Signal control trunk road-oriented connected vehicle trajectory planning method, device and medium

Abstract

The application provides a connected vehicle trajectory planning method and device for signal-controlled trunk roads, and relates to the technical field of intelligent transportation. The method comprises the following steps: acquiring signal timing information of each signal intersection of a target trunk road, vehicle speed guided by a green wave of each road section, and a state of a vehicle around the self vehicle, to provide basic data for trajectory planning; predicting the running trajectory of the vehicle around the self vehicle in a planning time domain according to the state of the vehicle around the self vehicle, and combining the signal timing information and the state of the self vehicle to perform forward inference, so as to determine a target passing decision of the self vehicle through a stop line of a downstream signal intersection; taking trajectory smoothness of the vehicle, followability of the vehicle speed guided by the green wave, and passing efficiency as optimization targets, comprehensively considering vehicle kinematics constraints, driving safety constraints, and stop line constraints corresponding to the target passing decision, to construct a vehicle trajectory optimization model; solving the model at each planning time, obtaining a planned trajectory of the self vehicle, executing a trajectory instruction at the next time, and rolling and updating the trajectory planning result at subsequent planning times.

Description

Technical Field

[0001] This invention relates to the field of intelligent transportation technology, and more specifically, to a method, device, and medium for planning the trajectory of connected vehicles on signal-controlled trunk lines. Background Technology

[0002] The rapid pace of urbanization has exacerbated urban traffic congestion. Signalized arterial roads, as the backbone of urban transportation, directly impact the overall operational efficiency of the road network. The development of intelligent connected vehicle technology has brought new opportunities for refined traffic management. Intelligent connected vehicles can not only perceive their own status and surrounding environment in real time, but also communicate with roadside facilities and other vehicles to obtain information such as signal timing and traffic flow status. This provides a technological foundation for improving the efficiency of arterial roads.

[0003] In existing technologies, most research on vehicle trajectory planning for signalized intersections focuses on single-point intersections, aiming to reduce start-stop and delays by optimizing the speed of intelligent connected vehicles passing through a single intersection. A few studies have considered arterial road scenarios, attempting to collaboratively optimize vehicle trajectories across multiple intersections to improve the overall traffic efficiency of arterial roads.

[0004] However, single-point optimization methods lack a global perspective on the coordinated control of multiple intersections along a main road. A vehicle may achieve optimization at one intersection but stop at the next due to a lack of coordination, failing to fully utilize the green wave bandwidth resources generated by the coordinated control of the main road. Furthermore, the few studies that consider main road scenarios suffer from high computational complexity and fail to effectively integrate with classical main road coordinated control theory, making them difficult to apply in practical engineering. Summary of the Invention

[0005] This invention provides a method, device, and medium for planning the trajectory of connected vehicles for signal control trunk lines, in order to improve at least one of the above-mentioned technical problems.

[0006] In a first aspect, the present invention provides a method for planning the trajectory of connected vehicles for signal control trunk lines, which includes steps S1 to S4.

[0007] S1. Obtain signal timing information for each signalized intersection on the target trunk line, green wave guidance speed for each road segment, and status information of surrounding vehicles that affect driving.

[0008] S2. Based on the status information of the surrounding vehicles, predict the trajectory of the surrounding vehicles in the planning time domain, and perform forward inference based on the signal timing information and the vehicle's status to determine the target passage decision of the vehicle through the stop line of the downstream signalized intersection.

[0009] S3. Taking vehicle trajectory smoothness, green wave guided vehicle speed following and traffic efficiency as optimization objectives, and combining vehicle kinematic constraints, driving safety constraints and stop line constraints corresponding to the objective traffic decisions, a vehicle trajectory optimization model is constructed.

[0010] S4. Solve the vehicle trajectory optimization model at each planning time to obtain the planned trajectory of the vehicle in the planning time domain, and execute the trajectory instruction for the next time step in the planned trajectory. Update the trajectory planning result in the next planning time step.

[0011] As a further aspect of the present invention, S1 includes:

[0012] The period, phase difference, and green light duration of each phase of each signalized intersection in the target trunk line are obtained through vehicle-to-infrastructure communication, and the green wave guidance speed corresponding to each road segment in the target trunk line is also obtained.

[0013] The vehicle obtains the position, speed, and acceleration of vehicles around it through onboard sensors and vehicle-to-vehicle communication.

[0014] The surrounding vehicles are identified as either intelligent connected vehicles or conventional vehicles based on their communication attributes, and the identification results are used as input for predicting the trajectories of the surrounding vehicles.

[0015] As a further aspect of the present invention, S2's prediction of the trajectories of surrounding vehicles in the planning time domain based on the state information of the surrounding vehicles includes: When the surrounding vehicles are intelligent connected vehicles, the planned trajectory shared by the intelligent connected vehicle in the future planning time domain is obtained through vehicle-to-vehicle communication.

[0016] When the surrounding vehicles are conventional vehicles, the intelligent driving model uses the current state of the conventional vehicle and the state of the vehicle in front of it to perform forward extrapolation step by step to obtain the predicted trajectory of the conventional vehicle in the future planning time domain.

