that both the target convoy vehicles and the environmental vehicle's CPV have an initial speed v at the start of the prediction. E and v CPV Traveling at a constant speed, when v cpv When x = 0, p Take the maximum value, at which point x p This indicates the maximum distance that needs to be considered in the area ahead:

[0094] x p =v E Δt+x th (1)

[0095] For the area behind the lane adjacent to the target queue of vehicles, environmental vehicles in this area may affect lane changing within the queue. Consider the following: Figure 5 In the scenario shown, at time t0, there is an environmental vehicle (TFV) behind the lane next to the target vehicle in the queue, with an initial distance of x from it. f At time t0+Δt, the environmental vehicle TFV travels to the rear of the target convoy of vehicles by x. th Within this range, the environmental vehicle TFV affects the lane changing of the target platoon of vehicles. Assume that the target platoon vehicles and the environmental vehicle TFV each have an initial speed v at the start of the prediction. E and v TFV Travel at a constant speed, x f It can be expressed by the following formula:

[0096] x f =(v TFV -v E )·Δt+x th (2)

[0097] When the speed of the environmental vehicle TFV is equal to the road speed limit v road And when the speed of the vehicles in the target convoy is lower than the speed limit of the adjacent lane, x f Take the maximum value, representing the limit range that needs to be considered for the area behind the adjacent lane; when the speed of the target platoon of vehicles exceeds the road speed limit, only the area behind the platoon at the start of prediction (x) is considered. th Vehicles within the designated area. This can be expressed by the following formula:

[0098]

[0099] According to the car-following theory, vehicles in the area behind the current lane of the queue will not affect the queue's movement, so they do not need to be considered.

[0100] Prediction unit 112 uses the object's historical information and future control inputs to predict the object's future state. The state transition equation for vehicle (m,n) is defined as follows:

[0101] X m,n (k+1)=AX m,n (k)+BU m,n (k) (4)

[0102] in,

[0103]

[0104] In the formula, ΔT represents the distance from the walk.

[0105] Since the motion of vehicles on highways is relatively stable, a uniform velocity model is used to predict the environmental vehicle state. The control variables are shown below:

[0106] U m,n (k)=(0 0 0) T (5)

[0107] Cost function and constraint element 113: Using safety, efficiency, and economy as optimization objectives, a cost function for predictive cruise and lane-changing decisions in the queue is established:

[0108]

[0109]

[0110] in,

[0111]

[0112]

[0113]

[0114] In the formula, N p For predicting the time domain, N c To control the time domain, ω f ,ω c ,ω o ,ω v ,ω a ,ω l These are: following weight, lane-changing weight, fuel consumption weight, speed weight, acceleration weight, and lane weight. For safe headway, Let (E,n) be the actual headway of the vehicle at step k. For safe collision time, Let θ be the actual collision time of vehicle (E,n), and let θ(i) be used to determine whether the queue changes lanes at step k. f(N) represents the actual collision time between the queue and the vehicle in the target lane at step k. E,i(k),T E,i (k) represents the fuel consumption model, indicating the fuel consumption per second of the vehicle (E,n) at step k, in g / s and N. E,n (k) represents the engine speed of the vehicle (E,n) at step k, T E,n (k) represents the engine torque of the vehicle (E,n) at step k, v des Indicates the desired speed of the queue. Used to determine the lane where the queue is located, α and β are constants.

[0115] The optimization unit 114 addresses the issue that the unsolved variables in the queuing predictive cruise and lane-changing decision algorithm include both continuous acceleration metrics and 0-1 lane-changing decision variables. Furthermore, both the cost function and constraints contain nonlinear components, making this a mixed-integer nonlinear programming problem. Solving such problems is extremely difficult and requires simplification. Additionally, considering the high risk of queuing lane changes, multiple lane changes within a short period are not recommended. Therefore, it is assumed that the queuing will change lanes at most once within the control time domain. Based on this assumption, and considering that the research scenario involves two lanes, the problem can be decomposed into N... c +1 sub-problems, corresponding to no lane change, lane change in step 1, ..., lane change in step N, respectively. c The step-by-step lane change is shown in the following formula:

[0116]

[0117] For each subproblem, the only quantity to be optimized is the acceleration sequence, which can be solved using nonlinear programming, while retaining the cost and acceleration sequence of each subproblem:

[0118]

[0119] Next, compare N c Given the cost of +1 subproblems, find the subproblem with the minimum cost. The longitudinal acceleration sequence and lateral lane-changing decision sequence corresponding to this subproblem are the final optimization results.

