Multi-vehicle queue cooperative control method based on dynamic model predictive control and medium

Abstract

The invention discloses a multi-vehicle queue cooperative control method based on dynamic model predictive control and a medium. The method comprises the following steps: establishing a vehicle dynamics model, a vehicle motion state equation, a communication topological structure and a queue state space model; establishing a two-stage control model which comprises a vehicle cut-in optimization model and a distributed model prediction controller; the vehicle cut-in optimization model comprises a target function and a queue stability constraint; the objective function comprises an adjustment cost function during cut-in; the optimization direction is to minimize the adjustment cost; solving and obtaining the optimal cut-in time and position of the cut-in vehicle; and dynamically controlling the vehicle by using the distributed model predictive controller. According to the cut-in vehicle control model, the energy consumption in the cut-in process is remarkably reduced while the cut-in efficiency and the safety are guaranteed, the stability and the cooperative capability of a queue system are enhanced, and an efficient and reliable solution is provided for multi-queue control in an intelligent traffic system.

Description

Technical Field

[0001] The present invention relates to the technical field of multi-vehicle queue cooperative control, and particularly relates to a multi-vehicle queue cooperative control method and medium based on dynamic model predictive control. Background Art

[0002] With the rapid development of intelligent transportation systems and autonomous driving technologies, the queue control of autonomous vehicles has gradually become an important research direction for improving traffic efficiency and driving safety. Existing technologies mainly focus on the control of a single queue, while there is less research on cooperative control in multi-queue systems. In a multi-queue environment, information interaction, cooperative control between different queue systems, and vehicle cut-in in different queues are the keys to achieving efficient operation of the system. The information flow topology structure in the queue system directly affects the stability, safety, and scalability of the queue. Especially in a multi-queue system, the unidirectionality of communication between queues and the topology structure are of great significance for the implementation of cooperative control. At the same time, it is a basic requirement for vehicles within a queue to maintain a stable spacing and speed, and vehicle cut-in involves the coordination of vehicles within the queue and the information flow interaction between queues. How to achieve high efficiency and economy while ensuring vehicle safety is a difficult and key problem in existing technologies. Summary of the Invention

[0003] The main objective of the present invention is to provide a multi-vehicle queue cooperative control method and medium based on dynamic model predictive control, aiming to solve the problem that when a vehicle cuts into a multi-queue system, it cannot achieve high efficiency and economy while ensuring vehicle safety.

[0004] To achieve the above objective, the multi-vehicle queue cooperative control method based on dynamic model predictive control proposed by the present invention includes the following steps: Establish a vehicle dynamics model, a vehicle motion state equation, a communication topology structure, a queue state space model, and a two-stage control model for a multi-queue system. The two-stage control model includes a vehicle cut-in optimization model and a distributed model predictive controller. The vehicle cut-in optimization model includes an objective function and queue stability constraints. The objective function includes a cut-in vehicle cost function and a queue cost function for measuring the adjustment costs required for the cut-in vehicle and the queue when the cut-in vehicle reaches the optimal cut-in position. The optimization direction is to minimize the adjustment cost. Obtain the states of the cut-in vehicle and the queue vehicles, and solve based on the vehicle cut-in optimization model to obtain the optimal cut-in time, the optimal cut-in position of the cut-in vehicle, and the queue expected trajectory. According to the queue expected trajectory, the optimal cut-in time, and the optimal cut-in position of the cut-in vehicle, use the distributed model predictive controller to perform dynamic control on the vehicle, so that the queue quickly returns to a dynamically stable state.

[0005] Preferably, the vehicle dynamics model includes the following expressions:

[0006] Air resistance:

[0007] Rolling resistance:

[0008] Where: represents the air density; represents the gravitational constant; , , , , and respectively represent the mass, speed, rolling resistance coefficient, frontal surface area and air resistance coefficient of the i-th vehicle in the k-th queue; , , and respectively represent the position, speed, acceleration and driving force or braking torque of the i-th vehicle in the -th queue at time t.

[0009] Preferably, the objective function of the vehicle cut-in optimization model includes: ; Where: represents the adjustment cost required for the cut-in vehicle and the queue when the cut-in vehicle reaches the optimal cut-in position; represents the dynamic adjustment cost of the cut-in operation to other vehicles in the queue; is the feasible solution domain of the optimal cut-in position, is the energy consumption at each possible cut-in point, and the solution with the minimum energy consumption is found from all results .

[0010] Preferably, the expression of the queue cost function includes:

[0011] Where: is the time step for the queue to complete the reserved space, is the time domain for the queue to complete the reserved space, and , is the time for the queue to complete the reserved space; represents the air density; represents the gravitational constant; , , , and respectively represent the mass, speed, rolling resistance coefficient, frontal surface area of the vehicle, and air resistance coefficient of the i-th vehicle in the k-th queue.

[0012] Preferably, the expression of the cut-in vehicle cost function includes:

[0013] In the formula, is the time step for the cut-in vehicle to reach the optimal cut-in point, is the longitudinal displacement distance of the vehicle within the optimal discrete time interval for the cut-in vehicle to reach, ; ; represents the air density; represents the gravitational constant; , , , , and respectively represent the mass, speed, rolling resistance coefficient, frontal surface area, quantity, and air resistance coefficient of the cut-in vehicle.