[0017] The planned trajectory and the predicted trajectory are used together as the boundary of the surrounding vehicle trajectories in the autonomous vehicle trajectory planning.

[0018] As a further aspect of the present invention, S2 performs forward inference based on the signal timing information and the vehicle status to determine the target passage decision for the vehicle to pass the stop line of the downstream signalized intersection, including: The forward inference trajectory is constructed based on the current state of the vehicle. The current state of the vehicle is: In the formula, Indicates that the car is The location at any given moment. Indicates that the car is The speed of time. Indicates that the car is Acceleration at any moment. This indicates the start time of the current plan.

[0019] Make the vehicle start with a constant acceleration Accelerate smoothly to cruising speed Then at the aforementioned cruising speed Travel at a constant speed.

[0020] Construct a safety boundary based on the trajectories of surrounding vehicles: In the formula, This indicates the position of the vehicle in front at the corresponding time. This indicates the time within the planning time domain. This represents the time displacement in the Newell model. This represents spatial displacement in the Newell model.

[0021] Determine whether the forward inference trajectory intrudes into the safety boundary. If the forward inference trajectory intrudes into the safety boundary, merge the forward inference trajectory with the safety boundary and use the merged trajectory as the new forward inference trajectory.

[0022] The relationship between the vehicle position and the stop line position at the end of the planning time domain is determined based on the forward inference trajectory that meets the safety distance requirements.

[0023] When satisfied The target traffic decision is that the vehicle will not pass the stop line within the planning time domain. This indicates the planning time domain. This indicates the vehicle's position at the end of the planning time domain. This indicates the location of the stop line at the downstream signal intersection.

[0024] When satisfied Then, calculate the time when the car crossed the stop line. And query the traffic light color at that moment. .in, Indicates the first The signal intersection The stop line is at The color change of the traffic lights at any given time. Indicates the stop line index. Indicates the index of the signalized intersection.

[0025] When satisfied The target passage decision is for the vehicle to pass the stop line within the planning time domain.

[0026] When satisfied The target passage decision is that the vehicle will not pass the stop line within the planning time domain.

[0027] As a further aspect of the present invention, the objective function of the vehicle trajectory optimization model is: .

[0028] In the formula, This means finding an optimal set of decision variables that minimizes the value of the function. This indicates the cost of smoothness. This represents the weighting coefficient corresponding to the smoothness cost. This indicates the cost of following. This represents the weighting coefficient corresponding to the following cost. This indicates the cost of efficiency. This represents the weighting coefficient corresponding to the efficiency cost.

[0029] The ride comfort cost is used to characterize the cumulative acceleration and jerk of the vehicle.

[0030] The following cost is used to characterize the deviation between the actual vehicle speed and the green wave-guided vehicle speed.

[0031] The efficiency cost is used to characterize the cumulative time before a vehicle passes the stop line.

[0032] As a further aspect of the present invention, the smoothness trade-off is: .

[0033] The following cost is: .

[0034] The efficiency cost is: .

[0035] In the formula, This indicates the start time of the current plan. This indicates the planning time domain. Indicates that the car is An auxiliary variable indicating whether the time has not yet passed the stop line. This indicates the time within the planning time domain. Indicates that the car is Acceleration at any moment. Indicates that the car is The acceleration of time. Indicates the time step. Indicates that the car is The speed of time. Indicates that the car is The green wave at any given time guides the vehicle speed.

[0036] As a further aspect of the present invention, the vehicle kinematic constraints include acceleration recursive constraints, velocity recursive constraints, displacement recursive constraints, and variable upper and lower limit constraints.

[0037] The acceleration recursion constraint is: .

[0038] The velocity recursion constraint is: .

[0039] The displacement recursion constraint is: .

[0040] The upper and lower limits of acceleration are: .

[0041] The upper and lower limits of the speed are constrained as follows: .

[0042] The upper and lower limits of the jerk are: .

[0043] In the formula, This indicates the time within the planning time domain. express The next discrete moment. Indicates that the car is Acceleration at any moment. Indicates that the car is The acceleration of time. Indicates the time step. Indicates that the car is The speed of time. Indicates that the car is The speed of time. Indicates that the car is The location at any given moment. Indicates that the car is The location at any given moment. This indicates the maximum allowable deceleration of the vehicle. This indicates the maximum allowable acceleration for the vehicle. This indicates the road speed limit. This indicates the lower limit of the permissible acceleration for the vehicle. This indicates the maximum allowable acceleration for the vehicle.

[0044] As a further aspect of the present invention, the driving safety constraints include safety distance constraints for vehicles ahead in the same lane, safety distance constraints for vehicles behind in the same lane, and speed limits for conflict zones.

[0045] The safety distance constraint for the vehicle in front in the same lane is: .

[0046] The following safety distance constraint for vehicles in the same lane is: .

[0047] The speed limit constraint in the conflict zone is as follows: .

[0048] In the formula, This indicates the position of the vehicle in front at the corresponding time. This represents the time displacement in the Newell model. This represents spatial displacement in the Newell model. This indicates the position of the vehicle following in the same lane at the corresponding time. This indicates the speed limit for vehicles within the intersection conflict zone after passing the stop line. This represents a large constant used to relax the constraints. Indicates that the car is An auxiliary variable indicating whether the time has not yet passed the stop line.