[0120]

[0121] Finally, the longitudinal acceleration sequence and lateral lane-changing decision sequence are sent to the lead vehicle of the target platoon. The vehicles in the target platoon then analyze and track the commands from the cloud control platform 1. This process is repeated in the next time step, forming a rolling closed-loop control.

[0122] The specific design of vehicle platform 2 will be described below:

[0123] like Figure 6 As shown, the command sent from cloud control platform 1 to the vehicle is a longitudinal acceleration sequence a. *Lateral lane change decision sequence lc * The trajectory planning algorithm module 21 can divide trajectory planning into two cases based on cloud instructions, lane change safety judgment results and safety degradation strategies: lane change trajectory planning and cruise trajectory planning. The following sections will explain each unit of the trajectory planning algorithm.

[0124] The cloud-based command parsing unit 211 parses the acceleration command and lane change decision command that the vehicle should currently be driving according to the system time. In this embodiment, the cloud control platform 1 interacts with the vehicle platform 2 at a frequency of 2Hz, and the queue predictive cruise and lane change decision algorithm module 11 operates with a discrete time step of 0.5s, with the longitudinal acceleration sequence and lateral lane change decision sequence having a length of 10. Due to the high vehicle speed under highway conditions, the trajectory planning algorithm module 21 and the queue controller module 22 operate at a high frequency of 100Hz. In this embodiment, based on the system time and the timestamps contained in the acceleration sequence and lane change decision sequence, the elements are stored in a queue linear table in order from nearest to farthest. The elements in the space follow the first-in-first-out principle and are output sequentially at the tail of the queue.

[0125] Lane change safety judgment unit 212, such as Figure 7 The scenario shown uses the lead vehicle (E,n) and the environmental vehicle (CPV) in the target convoy as examples to analyze the safe distance. At time t0, the environmental vehicle (CPV) brakes suddenly, and the lead vehicle (E,n) brakes after a reaction time. At time t1, the environmental vehicle (CPV) stops, and at time t2, the lead vehicle (E,n) stops, eventually coming to a stop behind the environmental vehicle (CPV). To ensure no collision occurs, the following conditions must be met:

[0126] x CPV (t1)-x E,n (t2)≥L CPV (14)

[0127] In the formula, x CPV (t1) and x E,n (t2) represents the front bumper positions of the environmental vehicle (CPV) and the navigator vehicle (E,n) when they are stopped, respectively. CPV This indicates the length of the environmental vehicle (CPV).

[0128] For environmentally friendly vehicles (CPVs), their braking process can be considered as a uniform deceleration process, x CPV (t1) can be expressed as follows:

[0129]

[0130] In the formula, b CPV This represents the maximum deceleration of CPV.

[0131] The lead vehicle (E,n) is an intelligent connected vehicle, and its acceleration at time t0 is a. E,n (t0), velocity is v E,n (t0), due to delays in vehicle data acquisition, processing, and control systems, the navigator (E,n) cannot directly enter braking mode and requires time τ. E,n Only then can it enter emergency braking mode and travel at maximum deceleration. At time t2, the lead car (E,n) stops, and the deceleration position of the lead car (E,n) to a stop can be calculated as follows:

[0132]

[0133] From equation (14-16), the required safe distance between the lead vehicle (E,n) and the environmental vehicle (CPV) can be calculated as follows:

[0134]

[0135] Similarly, D can be calculated. TPV and D TFV The following conditions must be met:

[0136]

[0137]

[0138] Safety degradation strategy unit 213, when the cloud control platform 1 decides to change lanes and the lane change safety judgment result is unsafe, uses the intelligent driver model to calculate the safety acceleration, performs cruise trajectory planning, and waits for the cloud command of the next cycle.

