Cooperative decision-making method in high-concurrency traffic scenario and related device

Abstract

This invention provides a collaborative decision-making method and related equipment for high-concurrency traffic scenarios. The method includes: dynamically dividing the functional areas of a road segment based on the monitored arrival traffic volume to obtain the lengths of the lane-changing zone and speed adjustment zone; grouping and guiding the speed calculation of vehicle queues entering the lane-changing zone and arranging them longitudinally in a staggered manner to generate lane-changing decisions; planning the longitudinal trajectory of vehicles entering the speed adjustment zone to obtain vehicle speed control commands; and jointly optimizing the signal timing parameters of downstream intersections to obtain the green light duration for each signal phase. This method adapts to changes in traffic flow by dynamically adjusting the length of functional zones, achieves rapid lane changing for multiple vehicles through a collaborative lane-changing strategy, and reduces vehicle delays by jointly optimizing vehicle trajectories and signal timing. It effectively solves the problem that fixed zoning methods cannot adapt to dynamic changes in traffic flow, improving the efficiency of road segment space resource utilization and intersection throughput.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a collaborative decision-making method and related equipment for high-concurrency traffic scenarios. Background Technology

[0002] With the acceleration of urbanization, urban road traffic congestion has become increasingly serious, especially on road sections between adjacent intersections. The lack of effective coordination between vehicle lane-changing needs and signal control leads to low traffic efficiency. Existing traffic flow optimization methods typically manage road sections using fixed functional zoning, that is, pre-delineating fixed-length lane-changing zones and speed adjustment zones.

[0003] This static zoning method cannot be adjusted according to changes in real-time traffic flow. When traffic flow fluctuates significantly, fixed functional zoning can easily lead to insufficient lane-changing area length or too short speed adjustment area, making it difficult for vehicles to complete the necessary lane changes within the limited road space, resulting in frequent vehicle stops, increased queuing delays, and decreased traffic efficiency. Summary of the Invention

[0004] The main objective of this invention is to solve the technical problem that in the processing of low-quality engineering drawings, the fixed threshold mechanism is sensitive to changes in the drawing scale and noise interference, resulting in low recognition accuracy and requiring a large amount of manual intervention. This invention provides a collaborative decision-making method for high-concurrency traffic scenarios, wherein the adaptive control method includes: Based on the monitored traffic volume arriving at the road segment, the functional areas of the road segment are dynamically divided to obtain the lengths of the vehicle lane-changing zone and the speed adjustment zone. Based on the length of the lane-changing zone, the vehicle queues entering the lane-changing zone are grouped and guided to calculate the target guidance speed for different vehicle queues. Based on the target guidance speed, the vehicle queues are longitudinally staggered to generate a vehicle lane-changing decision. Based on the length of the speed adjustment zone and the execution result of the vehicle lane-changing decision, longitudinal trajectory planning is performed on the vehicles entering the speed adjustment zone to obtain the vehicle speed control command. Based on the vehicle speed control command and the estimated time for the vehicle to pass the stop line, the signal timing parameters of the downstream intersection are jointly optimized to obtain the green light duration for each signal phase.

[0005] The present invention also provides a collaborative decision-making device for high-concurrency traffic scenarios, the collaborative decision-making device for high-concurrency traffic scenarios comprising: The dynamic zoning module is used to dynamically divide the functional areas of a road segment based on the monitored traffic volume, and to obtain the lengths of the lane-changing zone and the speed adjustment zone. The collaborative lane-changing module is used to calculate the grouping guidance speed of vehicle queues entering the vehicle lane-changing area based on the length of the vehicle lane-changing area, obtain the target guidance speed of different vehicle queues, and arrange each vehicle queue longitudinally in a staggered manner according to the target guidance speed to generate vehicle lane-changing decisions. The trajectory planning module is used to perform longitudinal trajectory planning for vehicles entering the speed adjustment zone based on the length of the speed adjustment zone and the execution result of the vehicle lane change decision, so as to obtain the vehicle speed control command. The signal optimization module is used to jointly optimize the signal timing parameters of the downstream intersection based on the vehicle speed control command and the estimated time for the vehicle to pass through the stop line, so as to obtain the green light duration for each signal phase.

[0006] The present invention also provides a collaborative decision-making device for high-concurrency traffic scenarios, comprising: a memory and at least one processor, wherein the memory stores instructions, and the memory and the at least one processor are interconnected via a line; the at least one processor invokes the instructions in the memory to cause the collaborative decision-making device for high-concurrency traffic scenarios to execute the steps of the above-described collaborative decision-making method for high-concurrency traffic scenarios.

[0007] The present invention also provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the steps of the above-described collaborative decision-making method for high-concurrency traffic scenarios.

[0008] The aforementioned collaborative decision-making method and related equipment for high-concurrency traffic scenarios dynamically divides the functional areas of road segments based on the monitored arrival traffic volume, obtaining the lengths of lane-changing zones and speed adjustment zones. It then groups and guides the speed calculations of vehicle queues entering the lane-changing zone, arranging them longitudinally in a staggered manner to generate lane-changing decisions. For vehicles entering the speed adjustment zone, it performs longitudinal trajectory planning to obtain speed control commands. Finally, it jointly optimizes the signal timing parameters of downstream intersections to obtain the green light duration for each signal phase. This method adapts to changes in traffic flow by dynamically adjusting the length of functional zones, achieves rapid lane-changing for multiple vehicles through collaborative lane-changing strategies, and reduces vehicle delays by jointly optimizing vehicle trajectories and signal timing. This effectively solves the problem that fixed-zone methods cannot adapt to dynamic changes in traffic flow, improving the efficiency of road segment space resource utilization and intersection throughput.

[0009] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained in accordance with the structures particularly pointed out in the description, claims and drawings.