[0049] As a further aspect of the present invention, the stop line constraint in S3 includes: A stop line constraint is established based on the target passage decision, and the stop line constraint includes a red light violation constraint and a decision result constraint.

[0050] The red light violation constraint is: .

[0051] In the formula, This indicates the location of the stop line at the downstream signal intersection. Indicates the first The signal intersection The stop line is at The color change of the traffic lights at any given time. Indicates the index of the signalized intersection. Indicates the stop line index.

[0052] The constraints on the decision outcome include: .

[0053] .

[0054] In the formula, This indicates the start time of the current plan. This indicates the planning time domain. This indicates the vehicle's position at the end of the planning time domain. Represents the target travel decision variable. This indicates that the plan will not pass through the stop line within the planning time domain. This indicates the stop line passed within the planning time domain.

[0055] As a further aspect of the present invention, S4 includes: At each planning moment, the latest signal timing information, green wave guided vehicle speed, vehicle status, surrounding vehicle status, and target traffic decision obtained in S2 are input into the vehicle trajectory optimization model constructed in S3.

[0056] Solve the vehicle trajectory optimization model to obtain the planned trajectory of the vehicle in the planning time domain.

[0057] Only the speed command or acceleration command for the next moment in the planned trajectory is sent to the vehicle actuator.

[0058] S1 to S4 are re-executed at the next planning time to form rolling time-domain control.

[0059] When the vehicle trajectory optimization model fails to find a feasible solution within a specified time, an intelligent driver model takes over vehicle control to perform safe following driving.

[0060] Secondly, the present invention provides a connected vehicle trajectory planning device for signal control trunk lines, which includes a processor, a memory, and a computer program stored in the memory.

[0061] The computer program can be executed by the processor to implement the connected vehicle trajectory planning method for signal control trunk lines as described in any part of the first aspect.

[0062] Thirdly, the present invention provides a computer-readable storage medium. The computer-readable storage medium includes a stored computer program.

[0063] When the computer program is executed, it controls the device containing the computer-readable storage medium to perform the connected vehicle trajectory planning method for signal control trunk lines as described in any section of the first aspect.

[0064] By adopting the above technical solution, the present invention can achieve the following technical effects: This invention integrates the green wave speed guidance generated by the coordinated control of trunk line signals directly into the objective function of connected vehicle trajectory planning. This synergizes the trajectory optimization of individual vehicles with the overall control objective of the trunk line, guiding traffic flow to fully utilize the green wave bandwidth and avoiding local optima caused by single-point optimization, effectively improving the overall traffic efficiency of the trunk line. Simultaneously, this invention can adapt to traffic flow environments where intelligent connected vehicles and conventional vehicles coexist: for conventional vehicles, a car-following model is used for trajectory prediction, enabling the vehicle to proactively adapt to random driving behaviors ahead during trajectory planning, improving the practicality and robustness of the method. Addressing the non-convexity problem caused by the coupling of discrete green light window selection and continuous smooth driving control in trajectory planning, this invention separates decision-making and optimization through forward inference, transforming the complex problem into an efficiently solvable convex optimization problem, meeting the requirements of real-time control. Furthermore, the model incorporates kinematic constraints, Newell model-based safety distance constraints, and a rolling time-domain online update mechanism to ensure driving safety, and adds penalty terms for acceleration and jerk to the objective function, effectively improving ride comfort. Attached Figure Description

[0065] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the specific embodiments of the present invention will be briefly introduced below. It should be understood that the following drawings only show some specific embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0066] Figure 1 This is a flowchart of the trajectory planning method for connected vehicles.

[0067] Figure 2 This is a schematic diagram of the layout of main roads and intersections.

[0068] Figure 3 It is a spatiotemporal trajectory map of a vehicle without trajectory planning.

[0069] Figure 4 It is a spatiotemporal trajectory diagram of the vehicle after applying the method of the present invention. Detailed Implementation

[0070] The technical solutions of the present invention will now be clearly and completely described with reference to the accompanying drawings in the embodiments of the present invention.

[0071] Example 1, please refer to Figures 1 to 4 The first embodiment of the present invention provides a method for planning the trajectory of connected vehicles for signal control trunk lines, which includes steps S1 to S4.

[0072] S1. Obtain the signal timing information of each signalized intersection in the target trunk road, the green wave guidance speed corresponding to each road segment, and the status information of surrounding vehicles affecting driving. Preferably, S1 includes S11 to S13.

[0073] S11. Obtain the cycle, phase difference, and green light duration of each signalized intersection in the target trunk line through vehicle-road communication, and obtain the green wave guidance speed corresponding to each road segment in the target trunk line.

[0074] S12. Obtain the position, speed, and acceleration of vehicles around the vehicle through onboard sensors and vehicle-to-vehicle communication.

[0075] S13. Identify the surrounding vehicles as intelligent connected vehicles or conventional vehicles based on their communication attributes, and use the identification results as input for predicting the trajectories of the surrounding vehicles.

[0076] Taking an urban trunk road with four signalized intersections as an example, such as Figure 2As shown. The main road runs from south to north for the uphill direction and from north to south for the downhill direction. Connected Automated Vehicles (CAVs) and Regular Vehicles (RVs) travel together on the road.