[0139] The intelligent driver model, proposed by Treiber et al. in 2000, is a car-following model consisting of two parts: the acceleration trend in free-flow conditions and the deceleration trend to avoid collisions with the vehicle in front. It can describe the car-following characteristics of a vehicle under different conditions.

[0140]

[0141] In the formula, a IDM a represents the acceleration of the target vehicle. max v represents the maximum acceleration of the vehicle. e v represents the speed of the target vehicle. des Let δ represent the target vehicle's desired speed, δ represent the acceleration exponent, and Δv represent the speed difference between the target vehicle and the leading vehicle. The expression is Δv = v e -v p v p The speed of the leading vehicle is represented by s. * (v eΔv) represents the expected following distance of the target vehicle, and s represents the actual following distance between the target vehicle and the leading vehicle. The expression is s = x p -x e -l p x p and x e These represent the longitudinal positions of the leading vehicle and the target vehicle, respectively. p Indicates the length of the lead vehicle, a max (1-(v e / v des ) δ ) represents the acceleration of the target vehicle in free flow, a max (s * (v e ,Δv) / s) 2 This indicates the braking deceleration used to prevent a collision with the vehicle in front.

[0142] The expression for the expected following distance is as follows:

[0143]

[0144] In the formula, s0 represents the stationary safe distance, T represents the safe headway, and b represents the comfortable deceleration.

[0145] The trajectory generation unit 214 comprises two parts: a cruising trajectory and a lane-changing trajectory. When the vehicle is cruising, a coordinate system is established with the front bumper of the lead vehicle in the queue at the initial moment of each departure distance as the origin, the vehicle's direction of travel as the x-axis, and the target lane direction as the y-axis. A uniform acceleration model is used to generate the cruising trajectory.

[0146]

[0147] In the formula, x(t) and y(t) represent the longitudinal and lateral displacements of the vehicle, respectively; x0, v0, and t0 represent the longitudinal position, longitudinal velocity, and time of the vehicle at the start of the planning, respectively; and a represents the desired acceleration, which may be the result of cloud-based acceleration analysis or the safe acceleration calculated by the intelligent driver model.

[0148] When a vehicle changes lanes, a coordinate system is established with the front bumper of the lead vehicle in the queue at the initial moment of the lane change as the origin, the vehicle's direction of travel as the x-axis, and the direction of the target lane as the y-axis. A double fifth-order polynomial is used to describe the lane change trajectory:

[0149]

[0150] Define the starting state of the lane change as P s The lane change endpoint state is P. e :

[0151]

[0152] In the formula, x, y, These represent the vehicle's longitudinal displacement, longitudinal velocity, longitudinal acceleration, lateral displacement, lateral velocity, and lateral acceleration, respectively.

[0153] Regarding lane-changing time, this embodiment selects t. lc =5s. For the starting state of a lane change, the speed and acceleration information can be measured by the onboard sensors. For the ending state of a lane change, assuming the lateral displacement is equal to the lane width, the lateral speed and lateral acceleration are 0, and the longitudinal displacement, longitudinal speed, and longitudinal acceleration are calculated by integrating the optimal acceleration sequence sent from the cloud within discrete time.

[0154] The lane-change trajectory function contains 12 unknown parameters. Given the starting state, ending state, and time of the lane change (a total of 13 quantities), the lane-change trajectory can be uniquely determined. To solve for the coefficients of the fifth-degree polynomial, the calculation process is transformed into matrix operations. Coefficient matrices A and B, and the state matrix P are defined. x and P y and the time parameter matrix T:

[0155]

[0156]

[0157]

[0158] Among them, t s Indicates the start time of lane change, t e Indicates the time when the lane change ends.

[0159] By combining equations (25-27), we can obtain the coefficient matrix of the polynomial curve:

[0160]

[0161] The queue distributed controller module 22 includes two parts: the queue vertical controller unit 221 and the queue horizontal controller unit 222. The two units are further described below.

[0162] Queue vertical controller 221, such as Figure 8 As shown, each vehicle is equipped with a model predictive controller. Vehicles share information using V2V technology. The leading vehicle transmits its position, speed, and acceleration information to the following vehicle. The following vehicle tracks the speed of the leading vehicle by solving an optimization problem based on its own vehicle status and the state of the leading vehicle, and maintains the desired inter-vehicle distance.