[0014] Preferably, the step of s500 includes the following steps; Perform trajectory planning and control on the cut-in vehicle using the distributed model predictive controller according to the optimal cut-in time and optimal cut-in position of the cut-in vehicle; Complete adjusting the speed and position of the vehicles in the queue using the distributed model predictive controller according to the expected queue trajectory; Use the distributed model predictive controller to complete the optimization of queue stability and dynamic behavior within the dynamic convergence time.

[0015] Preferably, the distributed model predictive controller includes a cut-in vehicle sub-controller. The optimization objective of the cut-in vehicle sub-controller is to minimize the deviation between the speed and acceleration of the cut-in vehicle. The distributed model predictive controller includes a cut-in time constraint and a cut-in distance constraint. The cut-in time constraint is used to constrain the cut-in vehicle to reach the optimal cut-in position at the optimal cut-in time, and the cut-in distance constraint is used to constrain the cut-in vehicle to maintain a dynamic safety distance from other vehicles.

[0016] Preferably, the communication topology structure includes a leader-follower tracking strategy between queues, and the leader vehicle of each queue only receives the information of the tail vehicle of the previous queue.

[0017] Preferably, the queue stability constraints include vehicle spacing constraints, queue spacing constraints, speed consistency constraints, and queue communication topology constraints; The expressions of the vehicle spacing constraint, the platoon spacing constraint, the speed consistency constraint, and the platoon communication topology constraint are as follows: ; ; ; ; In the formula: represents the position of the k th vehicle in the i th platoon; represents the position of the leader vehicle of platoon k; represents the position of the last vehicle of the previous platoon; represents the position of the k th vehicle at time t in the i th platoon; represents the initial speed; represents the minimum safe distance between any two vehicles; represents the minimum safe distance between any two platoons; represents the expected state of the leader vehicle of the kth platoon; represents the state information of other leader vehicles defined in the communication topology.

[0018] A medium stores executable instructions, and the executable instructions include program codes for executing the multi-vehicle platoon cooperative control method based on distributed model predictive control as described above when being executed.

[0019] In the technical solution of the present invention, precise control of the cutting-in vehicle is achieved: through the multi-dimensional control functions of position, speed, and acceleration, the cutting-in vehicle can accurately track the optimal cut-in point and the reference trajectory, avoiding deviations or unstable behaviors during the cut-in process, thereby improving the accuracy of vehicle cut-in.

[0020] Ensure the smoothness of the cut-in process: The acceleration control term can effectively limit the violent acceleration or deceleration behavior of the vehicle during the cut-in process, avoid interference of the violent movement of the vehicle to the system, and at the same time reduce the mechanical wear and energy consumption of the vehicle, ensuring the smoothness of vehicle movement.

[0021] Improve the stability of the platoon system: By optimizing the influence of the cutting-in vehicle on other vehicles in the platoon, reducing the adjustment frequency of the speed and acceleration of other vehicles in the platoon, ensuring that the platoon can quickly return to a stable state after the cutting-in vehicle is inserted, thereby improving the stability and coordination of the entire platoon. At the same time, by setting the safety distance constraint between the cutting-in vehicle and the platoon vehicles, the risk of conflict between vehicles during the cut-in process can be effectively reduced, ensuring the safety of the entire cut-in process.

[0022] Through a distributed model predictive controller, it supports distributed computing and real-time optimization: the combination of a distributed control architecture and a dynamic optimization model enables the system to respond to complex traffic dynamics in real time in a vehicle-to-everything (V2X) environment, with high scalability and can be applied to large-scale autonomous driving platoon systems. The cut-in vehicle control model of the present invention significantly reduces the energy consumption during the cut-in process while ensuring the cut-in efficiency and safety, and enhances the stability and cooperation ability of the platoon system, providing an efficient and reliable solution for multi-platoon control in an intelligent transportation system. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on the structures shown in these drawings.

[0024] Figure 1 It is a communication topology structure diagram of the leader vehicle - following vehicle in the multi-vehicle platoon cooperative control method of the dynamic model predictive control of the present invention.

[0025] Figure 2 It is a vehicle cut-in speed change diagram of the multi-platoon system in one embodiment of the multi-vehicle platoon cooperative control method of the dynamic model predictive control of the present invention.

[0026] Figure 3 It is a vehicle cut-in displacement change diagram of the multi-platoon system in one embodiment of the multi-vehicle platoon cooperative control method of the dynamic model predictive control of the present invention.

[0027] Figure 4 It is a weight matrix value table of the distributed model predictive controller in one embodiment of the multi-vehicle platoon cooperative control method of the dynamic model predictive control of the present invention.

[0028] The realization, functional features and advantages of the object of the present invention will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0029] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0030] In addition, in the present invention, descriptions such as "first", "second", etc. are only for descriptive purposes and should not be construed as indicating or implying their relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first", "second" may explicitly or implicitly include at least one of such features. In the description of the present invention, the meaning of "a plurality of" is at least two, such as two, three, etc., unless otherwise specifically defined.

[0031] In the present invention, unless otherwise clearly specified and defined, terms such as "connection", "fixation", etc. should be understood in a broad sense. For example, "fixation" can be a fixed connection, a detachable connection, or integrated; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and can be the communication inside two components or the interaction relationship between two components, unless otherwise clearly defined. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0032] In addition, the technical solutions between various embodiments of the present invention can be combined with each other, but it must be based on the ability of those of ordinary skill in the art to implement. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the protection scope required by the present invention.