[0010] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0011] Figure 1 This is a schematic diagram of the first embodiment of the collaborative decision-making method in a high-concurrency traffic scenario according to the present invention; Figure 2 This is a schematic diagram of the second embodiment of the collaborative decision-making method in a high-concurrency traffic scenario according to the present invention; Figure 3 This is a schematic diagram of one embodiment of the collaborative decision-making device in a high-concurrency traffic scenario according to the present invention; Figure 4 This is a schematic diagram of one embodiment of a collaborative decision-making device in a high-concurrency traffic scenario according to the present invention. Detailed Implementation

[0012] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0013] The terms "comprising" and "having," and any variations thereof, used in the embodiments of this invention are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the steps or units listed, but may optionally include other steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0014] To facilitate understanding of this embodiment, a collaborative decision-making method for high-concurrency traffic scenarios disclosed in this invention will first be described in detail. For example... Figure 1 As shown, this method includes the following steps: 101. Based on the monitored traffic volume arriving at the road segment, the functional areas of the road segment are dynamically divided to obtain the lengths of the vehicle lane-changing area and the speed adjustment area. In this embodiment, the step of dynamically dividing the functional areas of a road segment based on the monitored arrival traffic volume to obtain the lengths of the vehicle lane-changing area and the speed adjustment area includes: cumulatively updating the zoning index based on the relative change rate between the arrival traffic volume and the baseline traffic volume of the road segment; triggering a re-division of functional areas when the zoning index reaches a preset threshold; calculating the required lengths of the vehicle lane-changing area and the speed adjustment area based on the queue lengths of vehicles needing to change lanes and the waiting queue lengths of vehicles; and constructing and solving a linear programming model based on the constraint relationship between the total length of the road segment and the required lengths of the two functional areas to obtain the actual division lengths of the vehicle lane-changing area and the speed adjustment area.

[0015] Specifically, in a connected transportation environment, the functional areas of a road segment are virtually defined through vehicle-road cooperative communication, rather than relying on traditional physical markings or signs. The lane-changing zone refers to the area where vehicles can change lanes after entering the road segment from an upstream intersection. Within this zone, vehicles need to change from their entering lane to the target lane according to downstream turning requirements; for example, right-turning vehicles need to move to the outermost lane, and left-turning vehicles need to move to the innermost lane. The speed adjustment zone follows immediately after the lane-changing zone, extending to the stop line at the downstream intersection. Within this zone, vehicles have already completed their lane change, and their main task is to adjust their speed to match the signal timing at the downstream intersection.

[0016] The road segment controller collects arrival traffic volume data for each turning lane group every preset monitoring period (typically set to 5 minutes). This data covers the number of vehicles turning left, going straight, and turning right from the upstream intersection. After obtaining the traffic volume data for the current period, the system compares it with the baseline traffic volume, which is the value recorded when the area was last reclassified. The calculation method is to subtract the baseline traffic volume from the current traffic volume, take the absolute value, and then divide by the baseline traffic volume to obtain the relative change rate. This relative change rate is compared with a pre-configured threshold (e.g., 0.17). If it is greater than or equal to the threshold, the zoning index is incremented by 1; otherwise, it is reset to zero. This design prevents frequent adjustments caused by random fluctuations in traffic volume, and confirms the trend by accumulating multiple periods.

[0017] When the zoning index reaches a preset value (e.g., 3), it means that traffic volume changes have become significant and sustained, triggering a re-division of functional zones. After triggering, the traffic volume for the current period will be updated to a new baseline value for subsequent monitoring and comparison.

[0018] When redividing the intersection, the required length of each functional zone must first be calculated. The required length of the lane-changing zone depends on the queuing of vehicles needing to change lanes, taking into account the number of various turning vehicles entering the road segment, the length of each vehicle, and the safe distance between vehicles. The required length of the speed adjustment zone depends on the queuing of vehicles waiting to pass through the intersection, including the number of vehicles in each lane and the space required for vehicles to decelerate to a stop.

[0019] Since the total length of the road segment is fixed, the sum of the actual lengths of the two zones must equal the total road segment length. If the sum of the two required lengths does not exceed the road segment length, then the length of the lane-changing zone should be compressed as much as possible to leave more space for the speed adjustment zone; if the sum of the required lengths exceeds the required length, then the requirements of the lane-changing zone should be met as much as possible while ensuring that the speed adjustment zone is not less than its requirement. By processing these constraints using a linear programming model, the final partitioning result can be calculated.

[0020] After calculating the functional area division results, the road segment controller sends the current functional area division information to vehicles entering the road segment via roadside units (RSUs) using V2I communication. The message includes parameters such as the starting position and length of the lane-changing zone and the starting position and length of the speed adjustment zone. Upon receiving this information, connected autonomous vehicles use their onboard positioning system (such as GPS or BeiDou) to determine their current functional area and execute corresponding driving strategies. For connected human-driven vehicles, the onboard terminal will remind the driver of their current functional area via voice prompts or a head-up display, suggesting appropriate driving actions.

[0021] Furthermore, the step of cumulatively updating the zoning index based on the relative change rate between the arrival traffic volume and the baseline traffic volume of the road segment, and triggering the re-division of functional areas when the zoning index reaches a preset threshold, includes: collecting the arrival traffic volume of each turning lane group of the road segment according to a preset monitoring period, and calculating the ratio of the absolute value of the difference between the arrival traffic volume and the baseline traffic volume in the current monitoring period to the baseline traffic volume to obtain the relative change rate; updating the zoning index based on the comparison result of the relative change rate and the preset change rate threshold; and determining whether to trigger the re-division of functional areas based on the comparison result of the zoning index and the preset index threshold, wherein when the zoning index reaches the index threshold, the arrival traffic volume of the current monitoring period is updated to the new baseline traffic volume and the re-division is triggered.

[0022] Specifically, the road segment controller collects traffic volume data at pre-configured time intervals, typically 5 minutes in practice. However, this interval can be adjusted based on the traffic characteristics of different road segments. For example, it can be shortened for segments with drastic traffic flow changes and lengthened for segments with relatively stable traffic. At the end of each monitoring cycle, the controller uses sensors or cameras deployed on the road segment to count the number of various types of vehicles entering the road segment from upstream intersections. This count is categorized by turning demand, including left-turning vehicles, straight-going vehicles, and right-turning vehicles. The sum of these three types of vehicle counts yields the arrival traffic volume for the current monitoring cycle.