[0163] Model predictive control is a model-based closed-loop optimization control strategy that uses the model of the controlled object, the current state of the system, and future control inputs to predict the future trend of the controlled object and solve the optimization problem in the finite time domain online.

[0164] Based on feedback linearization theory, the third-order nonlinear model of the vehicle is linearized to model the longitudinal motion of the vehicle:

[0165]

[0166] In the formula, s represents the longitudinal displacement of the vehicle, v(t) represents the velocity of the vehicle, a(t) represents the acceleration of the vehicle, and τ represents the inertial hysteresis of the vehicle's longitudinal dynamics. des The desired acceleration of the vehicle is also the control input of the vehicle after feedback linearization.

[0167] Using a fixed-distance geometric configuration, the spacing error and speed error of the i-th vehicle are defined as follows:

[0168]

[0169] In the formula, l i-1 Let d represent the length of the (i-1)th vehicle. des This indicates the desired workshop distance.

[0170] Combining equations (29-30), the differential equation for the longitudinal error of the vehicle is established:

[0171] x(t)=Ax(t)+Bu(t)+Cω(t) (31)

[0172] In the formula:

[0173]

[0174] x=(e s e v a i u=a des ω=a i-1

[0175] Where x represents the state vector, u represents the control input, and ω represents the disturbance input.

[0176] The goal of queue control is to reduce spacing and velocity errors, while avoiding excessive acceleration and impact. The cost function is defined as follows:

[0177]

[0178]

[0179] In the formula, Let represent a quadratic function, and Q and R represent the system error and control input weight matrix, respectively.

[0180] The control sequence is then solved by solving a quadratic programming problem, and only the first control variable is applied to the system. This process is then repeated at the next sampling time.

[0181] Queue lateral controller 222, such as Figure 9 As shown, each vehicle is equipped with a linear quadratic controller. The lead vehicle sends the planned path to all following vehicles. The following vehicles calculate their own paths based on the geometry of the queuing and, combined with the vehicle status, use the lateral controller to calculate the front wheel angle to achieve path tracking.

[0182] The linear quadratic regulator is a typical algorithm for optimal control of linear systems. For linear systems:

[0183]

[0184] The control objective of a linear quadratic regulator is to select a suitable control law U. * (t) = -KX(t), which makes the following performance index function take the minimum value.

[0185]

[0186] In the formula, Q and R represent the state matrix and control matrix, respectively.

[0187] If A, B, Q, and R are known, the feedback control law can be solved by iteratively calculating the Riccati equation.

[0188] Using a linear two-degree-of-freedom model to describe the lateral dynamics of a vehicle, we can obtain the differential equations of the vehicle's lateral motion:

[0189]

[0190] In the formula, a and b represent the lengths from the center of mass to the front and rear axes, respectively, and v x and v y Let represent the longitudinal and lateral speeds of the vehicle at its center of mass in the vehicle coordinate system, respectively, and δ be the front wheel steering angle. α is the yaw angle, β is the centroid sideslip angle, and α is the yaw angle. f and α r C represents the slip angles of the front and rear wheels, respectively. f and C r I represents the lateral stiffness of the front and rear wheels, respectively. z This represents the moment of inertia of the center of mass about the z-axis.

[0191] Projecting the vehicle onto the reference path, the spacing error can be calculated using the distance e between the vehicle's centroid and the projection point. dThe heading error is expressed using the following formula:

[0192]

[0193] Based on lateral dynamics, geometric relationships, and Frenet's formula, the differential equation for the vehicle's lateral error can be obtained:

[0194]

[0195] In the formula,

[0196]

[0197] u=δ,

[0198] When ω(t) does not exist, the feedback control law u of the system can be derived. * (t)=-Kx(t).

