A fine trajectory level guidance and control method for transportation vehicle zero-waste

Abstract

The application relates to a fine trajectory-level guidance and control method for zero-waste of transport vehicles, comprising the following steps: collecting real-time vehicle running state information, grouping all vehicles in combination with road topological structures, so that each intersection control area corresponds to a group of vehicles to be controlled; for each group, a trajectory-level driving control optimization model is established, including an objective function, decision variables and constraint conditions; a customized Benders decomposition algorithm is used to optimize and solve each trajectory-level optimization model, and a plurality of corresponding traffic decision results are obtained; the traffic decision results are converted into lateral lane control instructions and longitudinal speed guidance instructions to control the driving trajectories of the vehicles in each group. Compared with the prior art, the application can effectively reduce the invalid driving time caused by multi-vehicle conflicts and unreasonable lane selection under the premise of ensuring intersection traffic safety, improve the operation efficiency of road transport vehicles and reduce the energy consumption of vehicle driving.

Description

Technical Field

[0001] This invention relates to the field of autonomous vehicle scheduling and control technology, and in particular to a refined trajectory-level guidance and control method for zero-waste transportation vehicles. Background Technology

[0002] With the continuous development of autonomous driving technology, vehicle-road cooperative technology, and vehicle intelligent control technology, road transport vehicles, relying on high-precision positioning, multi-source sensor fusion, and refined control capabilities, can achieve autonomous perception, autonomous decision-making, and precise execution in complex road environments. Especially in closed or semi-closed operating scenarios, the driving control system of autonomous vehicles can precisely control behaviors such as lane changing, overtaking, and speed adjustment, thus providing a technical foundation for achieving a more efficient road operation mode.

[0003] However, in real-world applications, when multiple road transport vehicles approach simultaneously at intersections or critical conflict zones, they are prone to competing for passage. If the vehicle's driving control system relies solely on local collision avoidance logic for responsive decision-making, it often leads to vehicles slowing down, stopping, or frequently adjusting their speed before reaching the conflict zone, resulting in operational efficiency losses. Furthermore, in the absence of a unified control strategy, vehicles may experience congestion or operational disruptions in localized sections due to inappropriate lane selection or uncoordinated overtaking strategies.

[0004] This will result in additional waiting time and unnecessary deceleration time for vehicles, creating a significant amount of wasted operating time, or "waste." "Waste" not only reduces the efficiency of individual vehicles but also weakens the technological advantages of autonomous vehicles in precise control. From the perspective of road transport vehicle driving control systems, existing technologies typically handle lane-changing control, overtaking control, and speed control separately, lacking a mechanism for unified optimization of the passage sequence and control variables of multiple vehicles at the trajectory level. This makes it difficult to achieve a systematic zero-waste control goal at intersections.

[0005] In summary, facing the problems of mutual congestion caused by road transport vehicles competing for road space and time resources, and the decline in local operating efficiency caused by unreasonable lane-changing decisions, existing driving control methods for road transport vehicles are unable to simultaneously coordinate vehicle overtaking decisions, lane selection, and speed control. They cannot obtain executable, refined trajectory control commands, which is not conducive to improving the stability and continuity of road transport vehicles. As a result, the problem of "empty consumption" of road transport vehicles still exists, which seriously restricts the transportation efficiency of road transport vehicles and increases unnecessary vehicle driving energy consumption. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a refined trajectory-level guidance and control method for zero-waste transportation vehicles. This method can achieve coordinated optimization of overtaking, lane changing and speed in multi-vehicle conflict scenarios such as intersections, effectively reduce waste caused by ineffective waiting and frequent deceleration, and improve the operating efficiency and driving stability of road transport vehicles.

[0007] The objective of this invention can be achieved through the following technical solution: a refined trajectory-level guidance and control method for zero-waste transportation vehicles, comprising the following steps:

[0008] S1. Real-time collection of the operating status information of each vehicle, combined with the road topology, grouping all vehicles so that each intersection control area corresponds to a group of vehicles to be controlled. This group of vehicles to be controlled will arrive at the same intersection within a preset time window and have potential spatiotemporal conflict relationships.

[0009] S2. For each group, establish a trajectory-level driving control optimization model, including objective function, decision variables and constraints. The objective function aims to minimize the total travel time of all vehicles in the group through the intersection.

[0010] S3. The customized Benders decomposition algorithm is used to optimize and solve each trajectory-level optimization model separately, and the corresponding multiple traffic decision results are output.