[0023] After obtaining this traffic volume data, it needs to be compared with the baseline traffic volume stored in the system. The baseline traffic volume is set to a very small number when the system is first started, such as 1, to avoid a zero denominator in division operations. Subsequently, after each trigger and reclassification, the baseline traffic volume is updated to the traffic volume value at the time of the trigger. The formula for calculating the relative rate of change is quite simple: subtract the baseline traffic volume from the current traffic volume, take the absolute value of the result, and then divide by the baseline traffic volume.

[0024] After calculating the relative rate of change, this value is compared with a rate of change threshold. The rate of change threshold is generally determined based on historical data statistical analysis; in this embodiment, it can be set to 0.17. The comparison logic is as follows: when the relative rate of change is greater than or equal to 0.17, the traffic volume is considered to have undergone a noteworthy change, and the current value of the zoning index is incremented by 1; if the relative rate of change is less than 0.17, it indicates that the traffic volume fluctuation is not significant and may just be normal random fluctuation, in which case the zoning index is directly reset to zero. This mechanism's design borrows from the filtering concept in signal processing; a single fluctuation will not immediately trigger a system response, but only multiple consecutive significant changes will trigger subsequent operations.

[0025] After the zoning index is updated, the next step is to determine whether it has reached the preset index threshold. In this embodiment, the index threshold is set to 3, meaning that a significant change in traffic volume needs to be detected for three consecutive monitoring cycles for the system to recognize it as a genuine trend rather than a random disturbance. If the zoning index is less than 3, no action is taken, and the system waits for the next monitoring cycle to continue observation; once the zoning index equals 3, the functional area re-division process is triggered.

[0026] When a re-division is triggered, two actions must be performed simultaneously. First, the arriving traffic volume for the current monitoring period is recorded and updated to a new baseline traffic volume for calculating the relative rate of change in subsequent periods. Second, the functional area division calculation module is invoked to recalculate the lengths of the lane-changing zone and speed adjustment zone based on the current traffic conditions. It's important to note that after the re-division is triggered, the zoning index will also be reset to 0, and the cumulative counting will restart.

[0027] This cumulative judgment-based triggering mechanism performs relatively stably in actual operation. For example, at the beginning of the morning rush hour, traffic volume rises rapidly from the low level at night, and a large relative rate of change is detected for several consecutive cycles. The zoning index quickly accumulates to 3, triggering timely zoning adjustments. During the peak plateau period, although traffic volume is high, it is relatively stable, and the relative rate of change remains small. The zoning index does not rise, and the system maintains the current zoning. This allows for a rapid response to significant changes in traffic flow while avoiding frequent adjustments due to short-term fluctuations, achieving a good balance between response speed and stability.

[0028] 102. Based on the length of the vehicle lane-changing zone, calculate the grouped guidance speed of the vehicle queues entering the vehicle lane-changing zone to obtain the target guidance speed of different vehicle queues, and arrange each vehicle queue longitudinally in a staggered manner according to the target guidance speed to generate vehicle lane-changing decisions. In this embodiment, when a vehicle enters the road segment from the upstream intersection, the road segment controller groups these vehicles according to their lane position upon entry. Generally, vehicles entering both the outer and inner lanes are assigned to the first vehicle queue group, while vehicles entering only the inner lane are assigned to the second vehicle queue group. This grouping method takes into account the differences in lane-changing needs of vehicles in different lane positions. Vehicles in the outer lane need to cross multiple lanes to turn left, while vehicles in the inner lane have a relatively shorter lane-changing distance.

[0029] After the vehicles are grouped, the system needs to calculate the actual queue length for each queue group. This calculation involves adding the length of each vehicle to the safe following distance between vehicles to obtain the longitudinal length of the entire queue on the road. For example, if the first queue group has 10 vehicles, with an average length of 5 meters per vehicle and a safe following distance of 2 meters, then the queue length would be 10 × (5 + 2) = 70 meters. Once the length data for each queue group is obtained, combined with the total length of the lane-changing area, the appropriate guidance speed for each queue group can be calculated.

[0030] The goal of the guidance speed calculation is to create a staggered arrangement between different lane groups in the longitudinal space, thus providing sufficient safety clearance for vehicles needing to change lanes. Specifically, a relatively high guidance speed is set for the first lane group, while a lower speed is used for the second lane group. The difference between the two speeds is determined by the length of the lane-changing zone; the longer the lane-changing zone, the larger the speed difference can be set, and the more lenient the conditions for vehicles to complete the staggered arrangement. The calculation must also consider the maximum and minimum speed limits of the road; the guidance speed cannot exceed these limits.

[0031] After each queuing group travels at a different guided speed for a period of time, a longitudinal offset will occur between the queuing groups. The system monitors this offset distance in real time, and when the offset reaches a preset threshold, it is considered that a sufficient staggered arrangement has been formed. At this point, for vehicles whose current lane does not meet the downstream turning requirements, if the longitudinal and lateral distances between them and vehicles in the target lane meet the safe lane-changing conditions, the system will generate a lane-changing decision for these vehicles. The lane-changing decision includes information such as the target lane number, the suggested lane-changing timing, and the speed to be maintained during the lane-changing process. This information is sent to the relevant vehicles via vehicle-to-infrastructure communication.

[0032] It's important to note that during the entire process, if a vehicle is already in a lane that meets its turning requirements, the system will not generate a lane-changing instruction for it. This vehicle simply needs to maintain the guide speed by following the other vehicles in the queue. Furthermore, for manually driven connected vehicles, the lane-changing decisions generated by the system are provided to the driver as suggestions, not mandatory. The driver can decide whether to adopt the suggestion based on the actual situation.

[0033] 103. Based on the length of the speed adjustment zone and the execution result of the vehicle lane-changing decision, perform longitudinal trajectory planning for the vehicle entering the speed adjustment zone to obtain the vehicle speed control command. In this embodiment, the process of performing longitudinal trajectory planning for vehicles entering the speed adjustment zone based on the length of the speed adjustment zone and the execution result of the vehicle lane-changing decision to obtain vehicle speed control instructions includes: calculating the shortest time required for each vehicle to pass the stop line without considering the influence of the vehicle in front, based on the vehicle type, current position, current speed, and number of vehicles queuing ahead; classifying vehicles into a set of vehicles that can pass and a set of vehicles that need to wait, based on the relationship between the shortest time and the remaining green light time of the current signal phase; and generating longitudinal speed control instructions for each vehicle based on the set to which the vehicle belongs and whether there are vehicles queuing ahead.