[0199] Due to the existence of ω(t), substituting the feedback control law into the transverse error differential equation reveals a steady-state error in the system. To eliminate this steady-state error, a feedforward control variable δ needs to be introduced. f ,Right now:

[0200] u(t)=-Kx(t)+δ f (39)

[0201] Substituting equation (38) into equation (37), let The feedforward control quantity can be solved as follows:

[0202]

[0203] In summary, the cloud-based predictive cruise and lane-changing control system based on the cloud control platform proposed in this invention has the following beneficial effects:

[0204] It allows target vehicles to establish connections with environmental vehicles, roadside infrastructure, and cloud control platforms, providing target vehicles with wide-area dynamic traffic information, thereby better predicting upcoming events and enabling vehicles to adjust their driving strategies more proactively. In addition, compared with the vehicle-road cooperative technology that existing predictive cruise control technology relies on, the participation of the cloud control platform can overcome its limitations in communication distance and perception range, improve the scope and predictability of decision-making and planning, and make it easier to solve control strategies in real time.

[0205] The deployed predictive cruise and lane-changing decision-making strategy improves the safety, efficiency, and economy of the target platoon's vehicle operation. Based on a model predictive control framework, this strategy uses the real-time status and future trends of vehicles in the surrounding environment as predictive information. Combined with the target platoon's vehicle operating status and vehicle model, it establishes an optimization problem with the long-term expected benefits of the target platoon's vehicle operation as the optimization objective. This collaboratively solves for the optimal cruise and lane-changing strategy of the platoon within the prediction time domain. Compared to existing predictive cruise control, this method expands the optimization dimensions and improves the optimality of the optimization results. Furthermore, since the optimization problem's objective is the long-term benefits of the target platoon's vehicles, it overcomes the short-sightedness of existing lane-changing decision-making technologies.

[0206] The deployed trajectory planning algorithm and distributed queue controller supplement the execution process of lane changing and cruising, ensuring the safety of the target queue of vehicles. By analyzing the braking process of the vehicles, a safe lane changing distance model for the queue is established, and a safety degradation strategy is designed for unsafe lane changing situations. The cruising trajectory and lane changing trajectory are established based on a uniform acceleration model and a fifth-order polynomial curve, respectively. In combination with the control objectives of the queue, longitudinal and transverse queue controllers are designed based on a distributed model predictive controller and a linear quadratic regulator, respectively, to ensure the tracking performance of the queue.

[0207] Next, referring to the accompanying drawings, a queue predictive cruise and lane change control method based on a cloud control platform, according to an embodiment of the present invention, is described.

[0208] Figure 10 This is a flowchart illustrating a cloud-based predictive cruise and lane-changing control method for queuing provided in an embodiment of the present invention.

[0209] like Figure 10 As shown, the platoon predictive cruise and lane change control method based on the cloud control platform includes the following steps:

[0210] In step S101, the real-time status and vehicle parameter information of the target queue vehicles are obtained through the edge cloud in the cloud control platform, and dynamic traffic environment information is collected through roadside infrastructure.

[0211] In step S102, the queue predictive cruise and lane change decision module in the cloud control platform is used to process real-time status, parameter information and dynamic traffic environment information to solve for the optimal longitudinal acceleration and optimal lateral lane change decision of the target queue vehicles, and then send them to the target queue vehicles.

[0212] Furthermore, in one embodiment of the present invention, the queue predictive cruise and lane-changing decision module includes: a dynamic sensing domain unit, a prediction unit, a cost function and constraint unit, and an optimization solution unit, wherein,

[0213] The dynamic perception domain unit is used to filter multiple environmental vehicles around the target queue of vehicles based on dynamic traffic environment information and real-time status, and to obtain the environmental vehicles that affect the driving of the target queue of vehicles within a preset prediction time domain.

[0214] The prediction module is used to establish the state transition equations of the target queue vehicles and the environment vehicles in the discrete space, so as to estimate the future states of the target queue vehicles and the environment vehicles.

[0215] The cost function and constraint unit are used to establish the cost function for predictive cruise and lane-changing decisions of the queue based on multiple preset optimization objectives and future states, and to constrain the vehicles in the target queue.

[0216] The optimization unit is used to classify the cost function of platoon predictive cruise and lane change decision into multiple nonlinear programming subproblems to solve for the optimal longitudinal acceleration and optimal lateral lane change decision, and then send the solution to the target platoon vehicles.

[0217] In step S103, the optimal longitudinal acceleration and optimal lateral lane change decision are received by the on-board platform of the target platoon vehicles.