[0011] S4. Convert the traffic decision results into lateral lane control commands and longitudinal speed guidance commands to control the driving trajectory of vehicles in each group accordingly.

[0012] Furthermore, the vehicle's operating status information in S1 includes the vehicle's current position, current lane, speed, acceleration, and information about the intersection control area it is about to enter.

[0013] Furthermore, the objective function in S2 is specifically:

[0014]

[0015] in, Indicates vehicle The travel time consumed within the intersection control area is used to describe the level of vehicle idle time at the intersection. The objective function value, Indicates vehicle The moment of entering the intersection, For vehicles The moment of arrival at the entrance to the intersection control area.

[0016] Furthermore, the decision variables in S2 include lateral decision variables. Longitudinal decision variables and priority passage decision variables The horizontal decision variables Specifically, it is a 0-1 discrete variable representing whether a vehicle performs a lane-changing or overtaking maneuver. Indicates vehicle Perform a lane change or overtaking maneuver. Indicates vehicle Maintain your current lane;

[0017] The vertical decision variables Specifically, vehicles Longitudinal speed control variables before entering and during passage through the intersection;

[0018] Priority decision variables Specifically, it refers to vehicles. With vehicles j The order of passage between them Indicates vehicle Superior to vehicles Passage Indicates vehicle Superior to vehicles Passage is permitted.

[0019] Furthermore, the constraints in S2 include vehicle kinematic constraints, time continuity constraints, conflict resolution constraints, and lateral control and priority relationship constraints. Among them, vehicle kinematic constraints are used to limit the range of values ​​for the average speed of the vehicle on the road segment.

[0020] Time continuity constraints are used to ensure consistency between the calculation of vehicle passage time at intersections and speed decisions;

[0021] Conflict resolution constraints are used to ensure that only one vehicle is allowed to occupy the same conflict zone at any given time.

[0022] Lateral control and associated constraints are used to ensure that a vehicle can only obtain right-of-way when it is performing a lane change or when it has legal driving conditions.

[0023] Furthermore, the time continuity constraint is specifically constructed through the following process:

[0024] The nonlinear relationship between the travel time of vehicle i through the intersection control area and its speed is determined as follows: ,in, This refers to a group of vehicles within the control area of ​​the same intersection.

[0025] Introducing continuous auxiliary variables Used to indicate vehicles Within the intersection control zone, the time for vehicles to enter the intersection is:

[0026]

[0027] Vehicle The permissible speed range within the intersection control zone is divided into: 3 adjacent discrete sub-intervals:

[0028]

[0029] in, Indicates the minimum speed limit for that section of road. Indicates the maximum speed limit for this section of road. , , Each subinterval corresponds to a velocity value segment, which is used to construct a piecewise linear approximation function;

[0030] To indicate which discrete interval the speed of vehicle r falls within, a 0-1 variable for interval selection is introduced. And add a range selection uniqueness constraint to ensure vehicle Select only one speed range at any given time:

[0031]

[0032] In each speed range Inside, to An approximation using a linear function:

[0033]

[0034] in, and For in the interval The linearization coefficients, pre-calculated, are used to approximate the variation trend of the original nonlinear function within this interval;

[0035] Piecewise linearized constraint expression, combined with interval selection variables Construct the following linear constraints to describe and Piecewise linear relationship between them:

[0036]

[0037]

[0038] To ensure speed variables Consistent with the selected interval, add the following interval constraints:

[0039]

[0040] in, k Number the sub-interval.

[0041] Furthermore, the solution process of the customized Benders decomposition algorithm in S3 includes:

[0042] S31. Construct the master problem, which only includes the first objective function corresponding to the horizontal decision variables and the master problem constraints;

[0043] S32. After the main problem is given discrete decision, a sub-problem is constructed. The sub-problem contains a corresponding second objective function and sub-problem constraints, which are used to solve for the optimal solution of vehicle longitudinal speed and travel time.

[0044] S33. Based on the optimal solution of vehicle longitudinal speed and travel time, generate the corresponding Benders cut and feed it back to the main problem. If the subproblem is feasible under the given discrete decision of the main problem, construct the optimality cut based on the dual information of the subproblem.

[0045] Otherwise, construct a feasibility cut to exclude driving control decision schemes that currently lead to conflict or fail to meet safety constraints;

[0046] S34. Repeat steps S31 to S33 until the difference between the target values ​​of the main problem and the sub-problems is less than a preset threshold, or the maximum number of iterations is reached, and output the decision result of the best driving control scheme.