[0077] First, real-time timing information (including cycle, phase difference, green light duration for each phase, etc.) of each signalized intersection in the target trunk line is obtained through vehicle-to-infrastructure communication, as well as the green wave guidance speed for each road segment.

[0078] Secondly, through onboard sensors (such as radar and cameras) and vehicle-to-vehicle communication, the system obtains status information of surrounding vehicles that affect driving, including the position, speed, and acceleration of surrounding vehicles, and identifies whether they are CAVs or RVs.

[0079] S2. Based on the status information of the surrounding vehicles, predict the trajectory of the surrounding vehicles in the planning time domain, and perform forward inference based on the signal timing information and the vehicle's status to determine the target passage decision of the vehicle through the stop line of the downstream signalized intersection.

[0080] Preferably, S2 includes S201 to S203.

[0081] S201. When the surrounding vehicles are connected automated vehicles (CAVs), obtain the planned trajectory shared by the connected automated vehicle in the future planning time domain through vehicle-to-vehicle communication.

[0082] S202. When the surrounding vehicles are regular vehicles (RVs), the intelligent driver model is used to perform forward extrapolation step by step based on the current state of the regular vehicle and the state of the vehicle in front of it, so as to obtain the predicted trajectory of the regular vehicle in the future planning time domain.

[0083] S203. The planned trajectory and the predicted trajectory are used together as the boundary of the surrounding vehicle trajectories in the vehicle trajectory planning.

[0084] S201 to S203 aim to construct a feasible solution space for subsequent precise trajectory optimization and make key high-level decisions. Surrounding vehicles are handled differently. For CAVs, their planned trajectories broadcast to the surrounding area over a future time domain are directly obtained through vehicle-to-vehicle communication. For RVs, an Intelligent Driver Model (IDM) is used to predict their future trajectories. The IDM model can calculate a reasonable following acceleration based on the state of the vehicle in front and its own speed. This embodiment starts from the current moment... Start with time step s is extrapolated forward to predict its time domain in the planning time domain. The trajectory within s.

[0085] Preferably, S2 further includes S204 to S212.

[0086] S204. Construct a forward inference trajectory based on the current state of the vehicle. The current state of the vehicle is: In the formula, Indicates that the car is The location at any given moment. Indicates that the car is The speed of time. Indicates that the car is Acceleration at any moment. This indicates the start time of the current plan.

[0087] S205, make the vehicle initially accelerate at a constant speed. Accelerate smoothly to cruising speed Then at the aforementioned cruising speed Travel at a constant speed.

[0088] To avoid the non-convexity problem caused by simultaneously optimizing "which green light window to choose" and "how to drive," we first make motion decisions. We then construct a path from the current state... The forward inference trajectory from the start. This trajectory assumes the vehicle initially accelerates at a constant speed. Accelerate uniformly to a cruising speed (Do not exceed the road speed limit), and then maintain a constant speed.

[0089] If the inferred trajectory, while satisfying the safe distance constraint from the vehicle in front, ensures that the vehicle does not cross the stop line within the planning time domain, the decision is to not proceed within the planning time domain. If the vehicle can cross the stop line, the next step is to determine whether the light is green when crossing the stop line. If it is green, the decision is to proceed within the planning time domain; if it is red, the decision is not to proceed.

[0090] S206. Construct a safety boundary based on the trajectories of surrounding vehicles: In the formula, This indicates the position of the vehicle in front at the corresponding time. This indicates the time within the planning time domain. This represents the time displacement in the Newell model. This represents spatial displacement in the Newell model.

[0091] Specifically, check whether this inferred trajectory will encroach on the safety boundary line constructed by the trajectory of the preceding vehicle (based on the Newell model). (Generation). If there is no intrusion, the security test is passed. If there is an intrusion, the inferred trajectory line is merged with the security boundary line (for example, taking the maximum displacement value at the same moment) to form a new, safe inferred trajectory.

[0092] S207. Determine whether the forward inference trajectory intrudes into the safety boundary. When the forward inference trajectory intrudes into the safety boundary, merge the forward inference trajectory with the safety boundary and use the merged trajectory as the new forward inference trajectory.

[0093] S208. Based on the forward inference trajectory that meets the safety distance requirements, determine the relationship between the vehicle position and the stop line position at the end of the planning time domain.

[0094] S209, when satisfied The target traffic decision is that the vehicle will not pass the stop line within the planning time domain. This indicates the planning time domain. This indicates the vehicle's position at the end of the planning time domain. This indicates the location of the stop line at the downstream signal intersection.

[0095] S210, when satisfied Then, calculate the time when the car crossed the stop line. And query the traffic light color at that moment. .in, Indicates the first The signal intersection The stop line is at The color change of the traffic lights at any given time. Indicates the stop line index. Indicates the index of the signalized intersection.

[0096] S211, when satisfied The target passage decision is for the vehicle to pass the stop line within the planning time domain.

[0097] S212, when satisfied The target passage decision is that the vehicle will not pass the stop line within the planning time domain.