[0034] Specifically, after a vehicle completes a lane-changing maneuver and enters the speed adjustment zone, the intersection controller begins longitudinal trajectory planning for these vehicles. The first step in planning is to acquire basic status information for each vehicle, including vehicle type, current location, real-time speed, and the queue situation ahead. The vehicle type is primarily distinguished between connected autonomous vehicles and connected human-driven vehicles. These two types of vehicles differ in reaction time; autonomous vehicles typically have a reaction time of around 0.1 seconds, while human-driven vehicles generally have a reaction time of 1 second or even longer. The current location can be obtained through the onboard positioning system, which calculates the actual distance between the vehicle and the stop line at the downstream intersection. The current speed is directly provided by the vehicle's speed sensor, while the number of vehicles in the queue ahead needs to be counted through roadside equipment or feedback from the vehicle in front.

[0035] Based on this information, the controller calculates the shortest time required for each vehicle to cross the stop line. The calculation considers the vehicle's acceleration performance and road speed limits, assuming the vehicle accelerates at its maximum permissible acceleration, maintains a constant speed once it reaches the road speed limit or the vehicle's performance limit, and finally decelerates appropriately to reach a safe speed to cross the intersection. The entire process involves the cumulative time of the acceleration, constant speed, and deceleration phases, plus the time required to travel the remaining distance at a constant speed. It's important to note that this shortest time calculation does not currently consider the impact of vehicles queuing ahead; it is an estimate under ideal conditions.

[0036] After calculating the shortest time, it needs to be compared with the remaining green light time for the current signal phase. The remaining green light time equals the total green light duration for that phase minus the time already used. For example, if the total green light duration for a certain phase is 45 seconds, and 20 seconds have already passed, then the remaining green light time is 25 seconds. If the vehicle's shortest passage time is less than or equal to the remaining green light time, it means that the vehicle has a chance to pass through the intersection before the green light ends, and it is included in the group of vehicles that can pass. Conversely, if the shortest passage time is greater than the remaining green light time, it means that even under ideal circumstances, it will not be able to make it to the green light, and it is placed in the group of vehicles that need to wait.

[0037] The system employs different speed control strategies for vehicles in these two sets. For vehicles in the passable vehicle set, it further determines whether there is a queue of vehicles ahead. If the road ahead is clear, or the distance to the vehicle in front is large enough, a control command to accelerate through can be generated, instructing the vehicle to accelerate to a higher speed, then maintain a constant speed for a period, and finally decelerate to a suitable speed for crossing the intersection. This trajectory allows vehicles to pass through intersections quickly, improving traffic efficiency. However, if there is already a queue of vehicles ahead, or the distance to the vehicle in front is relatively close, acceleration cannot be arbitrarily increased. In this case, the generated control command will require the vehicle to maintain a safe distance from the vehicle in front, and accelerate only when the vehicle in front starts moving or the distance increases.

[0038] For vehicles waiting in a queue, the control instructions are entirely different. Since these vehicles can't catch the current green light, there's no need to maintain high speed until the stop line and then brake suddenly; that's both uncomfortable and energy-consuming. A more reasonable approach is to reduce speed in advance, approaching the stop line slowly at a lower speed, and waiting for the next green light. The specific instructions will instruct the vehicle to first decelerate to a preset minimum speed, such as 4 meters per second, and then maintain this low speed until the light turns green before accelerating through. This control method avoids frequent sudden stops and starts, improving ride comfort and reducing energy consumption.

[0039] In practice, connected autonomous vehicles receive speed control commands and adjust the throttle and brakes accordingly. Connected human-driven vehicles, on the other hand, provide suggested speeds to the driver via an onboard display or voice prompts, allowing the driver to decide whether to adopt the suggested speed. Throughout the speed adjustment zone, the controller continuously tracks the vehicle's performance, updating control commands in real time if traffic signals change or unexpected situations arise.

[0040] 104. Based on the vehicle speed control command and the estimated time for the vehicle to pass the stop line, the signal timing parameters of the downstream intersection are jointly optimized to obtain the green light duration for each signal phase.

[0041] In this embodiment, the joint optimization of signal timing parameters at the downstream intersection based on the vehicle's speed control command and the estimated time for the vehicle to cross the stop line, to obtain the green light duration for each signal phase, includes: constructing the state space and control variable space of a dynamic programming model based on the green light duration constraints for each signal phase; based on the state space and control variable space, combined with the vehicle's speed control command and the estimated time to cross the stop line, obtaining the optimal value function and optimal control variable for each state by traversing each state and control variable combination and calculating vehicle delay and comfort indices; and obtaining the green light duration for each signal phase through the recursive and backtracking process of dynamic programming based on the optimal value function and optimal control variable.

[0042] Specifically, intersection controllers use dynamic programming to optimize signal timing. The basic idea of ​​dynamic programming is to break down the entire signal cycle into several stages, with each signal phase corresponding to one stage, and then establish a recursive relationship between these stages. In a typical four-phase intersection, the phase sequence is generally a cycle of left turn, straight ahead, left turn, straight ahead. Each phase ends with a yellow light and a full red light transition period; these transition times are fixed and not included in the optimization.

[0043] When constructing a dynamic programming model, state variables need to be defined first. In this embodiment, the state variable represents the total accumulated duration starting from the first phase. For example, at the end of the first phase, the state variable is equal to the green light duration of the first phase plus its transition time; at the end of the second phase, the state variable is the sum of the green light durations of the first two phases plus the two transition times, and so on. The state space is the set of all possible accumulated durations, and the range of this set is determined by the minimum and maximum green light durations of each phase. For example, if the green light duration of each phase must be at least 12 seconds and no more than 60 seconds, then at the end of the fourth phase, the value range of the state variable is between the minimum and maximum values ​​allocated to each of the four phases.