[0047] Furthermore, the first objective function is specifically:

[0048]

[0049] in, This is a lower bound estimate of the optimal objective value for the subproblem;

[0050] The constraints of the main problem include:

[0051]

[0052]

[0053] in, This represents the set of Benders cuts that have been generated. and For the first The cutting coefficients generated in the next iteration.

[0054] Furthermore, the second objective function is specifically:

[0055]

[0056] The sub-problem constraints include time continuity constraints and conflict resolution constraints.

[0057] Furthermore, the specific process of S33 is as follows:

[0058] Based on the solutions to the subproblems, the corresponding Benders cuts are generated and fed back to the main problem. If the subproblems fall within the given discrete decision... If the following is feasible, then construct an optimality cut based on the dual information of the subproblem:

[0059]

[0060] If the subproblem is in a given discrete decision If the following is not feasible, then construct a feasible section:

[0061]

[0062] in, Indicates with vehicles The relevant infeasibility indicator coefficient is used to describe the degree to which the vehicle participates in currently infeasible driving control decision-making schemes; The upper bound constant is determined based on the set of infeasible constraints of the subproblems. This constant is used to limit the recurrence of discrete decision combinations that lead to infeasibility. Feasibility cut is used to ensure that the main problem does not generate discrete decision combinations that are equivalent or similar to the current infeasible driving control decision scheme in subsequent iterations.

[0063] Compared with the prior art, the present invention has the following advantages:

[0064] This invention first groups all vehicles, ensuring that each intersection control area corresponds to a group of vehicles to be controlled. Then, for each group, a trajectory-level driving control optimization model is established. A customized Benders decomposition algorithm is then used to optimize and solve each trajectory-level optimization model, outputting multiple corresponding traffic decision results. Finally, these traffic decision results are converted into lateral lane control commands and longitudinal speed guidance commands to control the driving trajectories of vehicles within each group. This enables coordinated optimization of lateral lane control and longitudinal speed control, minimizing invalid travel time within the intersection control area, effectively reducing invalid travel time caused by multi-vehicle conflicts and unreasonable lane selection, and improving the operating efficiency and driving stability of road transport vehicles.

[0065] This invention uses each intersection control area as a decision-making unit. For each group corresponding to the intersection control area, a corresponding trajectory-level driving control optimization model is established. That is, multiple vehicles entering the same conflict area are uniformly modeled and collaboratively optimized. By solving each trajectory-level optimization model, multiple traffic decision results corresponding to different groups can be obtained. Using the lateral lane control command and longitudinal speed guidance command converted from the traffic decision results, the driving trajectory of vehicles in each group can be controlled accordingly. This allows vehicles to pre-determine traffic priority relationships and trajectory-level control decisions before entering the conflict area, thereby actively resolving potential conflicts and avoiding multiple vehicles waiting for each other or causing congestion.

[0066] This invention, when establishing a trajectory-level driving control optimization model, takes minimizing ineffective travel time as the optimization objective. It unifies lane change control, overtaking control, and longitudinal speed control into a unified model, moving away from isolated lateral decision-making and longitudinal adjustment. Instead, it models traffic priority relationships, integrating discrete lateral control decision variables with continuous longitudinal speed decision variables for coordinated optimization. This allows vehicles to acquire priority through reasonable lateral control strategies while avoiding unnecessary stops or sharp decelerations through precise speed adjustments. This shortens vehicle travel time within the intersection control area, effectively reducing localized congestion caused by unreasonable lane selection or speed fluctuations, and significantly reducing ineffective travel time within the intersection control area.

[0067] When establishing a trajectory-level driving control optimization model, this invention considers the nonlinear relationship between the vehicle's travel time through the intersection control area and its speed, which cannot be directly solved using linear or mixed-integer linear programming. Therefore, this invention linearizes the aforementioned nonlinear relationship to obtain a piecewise linearized constraint expression. Through linearization, the reciprocal relationship between the vehicle's longitudinal control speed and travel time is transformed into a piecewise linear form, so that the entire trajectory-level driving control optimization model is uniformly represented as a mixed-integer linear programming model, thereby meeting the real-time driving control decision requirements.