[0098] Specifically, based on the inferred trajectory obtained from passing security tests, it is determined that at the end of the planning time domain... Time, vehicle location Is it greater than the stop line position? .like The decision result was "not approved". This means that the vehicle cannot reach the stop line within the planned time range. Then, calculate the time when the vehicle crossed the stop line. Query Traffic light colors at all times If it's a green light The decision result is "passable" ( If it's a red light. The decision result was "not passable". ).

[0099] S3. Taking vehicle trajectory smoothness, green wave guided vehicle speed following and traffic efficiency as optimization objectives, and combining vehicle kinematic constraints, driving safety constraints and stop line constraints corresponding to the objective traffic decisions, a vehicle trajectory optimization model is constructed.

[0100] Specifically, based on the prediction and decision results of S2, a quadratic programming (QP) model is constructed.

[0101] The decision variable is: the jerk during the planning time domain. acceleration ,speed ,Location and 0-1 variables .

[0102] The objective function of the vehicle trajectory optimization model is: .

[0103] In the formula, This means finding an optimal set of decision variables that minimizes the value of the function. This indicates the cost of smoothness. This represents the weighting coefficient corresponding to the smoothness cost. This indicates the cost of following. This represents the weighting coefficient corresponding to the following cost. This indicates the cost of efficiency. This represents the weighting coefficient corresponding to the efficiency cost.

[0104] The smoothness cost is: It is used to characterize vehicle acceleration and the accumulation of jerk, and is designed to reduce abrupt acceleration / deceleration and shocks, thereby improving comfort.

[0105] The following cost is: It is used to characterize the deviation between the actual vehicle speed and the green wave guidance speed, and is designed to guide the vehicle speed to be close to the green wave speed recommended by the main line coordination.

[0106] The efficiency cost is: This is used to characterize the cumulative time before a vehicle crosses the stop line, and is designed to encourage vehicles to cross the stop line as quickly as possible to reduce travel time.

[0107] in, This indicates the start time of the current plan. This indicates the planning time domain. Indicates that the car is An auxiliary variable indicating whether the time has not yet passed the stop line. This indicates the time within the planning time domain. Indicates that the car is Acceleration at any moment. Indicates that the car is The acceleration of time. Indicates the time step. Indicates that the car is The speed of time. Indicates that the car is The green wave at any given time guides the vehicle speed.

[0108] in, It is an auxiliary variable, set to 1 when the vehicle has not yet crossed the stop line and to 0 after it has. This ensures that the optimization primarily focuses on the vehicle's movement along the road segment.

[0109] Based on the above embodiments, in an optional embodiment of the present invention, the vehicle kinematic constraints include acceleration recursive constraints, velocity recursive constraints, displacement recursive constraints, and variable upper and lower limit constraints.

[0110] The acceleration recursion constraint is: .

[0111] The velocity recursion constraint is: .

[0112] The displacement recursion constraint is: .

[0113] The upper and lower limits of acceleration are: .

[0114] The upper and lower limits of the speed are constrained as follows: .

[0115] The upper and lower limits of the jerk are: .

[0116] In the formula, This indicates the time within the planning time domain. express The next discrete moment. Indicates that the car is Acceleration at any moment. Indicates that the car is The acceleration of time. Indicates the time step. Indicates that the car is The speed of time. Indicates that the car is The speed of time. Indicates that the car is The location at any given moment. Indicates that the car is The location at any given moment. This indicates the maximum allowable deceleration of the vehicle. This indicates the maximum allowable acceleration for the vehicle. This indicates the road speed limit. This indicates the lower limit of the permissible acceleration for the vehicle. This indicates the maximum allowable acceleration for the vehicle.

[0117] Preferably, the driving safety constraints include safety distance constraints for vehicles ahead in the same lane, safety distance constraints for vehicles behind in the same lane, and speed limits in conflict zones.

[0118] The safety distance constraint for the vehicle in front in the same lane is: .

[0119] The following safety distance constraint for vehicles in the same lane is: To ensure that no danger is caused to the vehicle behind.

[0120] The speed limit constraint in the conflict zone is as follows: This means that after a vehicle passes the stop line, it must comply with the speed limit inside the intersection. .

[0121] Specifically, the safety distance constraint for the vehicle in front in the same lane is constructed based on the Newell car-following model.

[0122] in, This indicates the position of the vehicle in front at the corresponding time. This represents the time displacement in the Newell model. This represents spatial displacement in the Newell model. This indicates the position of the vehicle following in the same lane at the corresponding time. This indicates the speed limit for vehicles within the intersection conflict zone after passing the stop line. This represents a large constant used to relax the constraints. Indicates that the car is An auxiliary variable indicating whether the time has not yet passed the stop line.

[0123] Preferably, a stop line constraint is established based on the target traffic decision. The stop line constraint includes a red light violation constraint that prohibits crossing the stop line during a red light and a decision result constraint that ensures the vehicle's position at the end of the planning time domain remains consistent with the decision result in step S2.

[0124] The red light violation constraint is: In the formula, This indicates the location of the stop line at the downstream signal intersection. Indicates the first The signal intersection The stop line is at The color change of the traffic lights at any given time. Indicates the index of the signalized intersection. Indicates the stop line index.

[0125] The constraints on the decision outcome include: .