[0033] Please refer to Figures 1-4 , the present invention provides a multi-vehicle queue cooperative control method based on dynamic model prediction control, including the following steps: S100, establish a vehicle dynamics model, a vehicle motion state equation, a communication topology structure, a queue state space model, and a multi-queue system model; S200, establish a two-stage control model, the two-stage control model includes a vehicle cut-in optimization model and a distributed model predictive controller; the vehicle cut-in optimization model includes an objective function and a queue stability constraint; the objective function includes a cut-in vehicle cost function and a queue cost function for measuring the adjustment costs required for the cut-in vehicle and the queue when the cut-in vehicle reaches the optimal cut-in position; the optimization direction is to minimize the adjustment cost; S300, obtain the states of the cut-in vehicle and the queue vehicles, and solve based on the vehicle cut-in optimization model to obtain the optimal cut-in time, the optimal cut-in position of the cut-in vehicle, and the queue expected trajectory; S400, perform dynamic control on the vehicle according to the queue expected trajectory, the optimal cut-in time, and the optimal cut-in position of the cut-in vehicle by using the distributed model predictive controller, so that the queue quickly returns to a dynamically stable state.

[0034] In the technical solution of the present invention, precise control of the cutting-in vehicle is achieved: through a multi-dimensional control function of position, speed, and acceleration, the cutting-in vehicle can accurately track the optimal cut-in point and the reference trajectory, avoiding deviations or unstable behaviors during the cut-in process, thereby improving the accuracy of vehicle cut-in.

[0035] Ensure the smoothness of the cut-in process: The acceleration control term can effectively limit the violent acceleration or deceleration behavior of the vehicle during the cut-in process, avoid interference of the vehicle's violent movement to the system, and at the same time reduce the mechanical wear and energy consumption of the vehicle, ensuring the smoothness of vehicle movement.

[0036] Improve the stability of the queue system: By optimizing the influence of the cutting-in vehicle on other vehicles in the queue, reducing the adjustment frequency of the speed and acceleration of other vehicles in the queue, ensuring that the queue can quickly return to a stable state after the cutting-in vehicle is inserted, thereby improving the stability and coordination of the entire queue. At the same time, by setting the safety distance constraint between the cutting-in vehicle and the queue vehicles, the risk of conflict between vehicles during the cut-in process can be effectively reduced, ensuring the safety of the entire cut-in process.

[0037] Through the distributed model predictive controller, support for distributed computing and real-time optimization: The combination of the distributed control architecture and the dynamic optimization model enables the system to respond to complex traffic dynamics in real time in the vehicle networking environment, with high scalability and can be applied to large-scale autonomous driving queue systems. The cut-in vehicle control model of the present invention significantly reduces the energy consumption during the cut-in process while ensuring the cut-in efficiency and safety, and enhances the stability and cooperation ability of the queue system, providing an efficient and reliable solution for multi-queue control in intelligent transportation systems.

[0038] In another embodiment of the present invention, the vehicle dynamics model includes the following expressions: ; Air resistance: ; Rolling resistance: ; Where: Represents the air density; Represents the gravitational constant; , , , And Respectively represent the mass, speed, rolling resistance coefficient, front surface area of the vehicle, and air resistance coefficient of the i-th vehicle in the k-th queue; , , And respectively represent the position, speed, acceleration, and driving force or braking torque of the $i$-th vehicle in the $k$-th queue at time $t$.

[0039] Specifically, a multi-queue system model is constructed: Consider a multi-queue system composed of multiple vehicle queues, where each queue contains a lead vehicle and several following vehicles. The communication topology structure between and within the queues is described by a directed graph, enabling information to flow between the queues, thereby achieving the cooperative control of the multi-queue.

[0040] Define the set of vehicles within queue $k$ as: . For the entire system, which consists of multiple queues, the present invention defines the set of all vehicles as , which includes all the vehicles within the queues: ; Here represents the $i$-th vehicle in the $k$-th queue, where represents the lead vehicle, and other numbers represent the following vehicles, represents the last vehicle in queue $k$, and the cut-in vehicle is defined as .

[0041] Communication topology structure: A unidirectional communication topology structure is adopted within the queue, with the lead vehicle as the root node to ensure that all following vehicles can obtain the status information of the lead vehicle. A lead vehicle - last vehicle tracking strategy is adopted between the queues, and the lead vehicle of a queue receives the information of the last vehicle of the previous queue.

[0042] Vehicle dynamics model: The state of each vehicle is described by position, speed, acceleration, and driving force. A vehicle dynamics model is established through state - space equations, and combined with the physical constraint conditions of the vehicle, the vehicle cut - in is modeled.

[0043] Based on the state - space equations, considering the vehicle's position , speed , torque , control input , as well as physical constraints such as air resistance and rolling resistance, to ensure that the dynamic behavior of the vehicle can be fully characterized. Specifically as follows: ; Air resistance: ; Rolling resistance: ; Among them, is the air density, is the air resistance coefficient, is the frontal surface area of the vehicle, is the rolling resistance coefficient, $m$ is the vehicle mass, and $g$ is the gravitational constant.

[0044] Using a discrete-time model, the motion of the vehicle at each time step is represented by a state transition equation as follows: ; Wherein, , respectively represent the state and input of vehicle i in queue k at time t: ; .

[0045] In another embodiment of the present invention, the objective function of the vehicle cut-in optimization model includes: ; In the formula: represents the adjustment cost required for the cut-in vehicle and the queue when the cut-in vehicle reaches the optimal cut-in position; represents the dynamic adjustment cost of the cut-in operation to other vehicles in the queue; is the feasible solution domain of the optimal cut-in position, is the energy consumption at each possible cut-in point, and the solution with the minimum energy consumption is found from all the results .