[0068] This invention addresses the issues of large scale and tight coupling between discrete and continuous variables in the established trajectory-level driving control optimization model. Therefore, a customized Benders decomposition algorithm is constructed to solve the problem, which decomposes the original problem into a main problem and sub-problems. By iteratively generating optimal cuts and feasible cuts, the overall optimal driving control strategy is gradually approximated. This approach can significantly reduce computational complexity while ensuring solution accuracy, which is beneficial for meeting the online decision-making and control requirements in practical engineering applications. Attached Figure Description

[0069] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0070] Figure 2 This is a schematic diagram of the application process in an embodiment. Detailed Implementation

[0071] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0072] Example

[0073] To address the shortcomings of existing road transport vehicle driving control methods in effectively handling congestion caused by multiple vehicles competing for road space and time resources in conflict scenarios such as intersections, and the resulting decrease in operational efficiency due to unreasonable lane-changing decisions, this invention proposes a refined trajectory-level guidance and control method for zero-waste transport vehicles. Figure 1 As shown, it includes the following steps:

[0074] S1. Real-time collection of the operating status information of each vehicle, combined with the road topology, grouping all vehicles so that each intersection control area corresponds to a group of vehicles to be controlled. This group of vehicles to be controlled will arrive at the same intersection within a preset time window and have potential spatiotemporal conflict relationships.

[0075] S2. For each group, establish a trajectory-level driving control optimization model, including objective function, decision variables and constraints. The objective function aims to minimize the total travel time of all vehicles in the group through the intersection.

[0076] S3. The customized Benders decomposition algorithm is used to optimize and solve each trajectory-level optimization model separately, and the corresponding multiple traffic decision results are output.

[0077] S4. Convert the traffic decision results into lateral lane control commands and longitudinal speed guidance commands to control the driving trajectory of vehicles in each group accordingly.

[0078] This embodiment applies the above method to the driving control system of an existing road transport vehicle. The road transport vehicle is equipped with a variety of on-board sensors and an execution control unit. The on-board sensors and the execution control unit are respectively connected to the driving control system. The various on-board sensors are used to acquire the vehicle's operating status information in real time, while the execution control unit is used to execute the instructions received from the driving control system.

[0079] The application process of this embodiment is as follows: Figure 2 As shown, it specifically includes:

[0080] Step 1: Acquire the vehicle's operating status information in real time and send the acquired information to the decision module of the road transport vehicle driving control system. The vehicle's operating status information includes at least: the vehicle's current position coordinates, the current lane number, the driving speed, the acceleration, and the control area identifier of the intersection to be entered.

[0081] Step 2: The driving control system of the road transport vehicle identifies the intersection control area that the vehicle is about to enter based on the road topology and the current position of the vehicle, and divides the relevant vehicles into the same control group according to the conflict relationship. That is, based on the potential spatiotemporal conflict relationship between vehicles, vehicles that may enter the same conflict area at the same time within the preset prediction time window are grouped to form an intersection control group.

[0082] Step 3: For each intersection control group, establish a corresponding refined trajectory-level driving control optimization model to determine the traffic priority relationship between vehicles, lateral lane change decisions, and longitudinal speed decision commands. This trajectory-level driving control optimization model includes objective functions, decision variables, and constraints.

[0083] This embodiment first defines the model parameters, intermediate variables, and decision variables. The model parameters include: This refers to a group of road transport vehicles within the same intersection control area. This represents any two different road transport vehicles; Represents the set of conflict zones. Indicates a specific area of ​​conflict; Represents a sufficiently large positive number for linearization processing; This is the safe time interval for the intersection; For vehicles The time of arrival at the entrance to the intersection control area; Indicates the minimum speed limit for that section of road. Indicates the maximum speed limit for this section of road;

[0084] The intermediate variables include: The objective function value; Indicates vehicle The moment of entering the intersection;

[0085] The decision variables include: The 0-1 discrete variables represent the vehicle's lateral control decision. Indicates vehicle Perform lane change or overtaking control. Indicates vehicle Maintain your current lane;

[0086] This represents the target longitudinal speed control variable for the vehicle before entering the intersection and during its passage through the intersection.

[0087] Indicates vehicle Superior to vehicles Passage Indicates vehicle Superior to vehicles Passage is permitted.

[0088] Through the above decision variables, the coordinated optimization of vehicle lateral control, longitudinal speed control, and conflict priority control can be achieved.

[0089] The objective function is then constructed to minimize the total travel time of all vehicles through the intersection, thereby reducing vehicle waiting time and deceleration time at the intersection. This objective reflects zero-waste-oriented control at the driving control system level. The objective function takes the following form:

[0090]

[0091] in, Indicates vehicle Travel time consumed within the intersection area. Target time describes the ineffective travel time incurred by vehicles due to conflicts or improper speed control.