[0126] .

[0127] Specifically, Must be related to the decision outcome in S2 Consistent.

[0128] In the formula, This indicates the start time of the current plan. This indicates the planning time domain. This indicates the vehicle's position at the end of the planning time domain. Represents the target travel decision variable. This indicates that the plan will not pass through the stop line within the planning time domain. This indicates the stop line passed within the planning time domain.

[0129] S4. Solve the vehicle trajectory optimization model at each planning time to obtain the planned trajectory of the vehicle in the planning time domain, and execute the trajectory instruction for the next time step in the planned trajectory. Update the trajectory planning result in the next planning time step.

[0130] Specifically, the solution and execution employ a rolling time-domain control method. At each time step, the optimization model is re-solved based on the latest state information, and the trajectory within the planned time domain is updated. However, only the velocity or acceleration command for the next moment is issued to the vehicle actuators. If the solver cannot find a feasible solution within the specified time, vehicle control is taken over by the IDM model to ensure driving safety.

[0131] Preferably, S4 includes S41 to S45.

[0132] S41. At each planning moment, the latest signal timing information, green wave guided vehicle speed, vehicle status, surrounding vehicle status, and the target passage decision obtained in S2 are input into the vehicle trajectory optimization model constructed in S3.

[0133] S42. Solve the vehicle trajectory optimization model to obtain the planned trajectory of the vehicle in the planning time domain.

[0134] S43. Only the speed command or acceleration command for the next moment in the planned trajectory is sent to the vehicle actuator.

[0135] S44. Re-execute S1 to S4 at the next planning time to form rolling time-domain control.

[0136] S45. When the vehicle trajectory optimization model fails to find a feasible solution within a specified time, an intelligent driver model takes over vehicle control to perform safe following driving.

[0137] This embodiment employs a rolling time-domain control method. At each time step... The latest data obtained from S1 and the calculation results from S2 are input into the QP model constructed in S3, and the Gurobi solver is used for fast solution. After obtaining the optimal trajectory sequence, only the next time step is considered. The speed or acceleration command is issued to the vehicle's actuators. At any given moment, the latest data is retrieved, and a new cycle of "prediction-decision-optimization-execution" begins. If the solver fails to find a feasible solution within a specified time (e.g., 50ms) at any time, to ensure safety, vehicle control will automatically switch to the IDM model for car-following.

[0138] The method of this embodiment was verified using the SUMO simulation platform. The CAV penetration rate was set to 25%. The vehicle trajectories without trajectory planning (only signal timing optimization) and after applying the method of this invention were compared. Figure 3 and Figure 4 As shown.

[0139] Figure 3 In the middle, the vehicle did not travel at the green wave speed after passing through intersection 1, causing it to stop and wait before intersection 2, resulting in additional delays and stops.

[0140] Figure 4 In the simulation, the CAV (marked with a special symbol) applied the method of this invention. As can be seen, the CAV actively adjusted its speed on the road segment, smoothly following the green wave speed, thus passing through intersections 1 and 2 consecutively during the green light window. Following vehicles were also affected, achieving non-stop passage. Simulation data shows that after applying this method, the average delay time of mixed traffic flow decreased by 5.8%, and the average number of stops decreased by 7.5%, effectively improving the traffic efficiency of the main road.

[0141] In summary, this invention provides a trajectory planning method for connected vehicles (CAVs) on signal-controlled arterial roads. This method integrates arterial road signal timing, green wave speeds, and surrounding vehicle states, and employs a convexity processing strategy that separates decision-making and optimization. Under the premise of ensuring safety, it plans a smooth, efficient trajectory for CAVs that follows green wave guidance, thereby guiding mixed traffic flow to fully utilize the green wave bandwidth and improving the overall traffic efficiency of the arterial road. This achieves efficient, safe, and comfortable control of connected vehicles in complex mixed traffic flow environments, and has significant practical value for improving the performance of urban arterial traffic systems.

[0142] When applied to urban signal-controlled trunk lines, this invention can guide connected vehicle traffic to pass through intersections efficiently and smoothly within the green wave zone, significantly improving the overall operational efficiency of the trunk line traffic system.

[0143] This invention uses the green wave speed generated by trunk line signal coordination control as the guiding target for trajectory planning, so that the optimization of single vehicle trajectory is coordinated with the overall control target of trunk line (such as green wave bandwidth), guiding the traffic flow to make full use of green wave resources and avoiding the local optimum problem caused by single-point optimization.

[0144] This invention's method is applicable not only to CAVs but also considers mixed traffic flow environments by predicting the trajectories of RVs. When planning its own trajectory, the CAV can proactively adapt to the random driving behavior of the RVs ahead, improving the method's practicality and robustness.

[0145] By using the forward inference decision module, the discrete problem of "which green light window to choose" is separated from the continuous problem of "how to pass smoothly", and the complex non-convex optimization problem is transformed into a convex optimization problem (quadratic programming QP), which greatly improves the solution efficiency and meets the needs of real-time control.

[0146] The model ensures dynamic safety of vehicle trajectory through strict kinematic constraints, a Newell-based safe distance constraint, and online updates in the rolling time domain. Meanwhile, by introducing penalties for acceleration and jerk into the objective function, passenger comfort is effectively improved.