[0046] Based on the above, the optimization objective of the first stage (vehicle cut-in optimization model) can be expressed as: ; Wherein, P is the feasible solution domain of the optimal cut-in position. Considering the constraints of the states of the vehicles in front of and behind the cut-in vehicle in the actual process, the present invention calculates the energy consumption at each possible cut-in point and finds the solution with the minimum energy consumption from all the results, so as to determine the optimal cut-in position and strategy to provide the desired trajectory for the second stage (distributed model predictive controller for vehicle control); The desired trajectory of the cut-in vehicle ; and the desired trajectories of the vehicles in the multi-queue system .

[0047] In another embodiment of the present invention, the expression of the queue cost function includes: ;

[0048] In the formula: is the time step for the queue to complete the reserved space, is the time domain for the queue to complete the reserved space, and , is the time for the queue to complete the reserved space; represents the air density; represents the gravitational constant; , , , and respectively represent the mass, speed, rolling resistance coefficient, frontal surface area of the vehicle, and air resistance coefficient of the i-th vehicle in the k-th queue.

[0049] Specifically, calculate the driving energy consumption for each vehicle in the queue to vacate the cut-in position: ; Calculate the air resistance of each vehicle in the queue: ; Calculate the rolling resistance of each vehicle in the queue: ; Therefore, the queue adjustment objective function is: ; where k2 is the time step when the queue completes reserving space, Nt2 is the time domain when the queue completes reserving space, satisfying , and T2 is the time when the queue completes reserving space.

[0050] When the queue length , where L(k) is the length of the queue at time k, only the vehicles cutting into the target queue in the queue system change their states. At this time, the energy loss of reserving space in the queue only needs to calculate the energy loss of a single queue, which can be specifically expressed as: ; Queue adjustment objective function design. The queue cost function is used to measure the dynamic adjustment cost of the cut-in operation to other vehicles in the queue.

[0051] In another embodiment of the present invention, the expression of the cut-in vehicle cost function includes: ;

[0052] In the formula, represents the time step when the cut-in vehicle reaches the optimal cut-in point, represents the time domain when the cut-in vehicle reaches the optimal cut-in point, satisfying , represents the time when the cut-in vehicle completes the relative distance, represents the time interval the longitudinal displacement distance of the vehicle within, ; represents the air density; represents the gravitational constant; , , , , and respectively represent the mass, speed, rolling resistance coefficient, front surface area, number, and air resistance coefficient of the cutting-in vehicle.

[0053] Specifically, calculate the driving energy consumption of the cutting-in vehicle during the acceleration or deceleration from the current position to the cutting-in position: ; Calculate the air resistance of the cutting-in vehicle during the acceleration or deceleration from the current position to the cutting-in position: ; Calculate the rolling resistance of the cutting-in vehicle during the acceleration or deceleration from the current position to the cutting-in position: .

[0054] In another embodiment of the present invention, the steps of S400 include the following steps; S410, use the distributed model predictive controller to perform trajectory planning and control on the cutting-in vehicle according to the optimal cutting-in time and optimal cutting-in position of the cutting-in vehicle; S420, use the distributed model predictive controller to complete the adjustment of the speed and position of the vehicles in the queue according to the queue desired trajectory; S430, use the distributed model predictive controller to complete the queue stability and dynamic behavior optimization at the dynamic convergence time.

[0055] Specifically, in step S410, according to the cutting-in time, cutting-in position, and the desired trajectory of the leading vehicle output in the first stage, generate the reference trajectories of the cutting-in vehicle and other vehicles in the queue to ensure a smooth cutting-in process and meet the requirements of queue dynamic adjustment.

[0056] Input: optimal cutting-in time , optimal cutting-in position , the desired trajectory of the cutting-in vehicle , the desired trajectories of the queue vehicles ; Define , two state outputs, where represents the system state output predicting the future time steps based on the state at the current moment t; represents the assumed state output, which is obtained by time-shifting the optimal solution of the optimization problem at the previous moment, and its main role is to provide a preliminary estimate of the system behavior.

[0057] The steps of S420 include: S421, allocate independent control tasks for each vehicle, introduce the assumed trajectory state of the vehicle: ; Wherein: and respectively represent the assumed position and speed of vehicle i in queue k at time t; And the queue assumed trajectory state: ; Wherein: respectively represent the assumed position and speed of queue k.

[0058] To decouple the coupling state constraints between any two consecutive vehicles and consecutive queues.

[0059] S423, in-queue cooperative control: Dynamically adjust the queue state error. Through real-time control, adjust the positions and speeds of other vehicles in the queue to reserve sufficient space for the cutting-in vehicle while maintaining the overall dynamic stability of the queue.

[0060] ; Wherein, represents the system state output predicting the future time step lengths based on the state at the current time t, represents the desired trajectory generated within the prediction time domain; The design of the matrix Qi enables the vehicle control system to more precisely correct the position and speed errors, thereby improving the overall stability and queue stability.

[0061] Dynamic safety distance monitoring. Monitor the vehicle spacing in real time to ensure that the distances between all vehicles always meet the minimum safety distance requirements during the adjustment process to ensure operation safety.

[0062] ; Wherein, represents the distance between the ith vehicle and the vehicle in front.