[0044] The control variable is the actual green light duration allocated to each phase. For each state, there are multiple possible values ​​for the control variable. These values ​​must satisfy two constraints: one is that the green light duration itself must be between the minimum and maximum values, and the other is that their sum must reach the total duration corresponding to the current state. Collecting all green light duration values ​​that satisfy the constraints constitutes the control variable space for that state.

[0045] Once the state space and control variable space are determined, the traversal calculations can begin. For each combination of control variables in each state, the system needs to evaluate the quality of this combination. The evaluation metrics include two aspects: vehicle delay and driving comfort. Vehicle delay is calculated by comparing the actual time a vehicle takes to cross the stop line with the ideal time. If a vehicle could have crossed quickly but is delayed due to traffic lights, this delay is counted. Comfort is measured by the square integral of the vehicle's acceleration; the more drastic the acceleration change, the larger the integral value, indicating poorer comfort. When calculating these two metrics, all vehicles that can cross the stop line in the current phase must be considered. Their delay and comfort metrics are summed separately, and then weighted according to preset weighting coefficients to obtain the performance metrics for this state and control variable combination.

[0046] The core of dynamic programming lies in utilizing the recursive relationships between stages. For a certain state in the current stage, its optimal value function is equal to the value that minimizes the sum of "the performance index of the current stage plus the optimal value function of the corresponding state in the previous stage" among all possible control variables. In this way, starting from the first phase and working backward, the optimal value function of each state in each phase is calculated one by one, while recording the control variables used to reach the optimal value.

[0047] Once all phase calculations are complete, the backtracking phase begins. Backtracking starts from the optimal state of the last phase and proceeds backward, searching each phase one by one. Among all possible states of the last phase, the state with the smallest optimal value function is selected as the endpoint. Then, based on previously recorded information, the state is determined as to which previous phase it transitioned from, and the duration of the green light used during that transition. This process is repeated until the first phase is reached, thus determining the optimal green light duration for each phase.

[0048] It should be noted that this optimization process is generally performed near the end of each signal cycle, utilizing the brief interval between the yellow and all-red light periods to complete the calculation. The calculated green light duration will take effect in the next cycle, and this duration information will also be sent to vehicles about to enter the speed adjustment zone via the vehicle-to-infrastructure (V2I) network, allowing vehicle trajectory planning and signal timing to coordinate and achieve a better collaborative effect.

[0049] In this embodiment, the functional areas of a road segment are dynamically divided based on the monitored arrival traffic volume to obtain the lengths of the lane-changing zone and the speed adjustment zone. Vehicle queues entering the lane-changing zone are grouped, guided by speed calculations, and longitudinally staggered to generate lane-changing decisions. Vehicles entering the speed adjustment zone undergo longitudinal trajectory planning to obtain speed control commands. The signal timing parameters of downstream intersections are jointly optimized to obtain the green light duration for each signal phase. This method adapts to changes in traffic flow by dynamically adjusting the length of functional zones, achieves rapid lane changing for multiple vehicles through a collaborative lane-changing strategy, and reduces vehicle delays by jointly optimizing vehicle trajectories and signal timing. It effectively solves the problem that fixed-zone methods cannot adapt to dynamic changes in traffic flow, improving the efficiency of road segment space resource utilization and intersection throughput.

[0050] Please see Figure 2 Another embodiment of the collaborative decision-making method in high-concurrency traffic scenarios in this application includes: 201. Based on the monitored arrival traffic volume of the road segment, the functional areas of the road segment are dynamically divided to obtain the lengths of the vehicle lane-changing area and the speed adjustment area. In this embodiment, step 201 is similar to step 101, and will not be described again here.

[0051] 202. Based on the position of the vehicle in the lane when it enters the road section, divide the vehicles entering the lane changing area into multiple vehicle queue groups; In this embodiment, when the green light of a certain signal phase ends and a group of vehicles enters the road segment from the upstream intersection, the road segment controller immediately groups these vehicles. The grouping is based on the lane position of the vehicles when they first enter the road segment, rather than their final destination. This grouping method takes into account that vehicles in different lane positions face varying degrees of difficulty during lane changes, requiring different speed guidance strategies.

[0052] Specifically, suppose the road has three lanes, numbered 1, 2, and 3 from the outside in. Vehicles entering from lane 1 or lane 3 will be grouped into the first vehicle queue. This is because these two lanes are located at the edge of the road, and vehicles originating from these lanes often need to traverse a significant lateral distance to change lanes. For example, a right-turning vehicle entering from lane 1 might already be in a suitable lane; however, a left-turning vehicle would have to change lanes all the way to lane 3, making the lane-changing requirement quite strong. Conversely, left-turning vehicles entering from lane 3 generally don't need to change lanes, but right-turning vehicles face more difficulties.

[0053] The second vehicle queue group is relatively simpler to divide; it consists of vehicles entering from lane 2. This lane is located in the middle of the road, and vehicles starting from here don't have to travel too far to change lanes left or right, generally only needing to change lanes once to reach their destination. Therefore, they are grouped separately, and a slightly different speed guidance strategy can be adopted compared to the first group.

[0054] During the grouping process, the system acquires initial lane information for each vehicle through roadside detection devices. This information can come from intersection camera recognition or be reported by the vehicle itself via vehicle-to-infrastructure communication. Once a vehicle is determined to belong to a particular queue group, this grouping information, along with other vehicle attributes, is recorded in the system as input data for subsequent speed guidance calculations.

[0055] It should be noted that this grouping method still offers some flexibility in practical applications. If the road has only two lanes, the grouping logic simplifies to one group for the outer lanes and another for the inner lanes. If it's a four-lane or wider road, a third queue might be created, grouping the middle one or two lanes separately. The core idea is to determine the difficulty of lane changing based on lane position, grouping vehicles with similar difficulty into the same group for easier unified management.

[0056] After grouping, the system will count the number of vehicles in each queue group. This data will affect the subsequent calculation of queue length. For example, if the first queue group has 15 vehicles and the second queue group has 8 vehicles, the estimated road length occupied by the entire queue will be calculated by multiplying the number of vehicles by the length of each vehicle and the spacing between them. In addition, if a queue group has a particularly large number of vehicles, the system will also take this factor into account when setting the guidance speed to avoid queues becoming too long and exceeding the lane-changing area due to improper speed settings.