[0092] Then set constraints, including:

[0093] Vehicle kinematic constraints are used to limit the range of values ​​for the average speed of a vehicle traveling on a road segment:

[0094]

[0095] Time continuity constraints are used to ensure consistency between the calculated time of vehicle passage through the intersection and the longitudinal control variables.

[0096]

[0097] Conflict resolution constraints are used to ensure that within the same conflict zone... The two cars inside Only one vehicle is allowed to occupy the space at a time.

[0098]

[0099]

[0100] Lateral control and priority relationship constraints are used to ensure that a vehicle only gains priority when performing a lane change or when it has legal driving conditions:

[0101]

[0102] It should be noted that this scheme takes into account the time continuity constraint, including the vehicle The non-linear relationship between travel time and speed within an intersection control zone cannot be directly solved using linear or mixed-integer linear programming. Therefore, this solution linearizes this non-linear relationship, as detailed below:

[0103] Introducing continuous auxiliary variables Used to indicate vehicles For passage time within the intersection control area, the time when a vehicle enters the intersection is rewritten as follows:

[0104]

[0105] Vehicle The permissible speed range within the intersection control zone is divided into: 3 adjacent discrete sub-intervals:

[0106]

[0107] in, , , Each subinterval corresponds to a velocity value segment, which is used to construct a piecewise linear approximation function.

[0108] To indicate which discrete interval the speed of vehicle r falls within, a 0-1 variable for interval selection is introduced. And add a range selection uniqueness constraint to ensure vehicle Select only one speed range at any given time:

[0109]

[0110] Piecewise linear function approximation construction. In each velocity range... Inside, to An approximation using a linear function:

[0111]

[0112] in, and For in the interval The linearization coefficients, calculated in advance, are used to approximate the variation trend of the original nonlinear function within this interval.

[0113] Piecewise linearized constraint expression. Combined with interval selection variables. Construct the following linear constraints to describe and Piecewise linear relationship between them:

[0114]

[0115]

[0116] To ensure speed variables Consistent with the selected interval, add the following interval constraints:

[0117]

[0118] Through the above processing, the reciprocal relationship between the vehicle's longitudinal control speed and travel time can be transformed into a piecewise linear form, so that the entire trajectory-level driving control optimization model can be uniformly represented as a mixed integer linear programming model, thereby meeting the real-time driving control decision requirements.

[0119] Step 4: To address the issues of large scale and tight coupling between discrete and continuous variables in the trajectory-level driving control optimization model, this solution constructs a customized Benders decomposition algorithm to solve the problem. This algorithm decomposes the original problem into a main problem and sub-problems, and iteratively generates optimal cuts and feasibility cuts to gradually approach the overall optimal driving control strategy. The steps include:

[0120] Step 4.1: Construct the master problem to determine the vehicle lateral control decision variables, which includes the corresponding first objective function and constraints.

[0121] The first objective function is defined as:

[0122]

[0123] in, This is a lower bound estimate of the optimal objective value for the subproblem.

[0124] The main problem constraints include:

[0125]

[0126]

[0127] in, This represents the set of Benders cuts that have been generated. and For the first The cut coefficients generated in the next iteration.

[0128] Step 4.2: Given the main problem, make discrete decisions Next, we construct a subproblem to optimize the longitudinal velocity control variable and the time variable.

[0129] The second objective function corresponding to the subproblem is defined as:

[0130]

[0131] The constraints of the subproblem are:

[0132]

[0133]

[0134]

[0135]

[0136]

[0137]

[0138]

[0139] Step 4.3: Based on the solution results of the subproblems, generate the corresponding Benders cuts and feed them back to the main problem. If the subproblems are within the given discrete decision... The following is feasible: construct an optimality cut based on the dual information of the subproblems.

[0140]

[0141] If the subproblem is in a given discrete decision If the current approach is not feasible, a feasibility segment is constructed to exclude driving control schemes that cause conflict or fail to meet safety constraints.

[0142]

[0143] in Indicates with vehicles The relevant infeasibility indicator coefficient is used to describe the extent to which the vehicle participates in currently infeasible driving control schemes; This is an upper bound constant determined based on the set of infeasible constraints of the subproblems, used to restrict the recurrence of discrete decision combinations that lead to infeasibility. By utilizing feasibility cut, it can be ensured that in subsequent iterations, the main problem no longer generates discrete decision combinations equivalent to or similar to the currently infeasible driving control scheme.