[0218] Furthermore, in one embodiment of the present invention, the target convoy vehicles include a lead vehicle and multiple follower vehicles, wherein each vehicle in the target convoy is equipped with a model predictive controller and a linear quadratic regulator. The target convoy vehicles transmit their own position, speed, and acceleration to the follower vehicles and use the model predictive controller to calculate longitudinal control commands. The lead vehicle sends the trajectory planning route to each follower vehicle and uses the linear quadratic regulator to calculate lateral control commands.

[0219] In step S104, the trajectory planning algorithm module in the vehicle platform is used to plan the driving trajectory of the target queue of vehicles to obtain the trajectory planning route.

[0220] Furthermore, in one embodiment of the present invention, the trajectory planning algorithm module includes: a cloud instruction parsing unit, a lane change safety judgment unit, a safety degradation strategy unit, and a trajectory generation unit, wherein,

[0221] The cloud-based instruction parsing unit is used to analyze the optimal longitudinal acceleration and optimal lateral lane change decision to obtain the recommended acceleration and recommended lateral lane change decision that the lead vehicle should take.

[0222] The lane change safety judgment unit is used to analyze the drivable safe area of ​​the lead vehicle by recommending acceleration and recommending lateral lane change decision to determine whether the lane change meets the preset safety conditions.

[0223] The safety degradation strategy unit is used to obtain information about the vehicle ahead of the lead vehicle when the preset safety conditions are not met, and outputs a safety acceleration command for the target convoy vehicles with the goal of maintaining a safe headway.

[0224] The trajectory generation unit is used to plan the driving trajectory of the lead vehicle based on the uniform acceleration model and the fifth-order polynomial curve, the recommended acceleration, the recommended lateral lane change decision, the lane change safety judgment result, and the safety acceleration command, so as to obtain the trajectory planning route.

[0225] In step S105, the queue distributed controller module in the vehicle platform is used to track and control the vehicles in the target queue, and the real-time status of the vehicles in the target queue is uploaded to the cloud control platform to form a vehicle-cloud rolling closed-loop control.

[0226] Furthermore, in one embodiment of the present invention, the queue distributed controller module includes: a queue vertical controller unit and a queue horizontal controller unit, wherein,

[0227] The queue longitudinal controller unit is used to acquire the position, speed and acceleration of the vehicle in front and the state of the following vehicles in the target queue, so as to calculate the expected acceleration of each following vehicle. Each following vehicle maintains distance and tracks speed from the vehicle in front based on its corresponding expected acceleration.

[0228] The queue lateral controller unit is used to plan the route based on the trajectory and the state of each following vehicle to calculate the front wheel angle of each following vehicle. Each following vehicle tracks the lead vehicle according to its corresponding front wheel angle.

[0229] It should be noted that the foregoing explanation of the embodiment of the cloud-based predictive cruise and lane change control device also applies to the cloud-based predictive cruise and lane change control method of this embodiment, and will not be repeated here.

[0230] The platooning predictive cruise and lane-changing control method based on a cloud control platform proposed in this embodiment of the invention has the following advantages:

[0231] Beneficial effects:

[0232] It allows target vehicles to establish connections with environmental vehicles, roadside infrastructure, and cloud control platforms, providing target vehicles with wide-area dynamic traffic information, thereby better predicting upcoming events and enabling vehicles to adjust their driving strategies more proactively. In addition, compared with the vehicle-road cooperative technology that existing predictive cruise control technology relies on, the participation of the cloud control platform can overcome its limitations in communication distance and perception range, improve the scope and predictability of decision-making and planning, and make it easier to solve control strategies in real time.

[0233] The deployed predictive cruise and lane-changing decision-making strategy improves the safety, efficiency, and economy of the target platoon's vehicle operation. Based on a model predictive control framework, this strategy uses the real-time status and future trends of vehicles in the surrounding environment as predictive information. Combined with the target platoon's vehicle operating status and vehicle model, it establishes an optimization problem with the long-term expected benefits of the target platoon's vehicle operation as the optimization objective. This collaboratively solves for the optimal cruise and lane-changing strategy of the platoon within the prediction time domain. Compared to existing predictive cruise control, this method expands the optimization dimensions and improves the optimality of the optimization results. Furthermore, since the optimization problem's objective is the long-term benefits of the target platoon's vehicles, it overcomes the short-sightedness of existing lane-changing decision-making technologies.