[0057] 203. Based on the queue length of each vehicle queue group and the length of the vehicle lane-changing zone, calculate the target guidance speed between each vehicle queue group so that different vehicle queue groups are staggered in the longitudinal space. In this embodiment, after grouping the vehicles, the system needs to set appropriate driving speeds for different queues, allowing them to gradually increase the longitudinal distance as they move forward. The key to this speed setting is to ensure sufficient stagger between queues to create space for lane changes, while preventing the speed difference from being too large, which would cause some vehicles to travel too slowly or too fast, affecting overall traffic efficiency.

[0058] The starting point for speed calculation is determining the actual length of each queue group. Taking the first queue group as an example, assuming it contains 12 vehicles, each with an average length of 5 meters, and a safe distance of 2 meters between vehicles, the total length of the queue from beginning to end is 12 multiplied by 7, equaling 84 meters. The second queue group, with 6 vehicles, is calculated similarly to be 42 meters. This queue length data is crucial because determining whether the two queues are completely staggered involves checking if the longitudinal distance between the lead vehicle of the first queue group and the lead vehicle of the second queue group is approximately half the length of the first queue group.

[0059] After determining the target offset distance, the next step is to calculate the speed. The calculation method is as follows: the first and second queue groups start from the same starting point. If the first group's speed is slightly faster and the second group's speed is slightly slower, then after a period of time, the lead vehicle of the first group will overtake the lead vehicle of the second group. How far it overtakes depends on the speed difference and the travel time. Since the length of the lane-changing zone is known, we can calculate the speed difference between the two queues within that zone to achieve the desired offset effect.

[0060] In practice, the system typically assigns a guiding speed close to the road's maximum speed limit to the first queue group. For example, if the speed limit is 13 meters per second, the first queue group will run at this speed. The second queue group is set to a significantly lower speed, calculated using a formula. This formula takes into account factors such as the length of the lane-changing area, the target offset distance, and the speed of the first queue group. If the calculated speed of the second queue group is lower than the minimum speed limit, for example, 4 meters per second, then this lower limit is used, and the speed of the first queue group is adjusted accordingly to ensure that a necessary speed difference is maintained between the two queue groups.

[0061] Sometimes, a situation arises where all vehicles in the second queue haven't fully entered the road segment. In this case, speed calculation is divided into two phases. In the first phase, the first queue maintains its initial speed from when it exited the upstream intersection, and the second queue also maintains its initial speed. Formal speed guidance only begins after the last vehicle in the second queue enters the road segment. This is done to avoid speed inconsistencies within the queue. If some vehicles are already decelerating while others are still accelerating out of the intersection, unnecessary disturbances will occur within the queue.

[0062] Once the speed is set, the system will send the guidance speed information to the vehicles in each queue group. For connected autonomous vehicles, these speed values ​​will be directly input into the vehicle's control system as the target speed; for connected human-driven vehicles, the onboard terminal will display a suggested speed, reminding the driver to try to stay close to this speed.

[0063] 204. Based on the longitudinal staggered arrangement of each vehicle queue group, generate vehicle lane-changing decisions to meet lane-changing safety conditions. In this embodiment, the process of generating vehicle lane-changing decisions to meet lane-changing safety conditions based on the longitudinal misalignment of each vehicle queue group includes: determining whether a preset lane-changing initiation threshold has been reached based on the longitudinal misalignment distance between each vehicle queue group; when the lane-changing initiation threshold is reached, generating a first lane-changing decision from the entry lane to the target lane for vehicles that have entered the lane-changing area and whose current lane group does not meet the downstream turning requirements, wherein the target lane is an adjacent lane that meets the vehicle turning requirements; and calculating the deviation between the number of vehicles in each lane and the average number of vehicles in the same lane group based on the vehicle number distribution after the first lane-changing, generating a second lane-changing decision for some vehicles in lanes with a greater number of vehicles than the average number of vehicles to change lanes in lanes with a less than average number of vehicles.

[0064] Specifically, after two vehicle platoons have traveled at different guide speeds for a period of time, the system calculates the longitudinal misalignment distance between them in real time. This distance refers to the difference in the longitudinal direction between the lead vehicle of the first platoon and the lead vehicle of the second platoon. The calculation uses the real-time position data of the two vehicles, which can be obtained through the vehicle positioning system. Assuming that the lead vehicle of the first platoon has traveled to a position 80 meters from the start of the road segment, while the lead vehicle of the second platoon has only traveled 50 meters, then the longitudinal misalignment distance is 30 meters.

[0065] The system compares this real-time calculated misalignment distance with a pre-set lane-change initiation threshold. The lane-change initiation threshold is typically set to about half the maximum queue length of the first queue group. For example, if the maximum length of the first queue group is calculated to be 84 meters, the threshold can be set to 42 meters. When the actual misalignment distance reaches or exceeds this threshold, it means that a sufficient gap has been formed between the two queues, and vehicles can begin changing lanes.

[0066] The initial lane change is primarily to accommodate the turning needs of vehicles. The system checks whether each vehicle's current lane supports its turning intention at the downstream intersection. If not, a lane change is required. For example, if a vehicle intending to turn left is currently in the outermost lane, its target lane is the adjacent innermost lane, requiring it to merge to the left. If it needs to continue turning left, it may need to change lanes twice to reach the innermost left-turn lane. However, during the initial lane change phase, it generally only involves changing to an adjacent lane to avoid the risks associated with crossing multiple lanes at once.

[0067] When generating a lane-change decision, in addition to checking if the longitudinal misalignment meets the standard, it is also necessary to confirm the lateral and longitudinal safety clearances. Lateral safety clearance refers to the sufficient gap between vehicles in front and behind in the target lane, at least large enough to accommodate the length of the lane-changing vehicle plus the front and rear safety buffer distances. Longitudinal safety clearance refers to the distance maintained between the lane-changing vehicle and adjacent vehicles in the target lane in the longitudinal direction to prevent rear-end collisions during the lane change. Only when both conditions are met will the system issue a lane-change instruction to the vehicle.