[0147] Example 2: The present invention provides a connected vehicle trajectory planning device for signal control trunk lines, which includes a processor, a memory, and a computer program stored in the memory.

[0148] The computer program can be executed by the processor to implement the connected vehicle trajectory planning method for signal control trunk lines as described in any section of Embodiment 1.

[0149] The device can be a vehicle-side device / vehicle body, roadside device / edge computing device, cloud server / traffic control platform, vehicle-road-cloud collaborative system, or other computing device. Other computing devices include industrial computers, in-vehicle industrial control computers, traffic control servers, simulation platform servers, etc., as long as they have a processor and memory and can run the program, they can also be used as execution devices.

[0150] Example 3: The present invention provides a computer-readable storage medium. The computer-readable storage medium includes a stored computer program.

[0151] When the computer program runs, it controls the device containing the computer-readable storage medium to execute any segment of Embodiment 1, which describes a method for planning the trajectory of connected vehicles for signal control trunk lines.

[0152] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for trajectory planning of connected vehicles on signal-controlled trunk lines, characterized in that, include: S1. Obtain the signal timing information of each signalized intersection in the target trunk line, the green wave guidance speed corresponding to each road segment, and the status information of surrounding vehicles that affect driving around the vehicle. S2. Based on the status information of the surrounding vehicles, predict the trajectory of the surrounding vehicles in the planning time domain, and perform forward inference based on the signal timing information and the vehicle status to determine the target passage decision of the vehicle through the stop line of the downstream signalized intersection. S3. Taking vehicle trajectory smoothness, green wave guided vehicle speed following and traffic efficiency as optimization objectives, and combining vehicle kinematic constraints, driving safety constraints and stop line constraints corresponding to the target traffic decisions, a vehicle trajectory optimization model is constructed. S4. Solve the vehicle trajectory optimization model at each planning time to obtain the planned trajectory of the vehicle in the planning time domain, and execute the trajectory instruction for the next time step in the planned trajectory. Update the trajectory planning result in the next planning time step.

2. The method for planning the trajectory of connected vehicles for signal control trunk lines according to claim 1, characterized in that, S2 performs forward inference based on the signal timing information and the vehicle's status to determine the target passage decision for the vehicle to pass the stop line of the downstream signalized intersection, including: The forward inference trajectory is constructed based on the current state of the vehicle. The current state of the vehicle is: In the formula, Indicates that the car is The position at that moment; Indicates that the car is The speed of time; Indicates that the car is Acceleration at any moment; Indicates the start time of the current plan; Make the vehicle start with a constant acceleration Accelerate smoothly to cruising speed Then at the aforementioned cruising speed Travel at a constant speed; Construct a safety boundary based on the trajectories of surrounding vehicles: In the formula, Indicates the position of the vehicle in front at the corresponding time; Indicates the time within the planning time domain; This represents the time displacement in the Newell model; This represents spatial displacement in the Newell model; Determine whether the forward inference trajectory intrudes into the safety boundary. If the forward inference trajectory intrudes into the safety boundary, merge the forward inference trajectory with the safety boundary and use the merged trajectory as the new forward inference trajectory. The relationship between the vehicle position and the stop line position at the end of the planning time domain is determined based on the forward inference trajectory that meets the safety distance requirements. When satisfied The target traffic decision is that the vehicle will not pass the stop line within the planning time domain; where, Indicates the planning time domain; This indicates the vehicle's position at the end of the planning time domain; Indicates the location of the stop line at the downstream signal intersection; When satisfied Then, calculate the time when the car crossed the stop line. And query the traffic light color at that moment. ;in, Indicates the first The signal intersection The stop line is at The color change of the traffic lights at any given time; Indicates the stop line index; Indicates the crossroads index; When satisfied The target passage decision is for the vehicle to pass the stop line within the planning time domain; When satisfied The target passage decision is that the vehicle will not pass the stop line within the planning time domain.

3. The method for planning the trajectory of connected vehicles for signal control trunk lines according to claim 1, characterized in that, The objective function of the vehicle trajectory optimization model is: ； In the formula, This means finding an optimal set of decision variables that minimizes the value of the function. Indicates the cost of smoothness; This represents the weighting coefficient corresponding to the smoothness cost; Indicates the cost of following; This represents the weighting coefficient corresponding to the following cost; Indicates the cost of efficiency; Indicates the weighting coefficient corresponding to the efficiency cost; The ride comfort cost is used to characterize the accumulation of vehicle acceleration and jerk. The following cost is used to characterize the deviation between the actual vehicle speed and the green wave guided vehicle speed; The efficiency cost is used to characterize the cumulative time before a vehicle passes the stop line.

4. The method for planning the trajectory of connected vehicles for signal control trunk lines according to claim 3, characterized in that, The smoothness cost is: ; The following cost is: ; The efficiency cost is: ; In the formula, Indicates the start time of the current plan; Indicates the planning time domain; Indicates that the car is An auxiliary variable indicating whether the time has not yet passed the stop line; Indicates the time within the planning time domain; Indicates that the car is Acceleration at any moment; Indicates that the car is The acceleration of time; Indicates the time step; Indicates that the car is The speed of time; Indicates that the car is The green wave at any given time guides the vehicle speed.