[0063] Matrix is used to measure the relative position error of the vehicle with respect to the vehicle in front of it, helping to maintain the queue stability of the vehicle fleet and ensuring that the relative distances between vehicles do not change significantly due to disturbances.

[0064] Control input term. Apply a penalty to the control input of the vehicle to avoid drastic acceleration and deceleration. The weight matrix Ri is used to balance the magnitude of the control input ; Assumed state error. Used to measure the difference between the actual state of the vehicle and its assumed state, reflecting the ideal trajectory that node i should follow.

[0065] ; S424, Cooperative Control between Queues: Control of the leading vehicle in the queue. The leading vehicle adjusts its own state according to the reference trajectory, provides guidance for the dynamic adjustment of the queue, and ensures that the execution of the cut-in operation meets the expectations.

[0066] ; ; ; ; Control Constraints within the Queue: ; Among them, and respectively represent the system state output and the desired position of the leading vehicle in the k-th queue predicting the future time steps, represents the distance between the k-th queue and the previous queue, The leading vehicle 's state update depends on the state information of other leading vehicles defined in the communication topology . This constraint ensures that the behavior of the leading vehicle is restricted by the communication graph structure and can only obtain information from its neighbors in the communication topology.

[0067] Queue Length Change Constraint: ; Leading Vehicle State Terminal Constraint: ; Among them, represents the position of the k-th queue at time t, and respectively represent the speed of the leading vehicle and the speed of the trailing vehicle in the k-th queue at time t, represents the state of the leading vehicle in the k-th queue in the prediction time domain, represents the assumed state of the leading vehicle in the (k - 1)-th queue, represents the queue length of the (k - 1)-th queue at the prediction terminal.

[0068] S425, Monitoring of the Completion Time of the Cut-in Operation: Record the completion time of the cut-in operation and verify whether it is completed within the preset dynamic convergence time to ensure the efficiency and dynamic performance of the queue recovery.

[0069] S430, Analysis of the Queue State after Adjustment. After the cut-in operation is completed, analyze the queue speed consistency, vehicle spacing, and overall dynamic behavior to ensure that the queue returns to a stable state.

[0070] In another embodiment of the present invention, the distributed model predictive controller includes a cut-in vehicle sub-controller, the optimization objective of the cut-in vehicle sub-controller is to minimize the deviation between the speed and acceleration of the cut-in vehicle, and the distributed model predictive controller includes a cut-in time constraint and a cut-in distance constraint. The cut-in time constraint is used to constrain the cut-in vehicle to reach the optimal cut-in position at the optimal cut-in time, and the cut-in distance constraint is used to constrain the cut-in vehicle to maintain a dynamic safety distance from other vehicles.

[0071] Specifically, the goal of the cut-in vehicle control is to optimize the behavior of the cut-in vehicle by minimizing the deviation between vehicle speed and acceleration to achieve smooth and efficient queue cut-in. Specifically, it includes the following main control functions: ; ; ; in: , and They represent the current state of the cutting vehicle at time t and the predicted future state. velocity, acceleration and system state inputs for each time step; and They represent the current state of the cutting vehicle at time t and the predicted future state. The expected velocity and expected acceleration for each time step.

[0072] The cut-in distance constraints include: ;

[0073] in, , They represent the i-th vehicle and the total number of vehicles in the k-queue, represents the expected distance between the i-th vehicle and the i-1-th vehicle, It represents the deviation between the actual distance and the expected distance. In order to provide a certain degree of flexibility and tolerance to cope with minor disturbances and adjustments that may occur in the vehicle queue, the present invention allows has negative values, thus achieving the desired constant position spacing in the steady state.

[0074] Define each queue to be coupled to adjacent queues by the following constraints:

[0075] in, represents the last car in the k-1th queue, represents the leader car of the k-th queue, Represents the expected spacing between queue k-1 and queue k.

[0076] In another embodiment of the present invention, the communication topology includes a lead-follow tracking strategy between queues, and the lead vehicle of each queue only receives the information of the follow vehicle of the previous queue.

[0077] The communication topology includes: The communication matrix between queues: ; The global adjacency matrix: ; The global Laplacian matrix: ; Wherein, Represents the communication matrix between queue k and queue k-1 Indicates that the follow vehicle of queue k can send information to the lead vehicle of queue k-1.

[0078] Specifically, the present invention studies the unidirectional communication topology in a multi-queue environment. In each queue k, the present invention considers a directed graph , where Is the set of communication edges between different vehicles within queue k. Define a graph Contains a directed spanning tree with all lead vehicles as root nodes, where Is the set of lead vehicle nodes in the queue system, Is the set of communication edges between queues in the queue system to ensure information transfer and sharing between queues.

[0079] Graph The edge Indicates that vehicle i can receive the status information transmitted by vehicle j. The present invention defines As the set of follow vehicles that the i-th follow vehicle in the k-th queue can receive information from, and defines As the set of follow vehicles that the i-th follow vehicle in the k-th queue can send information to.

[0080] In another embodiment of the present invention, the queue stability constraints include vehicle spacing constraints, queue spacing constraints, speed consistency constraints, and queue communication topology constraints; The expressions of the vehicle spacing constraint, the queue spacing constraint, the speed consistency constraint, and the queue communication topology constraint are respectively: ; ; ; ; In the formula: represents the position of the i-th vehicle in the k-th queue; represents the position of the last vehicle in the k-th queue; represents the initial speed; represents the minimum safety distance between any two vehicles; represents the minimum safety distance between any two queues; represents the expected state of the leading vehicle in the k-th queue; represents the state information of other leading vehicles defined in the communication topology.