[0144] Step 4.4: Repeat steps 4.1 to 4.3 until the difference between the target values ​​of the main problem and the subproblems is less than a preset threshold, or the maximum number of iterations is reached. The algorithm then terminates and outputs the decision result of the optimal driving control scheme.

[0145] The customized Benders decomposition algorithm described above enables efficient solution and engineering feasibility for trajectory-level driving control problems in multi-vehicle conflict scenarios. Addressing the computational complexity caused by the high coupling between lateral discrete control variables and longitudinal continuous control variables in multi-vehicle, multi-conflict intersection conditions, this scheme constructs a customized Benders decomposition algorithm. This algorithm decomposes the original mixed-integer optimization problem into a main problem responsible for lateral control and traffic priority decisions, and a sub-problem responsible for longitudinal speed control and time optimization. Through an iterative generation mechanism of optimality cut and feasibility cut, efficient approximation and convergence of the multi-vehicle cooperative control problem are achieved. This decomposition strategy significantly reduces computational complexity while ensuring solution accuracy, enabling the proposed trajectory-level driving control method to possess real-time computing capabilities and engineering deployment feasibility, meeting the online decision-making requirements of road transport vehicle driving control systems in complex road scenarios.

[0146] Step 5: After the optimization solution is completed, the driving control system converts the obtained traffic decision results into lateral lane control commands and longitudinal speed guidance commands. Specifically, it converts the obtained lateral lane control decisions into lane change or lane keeping control commands and converts the longitudinal speed control variables into target speed trajectories, and sends them to the execution control unit of the road transport vehicle to generate an executable and refined driving trajectory.

[0147] Step 6: The vehicle execution control unit performs lateral and longitudinal control operations based on the received lateral lane control commands and longitudinal speed guidance commands, and feeds back the execution status to the driving control system in real time to update the vehicle's operating status and support subsequent rolling optimization.

[0148] After implementing the methods described in steps 1 to 6 above, the real-time operating status information of vehicles is first acquired, and the set of vehicles about to enter the same intersection control area is identified. Then, with the optimization objective of minimizing the vehicle's travel time within the control area, the vehicle lane change or overtaking decisions and speed control variables are jointly optimized. The nonlinear relationship between travel time and speed is processed using a piecewise linearization method, transforming the original problem into a mixed-integer linear programming model. Furthermore, a customized Benders decomposition algorithm is used for efficient solution, achieving decoupling optimization of traffic priority decisions and speed control variables. The optimization results are then converted into lane control commands and speed guidance commands, which are implemented by the vehicle execution control unit. Through this technical solution, while ensuring intersection traffic safety, the ineffective travel time caused by multi-vehicle conflicts and unreasonable lane selection can be effectively reduced, improving the operating efficiency and driving stability of road transport vehicles.

[0149] In practical applications, when the road transport vehicle driving control system identifies that a vehicle is about to enter the intersection control area, it performs unified conflict relationship identification and group control on vehicles entering the same control area. Based on the traffic priority relationship modeling method, it coordinates and optimizes the vehicle's lane change decision, overtaking control behavior and longitudinal driving speed, so that the traffic sequence determination and trajectory-level control command generation are completed before the vehicle actually enters the intersection control area.

[0150] In this embodiment, a refined trajectory-level driving control model for the intersection control area is constructed in step 3, and the nonlinear functional relationship between vehicle travel time and speed is piecewise linearized. This allows the vehicle to optimize and compress its travel time within the control area while satisfying speed constraints, conflict zone safety time interval constraints, and traffic priority constraints. This approach reduces unnecessary travel time caused by frequent deceleration, waiting, or stopping, improving the continuity and smoothness of vehicle passage through intersections.

[0151] In this embodiment, the customized Benders decomposition algorithm described in step 4 decomposes and optimizes the discrete lateral control decisions involving lane changing and overtaking behavior, along with continuous longitudinal control variables such as vehicle speed and travel time. Through an iterative solution mechanism of the main problem and sub-problems, even in complex intersection scenarios with multiple vehicles and conflict zones, a stable and feasible traffic priority scheme and trajectory-level control results can still be obtained within an acceptable computational time, enabling the driving control method to have real-time online application capabilities.