[0234] The deployed trajectory planning algorithm and distributed queue controller supplement the execution process of lane changing and cruising, ensuring the safety of the target queue of vehicles. By analyzing the braking process of the vehicles, a safe lane changing distance model for the queue is established, and a safety degradation strategy is designed for unsafe lane changing situations. The cruising trajectory and lane changing trajectory are established based on a uniform acceleration model and a fifth-order polynomial curve, respectively. In combination with the control objectives of the queue, longitudinal and transverse queue controllers are designed based on a distributed model predictive controller and a linear quadratic regulator, respectively, to ensure the tracking performance of the queue.

[0235] Figure 11 This is a schematic diagram of an electronic device provided in an embodiment of the present invention. The electronic device may include:

[0236] The memory 1101, the processor 1102, and the computer program stored on the memory 1101 and executable on the processor 1102.

[0237] When the processor 1102 executes the program, it implements the queue predictive cruise and lane change control method based on the cloud control platform provided in the above embodiments.

[0238] Furthermore, electronic devices also include:

[0239] Communication interface 1103 is used for communication between memory 1101 and processor 1102.

[0240] The memory 1101 is used to store computer programs that can run on the processor 1102.

[0241] The memory 1101 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0242] If the memory 1101, processor 1102, and communication interface 1103 are implemented independently, then the communication interface 1103, memory 1101, and processor 1102 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be divided into address buses, data buses, control buses, etc. For ease of representation, Figure 11 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0243] Optionally, in a specific implementation, if the memory 1101, processor 1102, and communication interface 1103 are integrated on a single chip, then the memory 1101, processor 1102, and communication interface 1103 can communicate with each other through an internal interface.

[0244] The processor 1102 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention.

[0245] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described cloud-based predictive cruise and lane-changing control method for queuing.

[0246] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0247] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0248] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0249] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0250] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0251] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0252] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0253] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A platoon predictive cruise and lane-changing control system based on a cloud control platform, characterized in that, include: The cloud control platform is equipped with a queuing predictive cruise and lane change decision module, which is used to acquire real-time status and parameter information of vehicles in the target queuing as well as dynamic traffic environment information. The queuing predictive cruise and lane change decision module processes the real-time status, parameter information and dynamic traffic environment information to solve the optimal longitudinal acceleration and optimal lateral lane change decision of the vehicles in the target queuing, and sends them to the vehicles in the target queuing. The vehicles in the target queuing include a lead vehicle and multiple follower vehicles. The vehicle-mounted platform is mounted on the vehicles in the target platoon. The lead vehicle's platform is equipped with a trajectory planning algorithm module. Each vehicle in the target platoon has a platoon distributed controller module. Based on the optimal longitudinal acceleration and the optimal lateral lane-changing decision, the trajectory planning algorithm module is used to plan the lead vehicle's trajectory to obtain a planned route. The platoon distributed controller module then tracks and controls the target platoon vehicles, and uploads the real-time status of the target platoon vehicles to the cloud control platform, forming a vehicle-cloud rolling closed-loop control system. Each vehicle in the target convoy is equipped with a model predictive controller and a linear quadratic regulator. The target convoy vehicles transmit their own position, speed, and acceleration to the following vehicles and use the model predictive controller to calculate longitudinal control commands. The lead vehicle sends the trajectory planning route to each following vehicle and uses the linear quadratic regulator to calculate lateral control commands. The trajectory planning algorithm module includes: a cloud instruction parsing unit, a lane change safety judgment unit, a safety degradation strategy unit, and a trajectory generation unit, wherein... The cloud-based instruction parsing unit is used to parse the optimal longitudinal acceleration and the optimal lateral lane change decision to obtain the recommended acceleration and recommended lateral lane change decision that the lead vehicle should drive. The lane change safety judgment unit is used to analyze the drivable safe area of ​​the lead vehicle through the recommended acceleration and the recommended lateral lane change decision, so as to determine whether the lane change meets the preset safety conditions. The safety degradation strategy unit is used to obtain the information of the vehicle in front of the lead vehicle when the preset safety conditions are not met, and output the safety acceleration command of the lead vehicle with the goal of maintaining a safe headway. The trajectory generation unit is used to plan the driving trajectory of the navigator based on the uniform acceleration model and the fifth-order polynomial curve, the recommended acceleration, the recommended lateral lane change decision, the lane change safety judgment result and the safety acceleration command, so as to obtain the planned trajectory route. The queue distributed controller module includes: a queue vertical controller unit and a queue horizontal controller unit, wherein... The queue longitudinal controller unit is used to acquire the position, speed and acceleration of the vehicle in front and the state of the following vehicles in the target queue, so as to calculate the expected acceleration of each following vehicle. Each following vehicle maintains the distance and tracks the speed of the vehicle in front according to its corresponding expected acceleration. The queue lateral controller unit is used to calculate the front wheel angle of each following vehicle and perform path tracking based on the trajectory planning route and the state of each following vehicle.