[0081] Specifically, the queue stability constraint design. To ensure that the queue can quickly return to a stable state after the cut-in operation, the first-stage optimization needs to consider the following queue stability constraints: Dynamic convergence time constraint. Ensure that the dynamic adjustment of the queue is completed within the convergence time as follows: ; Vehicle spacing constraint. Ensure that the distance between any two vehicles in the queue before and after the cut-in operation always satisfies the minimum safety distance: ; Queue spacing constraint: ; Speed consistency constraint. During the cut-in operation, the speeds of the vehicles in the queue need to be kept consistent to prevent excessive speed differences: ; A medium stores executable instructions, and the executable instructions include program codes for executing the multi-vehicle queue cooperative control method based on distributed model predictive control as described in any one of the above when being executed.

[0082] In another embodiment of the present invention, the steps of S300 include the following steps: S310, obtaining the states of the cut-in vehicle and the queue vehicles S311, vehicle state initialization: including the states of the leading vehicle, the following vehicle, the cut-in vehicle (queue), the vehicle prediction state, and the vehicle hypothesis state. Specifically: at moment, initialize the node state: ; where , represents the number of the vehicle in queue k.

[0083] Assume that the vehicles in each queue are in a uniform motion state, and for all vehicles and queues, the initial speeds are the same, that is: ; The initial acceleration of the queue vehicles is 0, i.e.: ; The distance between vehicles satisfies the minimum safety distance, and the distance between queues satisfies the minimum safety distance.

[0084] S312, Input state initialization: Initialize the assumed input sequence and the assumed state sequence for describing future input and state predictions.

[0085] S313, Queue length initialization: Define the queue length of each queue k as the distance between the leading vehicle and the trailing vehicle, i.e. At the initial moment .

[0086] S320, Solve based on the vehicle cut-in optimization model to obtain the optimal cut-in time, optimal cut-in position and the expected queue trajectory of the cut-in vehicle (Phase 1); The cut-in plan includes the cut-in point and the cut-in time. The cut-in point sin is the optimal position for the cut-in vehicle to enter the queue, calculated through Phase 1. The cut-in timing needs to be optimized under the conditions of satisfying the safety distance and queue stability.

[0087] S322, Expected trajectory sequences of the cut-in vehicle and the leading vehicle: For each queue k, generate the expected trajectory of the leading vehicle within the prediction time domain for guiding the cooperative control of the entire queue. The expected trajectory sequence is where: to ensure that the leading vehicle of the queue can maintain synchronization according to the distance between queues. The expected trajectory sequence of the cut-in vehicle is

[0088] In another embodiment of the present invention, the steps of S400 include the following steps: Solve the sub-optimization problem: For each vehicle node i, solve the sub-optimization problem at each moment t to obtain the optimal predicted control input sequence where .

[0089] Required data: (1) Queue cooperation data: The current measured state of the leading vehicle : The actual state of the leading vehicle in queue k at moment t.

[0090] The expected state of the queue : The expected state of the leading vehicle in the k-th queue, generated within the prediction horizon, is used to guide the cooperative control of the queue.

[0091] The state information of the domain queue : According to different strategies, it is used for cooperative optimization between multiple queues to ensure that the queue spacing between leading vehicles meets the requirements.

[0092] (2) Vehicle node data: The currently measured node state : The actual state of node i at time t.

[0093] Predicted state Sequence; Hypothetical state sequence : Includes the hypothetical state sequence of the leading vehicle of each queue; Expected trajectory sequence and : The data in step 2 is the solution result of phase one.

[0094] Solution process: a. Queue state determination: (1) Update the state sequence of the leading vehicle of each queue by applying the optimized input; At time t, apply the optimal control input obtained in the previous step to update the state of the leading vehicle of the queue to ensure that the leading vehicle can stay on the expected trajectory.

[0095] (2) Calculate the optimal state sequence of the leading vehicle of each queue within the prediction horizon ; Initial state According to the optimal control input , calculate the state update of the vehicle within the prediction horizon.

[0096] ; (3) Calculate the hypothetical input sequence for the next step of the leading vehicle of the queue: Within the prediction horizon, update the hypothetical input sequence of the vehicle ; For , the hypothetical input is: ; For , the hypothetical input is: ; (4) Hypothetical state sequence prediction: Based on the hypothetical input sequence calculate the hypothetical state sequence : ; (5) Update queue length: ; In the connected environment, the assumed state sequence and the assumed input sequence are sent to adjacent queue nodes that can receive information. By obtaining information about neighboring vehicles and queues, each node synchronously adjusts and optimizes its control input to ensure the overall effect of collaborative optimization.

[0097] b. Node state update: (1) For each vehicle node i, use to control the state update of the vehicle at the current moment, where .

[0098] (2) Calculate the optimal state sequence within the prediction horizon;

[0099] Initial state Calculate the state update of the vehicle within the prediction horizon according to the optimal control input .

[0100] ; Initial state: ; (3) Calculate the next assumed input sequence: Update the assumed input sequence of the vehicle within the prediction horizon; For , the assumed input is: ; For , the assumed input is: ; (4) Assumed state sequence prediction: The assumed state sequence is: ; c. In the connected environment, the assumed state sequence and the assumed input sequence of the queue node are sent to adjacent queue nodes that can receive information. The assumed state sequence and the assumed input sequence of the vehicle node are sent to adjacent queue nodes that can receive information. By obtaining information about neighboring vehicles and queues, each node synchronously adjusts and optimizes its control input to ensure the overall effect of collaborative optimization.