[0152] Under the above implementation method, the passage sequence of vehicles within the intersection control area is rationally coordinated, effectively avoiding mutual congestion caused by multiple vehicles entering the same conflict area simultaneously, and significantly reducing the additional waiting time caused by conflict avoidance. Simultaneously, by converting the priority decision results into clear lane control and speed control commands, and having the vehicle execution control unit perform lateral and longitudinal control, vehicles can pass through the intersection control area with a continuous and smooth trajectory, reducing local congestion caused by unreasonable lane selection or speed fluctuations, improving vehicle operational stability, and thus increasing transportation efficiency and reducing driving energy consumption.

[0153] In summary, the trajectory-level guidance and control method for road transport vehicles described in this specific embodiment effectively reduces the ineffective travel time of vehicles within the control area and adjacent road segments, improves single-vehicle operating efficiency and road transport efficiency, and fully leverages the technical advantages of autonomous vehicles in precise control and collaborative execution, demonstrating good engineering implementation results and application value. This solution uses the intersection and its internal conflict zone as the core unit of driving control, upgrading the traditional vehicle traffic control method based on path rules or fixed priority strategies to a trajectory-level collaborative control method oriented towards conflict zones. It fully utilizes the high-precision control capabilities of autonomous vehicles in longitudinal speed adjustment and lateral lane changing, using the intersection control area as the decision-making unit to uniformly model and collaboratively optimize multiple vehicles entering the same conflict zone, completing the determination of traffic priority relationships and trajectory-level control decisions before vehicles enter the conflict zone. Compared to existing control methods that rely on post-event avoidance or static priority rules, this method can proactively resolve potential conflicts at the driving control level, avoid multiple vehicles waiting for each other or blocking each other, reduce the invalid driving time caused by vehicles waiting, slowing down or stopping from the source, improve the stability and continuity of road transport vehicle operation, and thus improve transportation efficiency and reduce driving energy consumption.

[0154] The above are merely preferred embodiments of the present invention and do not constitute any limitation on the present invention. Any equivalent substitutions or modifications made by those skilled in the art to the technical solutions and content disclosed in the present invention without departing from the scope of the present invention shall be deemed to have remained within the protection scope of the present invention.

Claims

1. A refined trajectory-level guidance and control method for zero-waste transportation vehicles, characterized in that, Includes the following steps: S1. Real-time collection of the operating status information of each vehicle, combined with the road topology, grouping all vehicles so that each intersection control area corresponds to a group of vehicles to be controlled. This group of vehicles to be controlled will arrive at the same intersection within a preset time window and have potential spatiotemporal conflict relationships. S2. For each group, establish a trajectory-level driving control optimization model, including objective function, decision variables and constraints. The objective function aims to minimize the total travel time of all vehicles in the group through the intersection. S3. The customized Benders decomposition algorithm is used to optimize and solve each trajectory-level optimization model separately, and the corresponding multiple traffic decision results are output. S4. Convert the traffic decision results into lateral lane control instructions and longitudinal speed guidance instructions to control the driving trajectory of vehicles in each group accordingly. The lateral lane control instructions include lane change instructions and overtaking instructions. The solution process of the customized Benders decomposition algorithm in S3 includes: S31. Construct the master problem, which only includes the first objective function corresponding to the horizontal decision variables and the master problem constraints; S32. After the main problem is given discrete decision, a sub-problem is constructed. The sub-problem contains a corresponding second objective function and sub-problem constraints, which are used to solve for the optimal solution of vehicle longitudinal speed and travel time. S33. Based on the optimal solution of vehicle longitudinal speed and travel time, generate the corresponding Benders cut and feed it back to the main problem. If the subproblem is feasible under the given discrete decision of the main problem, construct the optimality cut based on the dual information of the subproblem. Otherwise, construct a feasibility cut to exclude driving control decision schemes that currently lead to conflict or fail to meet safety constraints; S34. Repeat steps S31 to S33 until the difference between the target values ​​of the main problem and the sub-problems is less than a preset threshold, or the maximum number of iterations is reached, and output the decision result of the best driving control scheme.

2. The refined trajectory-level guidance and control method for zero-waste transportation vehicles according to claim 1, characterized in that, The vehicle's operating status information in S1 includes the vehicle's current position, current lane, speed, acceleration, and the control area of ​​the intersection it is about to enter. Specifically, S1 involves classifying related vehicles into groups corresponding to the same intersection based on the vehicle's current location, the control area of ​​the intersection it is about to enter, and the road topology, according to their conflict relationships.