[0068] The lane-change instruction includes the target lane number, the suggested timing for the lane change, and the speed to be maintained during the lane change. After receiving the instruction, the connected autonomous vehicle will automatically execute the lane change operation, while the driver of the connected human-driven vehicle will see a prompt message on the dashboard or head-up display, with text such as "It is recommended to change lanes to the left to lane 2," accompanied by a directional arrow icon.

[0069] After the first lane change is completed, the distribution of vehicles in each lane may become uneven. For example, if a straight-ahead lane group contains two lanes, with 10 vehicles in the first lane and 6 vehicles in the second lane, the average number of vehicles per lane should be 8. If the system detects that the first lane has 2 more vehicles than the average and the second lane has 2 fewer vehicles, it will initiate a second lane change to balance this difference.

[0070] The second lane change doesn't consider the turning needs of vehicles because after the first lane change, all vehicles are already in the lane group that matches their turning intentions. This lane change is purely to make the number of vehicles in each lane within the same lane group as close as possible, avoiding one lane being too congested while adjacent lanes are empty. Vehicles choosing to change lanes generally start from the end of the more congested lane, moving to the less congested lane. Using the previous example, the last two vehicles in the first lane will receive an instruction to merge into the second lane. After this adjustment, both lanes have eight vehicles each, achieving a relatively ideal equilibrium.

[0071] 205. Based on the length of the speed adjustment zone and the execution result of the vehicle lane-changing decision, perform longitudinal trajectory planning for the vehicle entering the speed adjustment zone to obtain the vehicle speed control command. 206. Based on the vehicle speed control command and the estimated time for the vehicle to pass the stop line, the signal timing parameters of the downstream intersection are jointly optimized to obtain the green light duration for each signal phase.

[0072] In this embodiment, steps 205-206 are similar to steps 103-104, and will not be described again here.

[0073] In this embodiment, the functional areas of a road segment are dynamically divided based on the monitored arrival traffic volume to obtain the lengths of the lane-changing zone and the speed adjustment zone. Vehicle queues entering the lane-changing zone are grouped, guided by speed calculations, and longitudinally staggered to generate lane-changing decisions. Vehicles entering the speed adjustment zone undergo longitudinal trajectory planning to obtain speed control commands. The signal timing parameters of downstream intersections are jointly optimized to obtain the green light duration for each signal phase. This method adapts to changes in traffic flow by dynamically adjusting the length of functional zones, achieves rapid lane changing for multiple vehicles through a collaborative lane-changing strategy, and reduces vehicle delays by jointly optimizing vehicle trajectories and signal timing. It effectively solves the problem that fixed-zone methods cannot adapt to dynamic changes in traffic flow, improving the efficiency of road segment space resource utilization and intersection throughput.

[0074] The collaborative decision-making method for high-concurrency traffic scenarios in this embodiment of the invention has been described above. The collaborative decision-making device for high-concurrency traffic scenarios in this embodiment of the invention is described below. Please refer to [link to relevant documentation] for details on this collaborative decision-making device for high-concurrency traffic scenarios. Figure 3One embodiment of the collaborative decision-making device for high-concurrency traffic scenarios in this invention includes: The dynamic partitioning module 301 is used to dynamically divide the functional areas of the road segment according to the monitored traffic volume of the road segment, and obtain the length of the vehicle lane changing area and speed adjustment area. The collaborative lane-changing module 302 is used to calculate the grouping guidance speed of the vehicle queues entering the vehicle lane-changing area based on the length of the vehicle lane-changing area, obtain the target guidance speed of different vehicle queues, and arrange each vehicle queue longitudinally in a staggered manner according to the target guidance speed to generate vehicle lane-changing decisions. The trajectory planning module 303 is used to perform longitudinal trajectory planning for vehicles entering the speed adjustment zone based on the length of the speed adjustment zone and the execution result of the vehicle lane change decision, so as to obtain the vehicle speed control command. The signal optimization module 304 is used to jointly optimize the signal timing parameters of the downstream intersection based on the vehicle speed control command and the estimated time for the vehicle to pass the stop line, so as to obtain the green light duration of each signal phase.

[0075] In this embodiment of the invention, the collaborative decision-making device for high-concurrency traffic scenarios operates the aforementioned collaborative decision-making method for high-concurrency traffic scenarios. The device dynamically divides the functional areas of a road segment based on the monitored traffic volume, obtaining the lengths of the lane-changing zone and speed adjustment zone. It then groups and guides the speed calculations of vehicle queues entering the lane-changing zone, arranging them longitudinally in a staggered manner to generate lane-changing decisions. For vehicles entering the speed adjustment zone, it performs longitudinal trajectory planning to obtain speed control commands. Finally, it jointly optimizes the signal timing parameters of downstream intersections to obtain the green light duration for each signal phase. This method adapts to changes in traffic flow by dynamically adjusting the length of functional zones, achieves rapid lane changing for multiple vehicles through a collaborative lane-changing strategy, and reduces vehicle delays by jointly optimizing vehicle trajectories and signal timing. This effectively solves the problem that fixed-zone methods cannot adapt to dynamic changes in traffic flow, improving the efficiency of road segment space resource utilization and intersection throughput.

[0076] above Figure 3 The collaborative decision-making device in high-concurrency traffic scenarios in this embodiment of the invention is described in detail from the perspective of unitized functional entities. The collaborative decision-making device in high-concurrency traffic scenarios in this embodiment of the invention is described in detail from the perspective of hardware processing.

[0077] Figure 4This is a schematic diagram of the structure of a collaborative decision-making device for high-concurrency traffic scenarios provided in an embodiment of the present invention. The collaborative decision-making device 400 for high-concurrency traffic scenarios can vary significantly due to different configurations or performance. It may include one or more central processing units (CPUs) 410 (e.g., one or more processors) and a memory 420, and one or more storage media 430 (e.g., one or more mass storage devices) for storing application programs 433 or data 432. The memory 420 and storage media 430 can be temporary or persistent storage. The program stored in the storage media 430 may include one or more units (not shown in the diagram), each unit may include a series of instruction operations on the collaborative decision-making device 400 for high-concurrency traffic scenarios. Furthermore, the processor 410 may be configured to communicate with the storage media 430 and execute a series of instruction operations in the storage media 430 on the collaborative decision-making device 400 for high-concurrency traffic scenarios to implement the steps of the collaborative decision-making method for high-concurrency traffic scenarios described above.