[0101] d. Repeat steps a~c. At each time step, update the control input and state of queue node k and vehicle node i to ensure that all queues and vehicles maintain synchronization and stability among the queues and vehicles.

[0102] In another embodiment of the present invention, the simulation background and parameter settings are as follows: In this simulation, three queues (Red Queue 1, Blue Queue 2, and Green Queue 3) are set up, with a total of 25 vehicles (9, 7, 7). The cutting-in vehicle starts cutting in from an adjacent lane outside the queue.

[0103] The mass of all vehicles is 1500 kg, the driving delay is 0.7 s, the air resistance coefficient is 1, and the tire radius is 0.35 m. The system time step is 0.1 s, the friction coefficient is 0.01, the transmission efficiency is 0.96, the gravitational acceleration is taken as 9.8 m / s², and the prediction horizon is 20 steps.

[0104] The desired safety distance between vehicles is set to 10 m, the desired distance between queues is 30 m, and the cutting-in vehicle needs to maintain a minimum distance of 10 m from the front and rear vehicles near the insertion point to ensure safe cutting in. , .

[0105] The weight matrices of the remaining distributed model predictive controllers are set as Figure 4 .

[0106] The above parameters are used to test the impact of the cutting-in vehicle on the queue system under dynamic optimization conditions and analyze the performance of different optimization strategies under cooperative control.

[0107] The core objective of the first stage is to gradually transition the cutting-in vehicle from its initial state to an ideal state that meets the cutting-in requirements by dynamically adjusting its speed, position, and power state, laying a foundation for the subsequent formal cutting-in operation. The control strategy in this stage focuses on smoothness and safety, ensuring that the impact of the cutting-in vehicle on the target queue and other vehicles is minimized.

[0108] In terms of speed adjustment, the cutting-in vehicle smoothly accelerates from an initial speed of 17.66 m / s to a target speed of 18 m / s, achieving complete synchronization with the speed of the target queue and avoiding dynamic instability caused by speed differences. In terms of position adjustment, the cutting-in vehicle gradually approaches from a position approximately 200 meters away from the vehicle in front of the target queue to a distance of only 0.5 safety intervals, meeting the safety requirements and leaving enough space for subsequent cutting-in actions.

[0109] The output results of the first stage show that the cutting-in vehicle can achieve dynamic synchronization with the target queue, and its speed, position, and power state can all reach the ideal conditions for cutting in. This not only provides a favorable initial state for the cutting-in operation in the subsequent stage but also verifies the effectiveness and robustness of the control strategy in dynamic adjustment.

[0110] From Figure 2It can be observed that after the dynamic adjustment in the 6 seconds before the cut-in vehicle completes, its speed has gradually approached the speed of the target queue (20 m / s). In the second stage, the cut-in vehicle further fine-tunes its speed and finally completely matches the speed of the target queue. In the second stage, the cut-in vehicle gradually increases from a speed slightly lower than the target queue (about 19.5 m / s) to the fully synchronized 20 m / s. This process shows smooth acceleration characteristics without violent fluctuations.

[0111] Dynamic adaptation of platoon vehicles: The target queue (platoon 2) maintained a high degree of dynamic stability during the cut-in process, showing only minimal speed fluctuations (less than 0.5 m / s). This indicates that the impact of the cut-in vehicle's operation on the speed perturbation of the queue is controlled to a minimum. The speed curves of the other queues (platoon 1 and platoon 3) remained completely stable throughout the second stage, indicating that the impact of the cut-in operation is limited within the target queue, further verifying the isolation and stability of the control strategy for the queue system.

[0112] From Figure 3 It can be seen that in the second stage, the cut-in vehicle gradually completes the cut-in action from the preset position into the target queue. Position change of the cut-in vehicle: At the 6th second, the cut-in vehicle is 0.5 safety spacings away from the vehicle in front of the target queue. In the subsequent second stage, the cut-in vehicle gradually integrates into the target queue through precise longitudinal position adjustment and finally stabilizes between two target queue vehicles, forming a uniform queue spacing.

[0113] Coordinated adjustment of platoon vehicles: The target queue vehicles actively adjust their positions to provide enough space for the cut-in vehicle. This coordinated adaptation behavior effectively avoids congestion within the queue caused by the cut-in operation.

[0114] Stability of other queues: The other queues (platoon 1 and platoon 3) still maintain completely stable trajectories in the second stage, further proving that the dynamic impact of the cut-in operation is well confined within the target queue range.

[0115] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structural transformation made under the concept of the present invention by using the content of the specification and drawings of the present invention, or directly / indirectly applied in other related technical fields, is included in the patent protection scope of the present invention.