5. The method for planning the trajectory of connected vehicles for signal control trunk lines according to claim 1, characterized in that, The vehicle kinematic constraints include acceleration recursive constraints, velocity recursive constraints, displacement recursive constraints, and upper and lower limit constraints for variables. The acceleration recursion constraint is: ; The velocity recursion constraint is: ; The displacement recursion constraint is: ; The upper and lower limits of acceleration are: ; The upper and lower limits of the speed are constrained as follows: ; The upper and lower limits of the jerk are: ; In the formula, Indicates the time within the planning time domain; express The next discrete moment; Indicates that the car is Acceleration at any moment; Indicates that the car is The acceleration of time; Indicates the time step; Indicates that the car is The speed of time; Indicates that the car is The speed of time; Indicates that the car is The position at that moment; Indicates that the car is The position at that moment; This indicates the maximum allowable deceleration for the vehicle. This indicates the maximum allowable acceleration for the vehicle. Indicates the road speed limit; This indicates the lower limit of the permissible acceleration for the vehicle. This indicates the maximum allowable acceleration for the vehicle.

6. The method for planning the trajectory of connected vehicles for signal control trunk lines according to claim 5, characterized in that, The driving safety constraints include the safety distance constraints for vehicles ahead in the same lane, the safety distance constraints for vehicles behind in the same lane, and the speed limit constraints for conflict zones; The safety distance constraint for the vehicle in front in the same lane is: ; The following safety distance constraint for vehicles in the same lane is: ; The speed limit constraint in the conflict zone is as follows: ; In the formula, Indicates the position of the vehicle in front at the corresponding time; This represents the time displacement in the Newell model; This represents spatial displacement in the Newell model; Indicates the position of the vehicle following in the same lane at the corresponding time; This indicates the speed limit for vehicles within the intersection conflict zone after they have passed the stop line; This represents a large constant used to relax the constraints; Indicates that the car is An auxiliary variable indicating whether the time has not yet passed the stop line; The stop line constraints in S3 include: Based on the target passage decision, a stop line constraint is established, which includes a red light violation constraint and a decision result constraint. The red light violation constraint is: ； In the formula, Indicates the location of the stop line at the downstream signal intersection; Indicates the first The signal intersection The stop line is at The color change of the traffic lights at any given time; Indicates the crossroads index; Indicates the stop line index; The constraints on the decision outcome include: ； ； In the formula, Indicates the start time of the current plan; Indicates the planning time domain; This indicates the vehicle's position at the end of the planning time domain; Represents the target travel decision variable. This indicates that the plan will not pass through the stop line within the planning time domain. This indicates the stop line passed within the planning time domain.

7. A method for planning the trajectory of connected vehicles for signal control trunk lines according to any one of claims 1 to 6, characterized in that, S4 include: At each planning moment, the latest signal timing information, green wave guided vehicle speed, vehicle status, surrounding vehicle status, and target passage decision obtained in S2 are input into the vehicle trajectory optimization model constructed in S3. Solve the vehicle trajectory optimization model to obtain the planned trajectory of the vehicle in the planning time domain; Only the speed command or acceleration command for the next moment in the planned trajectory is sent to the vehicle actuator; S1 to S4 are re-executed at the next planning time to form rolling time-domain control; When the vehicle trajectory optimization model fails to find a feasible solution within a specified time, an intelligent driver model takes over vehicle control to perform safe following driving.

8. A method for planning the trajectory of connected vehicles for signal-controlled trunk lines according to any one of claims 1 to 6, characterized in that, S2 predicts the trajectories of surrounding vehicles in the planning time domain based on the state information of the surrounding vehicles, including: When the surrounding vehicles are intelligent connected vehicles, the planned trajectory shared by the intelligent connected vehicle in the future planning time domain is obtained through vehicle-to-vehicle communication; When the surrounding vehicles are conventional vehicles, the intelligent driving model is used to perform forward extrapolation step by step based on the current state of the conventional vehicle and the state of the vehicle in front of it, so as to obtain the predicted trajectory of the conventional vehicle in the future planning time domain. The planned trajectory and the predicted trajectory are used together as the boundary of the surrounding vehicle trajectories in the autonomous vehicle trajectory planning; S1 includes: The period, phase difference, and green light duration of each phase of each signalized intersection in the target trunk line are obtained through vehicle-road communication, and the green wave guidance speed corresponding to each road segment in the target trunk line is also obtained. The vehicle obtains the position, speed, and acceleration of vehicles around it through onboard sensors and vehicle-to-vehicle communication; The surrounding vehicles are identified as either intelligent connected vehicles or conventional vehicles based on their communication attributes, and the identification results are used as input for predicting the trajectories of the surrounding vehicles.

9. A connected vehicle trajectory planning device for signal control trunk lines, characterized in that, Includes a processor, a memory, and a computer program stored in the memory; The computer program can be executed by the processor to implement the connected vehicle trajectory planning method for signal control trunk lines as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program; When the computer program is executed, it controls the device containing the computer-readable storage medium to perform the connected vehicle trajectory planning method according to any one of claims 1 to 8.

Images