2. The queue predictive cruise and lane-changing control system based on a cloud control platform according to claim 1, characterized in that, The cloud control platform obtains the real-time status and vehicle parameter information of the target queue of vehicles through the edge cloud.

3. The platoon predictive cruise and lane-changing control system based on a cloud control platform according to claim 1, characterized in that, Also includes: Roadside infrastructure, which is connected to the cloud control platform, is used to collect the dynamic traffic environment information and send the dynamic traffic environment information to the cloud control platform; The vehicle-mounted intelligent remote information processing terminal T-BOX is mounted on the vehicles in the target convoy and is used to enable the vehicle-mounted platform to communicate and process information with the cloud control platform.

4. The platoon predictive cruise and lane-changing control system based on a cloud control platform according to claim 1, characterized in that, The queue predictive cruise and lane-changing decision module includes: a dynamic sensing domain unit, a prediction unit, a cost function and constraint unit, and an optimization solution unit, wherein... The dynamic perception domain unit is used to filter multiple environmental vehicles around the target queue of vehicles based on the dynamic traffic environment information and the real-time status, and to obtain the environmental vehicles that affect the driving of the target queue of vehicles within a preset prediction time domain. The prediction unit is used to establish the state transition equations of the target queue vehicles and the environmental vehicles in the discrete space, so as to estimate the future states of the target queue vehicles and the environmental vehicles. The cost function and constraint unit are used to establish a queue predictive cruise and lane-changing decision cost function based on multiple preset optimization objectives and the future state, and to constrain the target queue vehicles. The optimization solution unit is used to classify the cost function of the platoon predictive cruise and lane change decision into multiple nonlinear programming subproblems to solve for the optimal longitudinal acceleration and optimal lateral lane change decision, and then send them to the target platoon vehicles.

5. A method for predictive cruise and lane-changing control based on a cloud control platform, characterized in that, The platooning predictive cruise and lane-changing control system based on a cloud control platform, as described in any one of claims 1-4, comprises the following steps: The real-time status and vehicle parameter information of the target queue vehicles are obtained through the edge cloud in the cloud control platform, and dynamic traffic environment information is collected through roadside infrastructure. The queuing predictive cruise and lane-changing decision module in the cloud control platform processes the real-time status, parameter information, and dynamic traffic environment information to solve for the optimal longitudinal acceleration and optimal lateral lane-changing decision of the target queuing vehicles, and sends the decision to the target queuing vehicles. The optimal longitudinal acceleration and the optimal lateral lane-changing decision are received by the onboard platform of the vehicles in the target platoon. The trajectory planning algorithm module in the vehicle platform is used to plan the driving trajectory of the target convoy of vehicles to obtain the trajectory planning route. The vehicle-mounted platform uses a queue distributed controller module to track and control the vehicles in the target queue, and uploads the real-time status of the vehicles in the target queue to the cloud control platform, forming a vehicle-cloud rolling closed-loop control.

6. An electronic device, characterized in that, include: The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the cloud-based predictive cruise and lane-change control method as described in claim 5.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement the cloud-based predictive cruise and lane-changing control method for queuing as described in claim 5.

Images