3. The refined trajectory-level guidance and control method for zero-waste transportation vehicles according to claim 1, characterized in that, The objective function in S2 is specifically: in, Indicates vehicle The travel time consumed within the intersection control area is used to describe the level of vehicle idle time at the intersection. The objective function value, Indicates vehicle The moment of entering the intersection, For vehicles The moment of arrival at the entrance to the intersection control area.

4. The refined trajectory-level guidance and control method for zero-waste transportation vehicles according to claim 3, characterized in that, The decision variables in S2 include lateral decision variables. Longitudinal decision variables and priority passage decision variables The horizontal decision variables Specifically, it is a 0-1 discrete variable representing whether a vehicle performs a lane-changing or overtaking maneuver. Indicates vehicle Perform a lane change or overtaking maneuver. Indicates vehicle Maintain your current lane; The vertical decision variables Specifically, vehicles Longitudinal speed control variables before entering and during passage through the intersection; Priority decision variables Specifically, it refers to vehicles. With vehicles j The order of passage between them Indicates vehicle Superior to vehicles Passage Indicates vehicle Superior to vehicles Passage is permitted.

5. A refined trajectory-level guidance and control method for zero-waste transportation vehicles according to claim 4, characterized in that, The constraints in S2 include vehicle kinematic constraints, time continuity constraints, conflict resolution constraints, and lateral control and priority relationship constraints. Among them, vehicle kinematic constraints are used to limit the range of values ​​for the average speed of vehicles traveling on road segments. Time continuity constraints are used to ensure consistency between the calculation of vehicle passage time at intersections and speed decisions; Conflict resolution constraints are used to ensure that only one vehicle is allowed to occupy the same conflict zone at any given time. Lateral control and associated constraints are used to ensure that a vehicle can only obtain right-of-way when it is performing a lane change or when it has legal driving conditions.

6. A refined trajectory-level guidance and control method for zero-waste transportation vehicles according to claim 5, characterized in that, The time continuity constraint is specifically constructed through the following process: The nonlinear relationship between the travel time of vehicle i through the intersection control area and its speed is determined as follows: ,in, This refers to a group of vehicles within the control area of ​​the same intersection. Introducing continuous auxiliary variables Used to indicate vehicles Within the intersection control zone, the time for vehicles to enter the intersection is: Vehicle The permissible speed range within the intersection control zone is divided into: 3 adjacent discrete sub-intervals: in, Indicates the minimum speed limit for that section of road. Indicates the maximum speed limit for this section of road. , , Each subinterval corresponds to a velocity value segment, which is used to construct a piecewise linear approximation function; To indicate which discrete interval the speed of vehicle r falls within, a 0-1 variable for interval selection is introduced. And add a range selection uniqueness constraint to ensure vehicle Select only one speed range at any given time: In each speed range Inside, to An approximation using a linear function: in, and In the interval The linearization coefficients, pre-calculated, are used to approximate the variation trend of the original nonlinear function within this interval; Piecewise linearized constraint expression, combined with interval selection variables Construct the following linear constraints to describe and Piecewise linear relationship between them: To ensure speed variables Consistent with the selected interval, add the following interval constraints: in, k Use the sub-interval numbering.

7. A refined trajectory-level guidance and control method for zero-waste transportation vehicles according to claim 6, characterized in that, The first objective function is specifically: in, This is a lower bound estimate of the optimal objective value for the subproblem; The constraints of the main problem include: in, This represents the set of Benders cuts that have been generated. and For the first The cut coefficients generated in the next iteration.

8. A refined trajectory-level guidance and control method for zero-waste transportation vehicles according to claim 7, characterized in that, The second objective function is as follows: The sub-problem constraints include time continuity constraints and conflict resolution constraints.

9. A refined trajectory-level guidance and control method for zero-waste transportation vehicles according to claim 8, characterized in that, The specific process of S33 is as follows: Based on the solutions to the subproblems, the corresponding Benders cuts are generated and fed back to the main problem. If the subproblems fall within the given discrete decision... If the following is feasible, then construct an optimality cut based on the dual information of the subproblem: If the subproblem is in a given discrete decision If the following is not feasible, then construct a feasible section: in, Indicates with vehicles The relevant infeasibility indicator coefficient is used to describe the degree to which the vehicle participates in currently infeasible driving control decision-making schemes; The upper bound constant is determined based on the set of infeasible constraints of the subproblems. This constant is used to limit the recurrence of discrete decision combinations that lead to infeasibility. Feasibility cut is used to ensure that the main problem does not generate discrete decision combinations that are equivalent or similar to the current infeasible driving control decision scheme in subsequent iterations.

Images