Claims

1. A multi-vehicle queue cooperative control method based on dynamic model predictive control, characterized in that It includes the following steps: Establish a vehicle dynamics model, a vehicle motion state equation, a communication topology structure, a queue state space model, and a multi-queue system model; Establish a two-stage control model, where the two-stage control model includes a vehicle cut-in optimization model and a distributed model predictive controller; the vehicle cut-in optimization model includes an objective function and queue stability constraints; the objective function includes a cut-in vehicle cost function and a queue cost function for measuring the adjustment costs required for the cut-in vehicle and the queue when the cut-in vehicle reaches the optimal cut-in position; the optimization direction is to minimize the adjustment cost; Obtain the states of the cut-in vehicle and the queue vehicles, and solve based on the vehicle cut-in optimization model to obtain the optimal cut-in time, the optimal cut-in position of the cut-in vehicle, and the queue expected trajectory; Use the distributed model predictive controller to perform dynamic control on the vehicle according to the queue expected trajectory, the optimal cut-in time, and the optimal cut-in position of the cut-in vehicle, so that the queue quickly returns to the dynamic stable state.

2. The multi-vehicle queue cooperative control method based on dynamic model predictive control according to claim 1, characterized in that, The vehicle dynamics model includes the following expressions: ； Air resistance: ; Rolling resistance: ; In the formula: In the formula: represents the air density; represents the gravitational constant; , , , and respectively represent the mass, speed, rolling resistance coefficient, frontal surface area of the vehicle, and air drag coefficient of the i-th vehicle in the k-th queue; , , and respectively represent the position, speed, acceleration, and driving force or braking torque of the i-th vehicle in the k-th queue at time t.

3. The multi-vehicle queue cooperative control method based on dynamic model predictive control according to claim 1, characterized in that The objective function of the vehicle cut-in optimization model includes: ； Wherein: represents the adjustment cost required for the cutting-in vehicle and the queue when the cutting-in vehicle reaches the optimal cutting-in position; represents the dynamic adjustment cost of the cutting-in operation to other vehicles in the queue; P is the feasible solution domain of the optimal cutting-in position, is the energy consumption at each possible cutting-in point, and the solution with the minimum energy consumption is found from all results .

4. The multi-vehicle queue cooperative control method based on dynamic model predictive control according to claim 1, characterized in that, The expression of the queue cost function includes: ； In the formula: is the time step for the queue to complete the reserved space, is the time domain for the queue to complete the reserved space, and , is the time for the queue to complete the reserved space; represents the air density; represents the gravitational constant; , , , and respectively represent the mass, speed, rolling resistance coefficient, frontal surface area of the vehicle, and air resistance coefficient of the i-th vehicle in the k-th queue, is the unit discrete time interval is the longitudinal displacement distance of the vehicle within it.

5. The multi-vehicle queue cooperative control method based on dynamic model predictive control according to claim 1, characterized in that The expression of the cut-in vehicle cost function includes: ； Wherein, is the time step when the cutting vehicle reaches the optimal cutting point, is the optimal discrete time interval when the cutting vehicle reaches, and within this interval is the longitudinal displacement distance of the vehicle, ; represents the air density; represents the gravitational constant; , , , , and respectively represent the mass, speed, rolling resistance coefficient, front surface area, quantity and air resistance coefficient of the cutting vehicle.

6. The multi-vehicle queue cooperative control method based on dynamic model predictive control according to claim 1, wherein The step of using the distributed model predictive controller to perform dynamic control on the vehicle according to the queue expected trajectory, the optimal cut-in time, and the optimal cut-in position of the cut-in vehicle, so that the queue quickly returns to the dynamic stable state, includes the following steps; Use the distributed model predictive controller to perform trajectory planning and control on the cut-in vehicle according to the optimal cut-in time and the optimal cut-in position of the cut-in vehicle; Use the distributed model predictive controller to complete the adjustment of the speeds and positions of the vehicles in the queue according to the queue expected trajectory; Use the distributed model predictive controller to complete the optimization of queue stability and dynamic behavior within the dynamic convergence time.

7. The multi-vehicle queue cooperative control method based on dynamic model predictive control according to claim 6, characterized in that The distributed model predictive controller includes a cut-in vehicle sub-controller, the optimization objective of the cut-in vehicle sub-controller is to minimize the deviation between the speed and acceleration of the cut-in vehicle, the distributed model predictive controller includes a cut-in time constraint and a cut-in distance constraint, the cut-in time constraint is used to constrain the cut-in vehicle to reach the optimal cut-in position at the optimal cut-in time, and the cut-in distance constraint is used to constrain the cut-in vehicle to maintain a dynamic safety distance from other vehicles.

8. The multi-vehicle queue cooperative control method based on dynamic model predictive control according to claim 6, characterized in that The communication topology structure includes the lead vehicle-following vehicle tracking strategy between queues, and the lead vehicle of each queue only receives the information of the rear vehicle of the previous queue.

9. The multi-vehicle queue cooperative control method based on dynamic model predictive control according to any one of claims 1-8, characterized in that The queue stability constraints include vehicle spacing constraints, queue spacing constraints, speed consistency constraints, and queue communication topology constraints; The expressions of the vehicle spacing constraint, the queue spacing constraint, the speed consistency constraint, and the queue communication topology constraint are respectively: ； ； ； Wherein: represents the position of the k th vehicle in the i th queue; represents the position of the leading vehicle in the kth queue; represents the position of the trailing vehicle in the previous queue; represents the position of the k th vehicle at time t in the i th queue; represents the initial speed; represents the minimum safe distance between any two vehicles; represents the minimum safe distance between any two queues; represents the expected state of the leading vehicle in the kth queue; represents the state information of other leading vehicles defined in the communication topology.

10. A medium, characterized in that, Stores executable instructions, and the executable instructions contain program codes for executing the multi-vehicle queue cooperative control method based on distributed model predictive control as described in any one of claims 1-9 when being executed.

Images