[0078] The collaborative decision-making device 400 for high-concurrency traffic scenarios may also include one or more power supplies 440, one or more wired or wireless network interfaces 450, one or more input / output interfaces 460, and / or one or more operating systems 431, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 4 The illustrated collaborative decision-making device structure for high-concurrency traffic scenarios does not constitute a limitation on the collaborative decision-making device for high-concurrency traffic scenarios provided by the present invention. It may include more or fewer components than illustrated, or combine certain components, or have different component arrangements.

[0079] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the collaborative decision-making method in the high-concurrency traffic scenario.

[0080] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system, device, or unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0081] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0082] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A collaborative decision-making method for high-concurrency traffic scenarios, characterized in that, The collaborative decision-making method for high-concurrency traffic scenarios includes: Based on the monitored traffic volume arriving at the road segment, the functional areas of the road segment are dynamically divided to obtain the lengths of the vehicle lane-changing zone and the speed adjustment zone. Based on the length of the lane-changing zone, the vehicle queues entering the lane-changing zone are grouped and guided to calculate the target guidance speed for different vehicle queues. Based on the target guidance speed, the vehicle queues are longitudinally staggered to generate a vehicle lane-changing decision. Based on the length of the speed adjustment zone and the execution result of the vehicle lane-changing decision, longitudinal trajectory planning is performed on the vehicles entering the speed adjustment zone to obtain the vehicle speed control command. Based on the vehicle speed control command and the estimated time for the vehicle to pass the stop line, the signal timing parameters of the downstream intersection are jointly optimized to obtain the green light duration for each signal phase.

2. The collaborative decision-making method for high-concurrency traffic scenarios according to claim 1, characterized in that, The process of dynamically dividing the functional areas of a road segment based on the monitored arrival traffic volume, resulting in the lengths of the lane-changing zone and speed adjustment zone, includes: The zoning index is cumulatively updated based on the relative change rate between the arrival traffic volume of the road segment and the baseline traffic volume. When the zoning index reaches a preset threshold, the functional area is re-divided. Based on the required queue length of vehicles changing lanes and the queue length of waiting vehicles, calculate the required length of the vehicle lane-changing area and the required length of the speed adjustment area respectively; Based on the constraint relationship between the total length of the road segment and the required length of the two functional zones, a linear programming model is constructed and solved to obtain the actual division length of the vehicle lane-changing zone and the speed adjustment zone.

3. The collaborative decision-making method for high-concurrency traffic scenarios according to claim 2, characterized in that, The step of cumulatively updating the zoning index based on the relative change rate between the arrival traffic volume and the baseline traffic volume of a road segment, and triggering the re-division of functional areas when the zoning index reaches a preset threshold, includes: According to the preset monitoring cycle, the arrival traffic volume of each turning lane group of the road segment is collected, and the ratio of the absolute value of the difference between the arrival traffic volume of the current monitoring cycle and the baseline traffic volume to the baseline traffic volume is calculated to obtain the relative change rate. The partition index is updated based on the comparison between the relative rate of change and the preset rate of change threshold. Based on the comparison result between the zoning index and the preset index threshold, it is determined whether to trigger the re-division of functional areas. When the zoning index reaches the index threshold, the arrival traffic volume of the current monitoring period is updated to the new baseline traffic volume and the re-division is triggered.

4. The collaborative decision-making method for high-concurrency traffic scenarios according to claim 1, characterized in that, The process of calculating the grouping guidance speed for vehicle queues entering the lane-changing zone based on the length of the lane-changing zone, obtaining the target guidance speed for different vehicle queues, and then longitudinally staggering the vehicle queues according to the target guidance speed to generate vehicle lane-changing decisions includes: Based on the position of the vehicle in the lane when it enters the road segment, vehicles entering the lane-changing area are divided into multiple vehicle queue groups. Based on the queue length of each vehicle queue group and the length of the vehicle lane change zone, the target guidance speed between each vehicle queue group is calculated so that different vehicle queue groups are staggered in the longitudinal space. Based on the longitudinal staggered arrangement of each vehicle queue group, a vehicle lane-changing decision is generated to meet the lane-changing safety conditions.

5. The collaborative decision-making method for high-concurrency traffic scenarios according to claim 4, characterized in that, The process of generating lane-changing decisions for vehicles to meet lane-changing safety conditions based on the longitudinal staggered arrangement of each vehicle queue group includes: Based on the longitudinal misalignment distance between each vehicle queue group, it is determined whether the preset lane change initiation threshold has been reached. When the lane change initiation threshold is reached, the first lane change decision is generated for vehicles that have entered the vehicle lane change area and whose current lane group does not meet the downstream turning requirements, from the entry lane to the target lane, wherein the target lane is an adjacent lane that meets the vehicle turning requirements. Based on the vehicle distribution in each lane after the first lane change, the deviation between the number of vehicles in each lane in the same lane group and the average number of vehicles in that lane group is calculated. A second lane change decision is then generated for some vehicles in lanes with a greater number of vehicles than the average number of vehicles, allowing them to switch to lanes with a less than average number of vehicles.

6. The collaborative decision-making method for high-concurrency traffic scenarios according to claim 1, characterized in that, The process of planning the longitudinal trajectory of vehicles entering the speed adjustment zone based on the length of the speed adjustment zone and the execution result of the vehicle lane-changing decision, to obtain the vehicle speed control command, includes: Based on the vehicle type entering the speed adjustment zone, current position, current speed, and number of vehicles in the queue ahead, calculate the shortest time required for each vehicle to cross the stop line without considering the influence of the vehicle in front. Based on the relationship between the shortest time and the remaining green light time of the current signal phase, vehicles are divided into a group of vehicles that can pass and a group of vehicles that need to wait. Based on the vehicle's group and whether there are vehicles queuing ahead, longitudinal speed control commands are generated for each vehicle.

7. The collaborative decision-making method for high-concurrency traffic scenarios according to claim 1, characterized in that, The process of jointly optimizing the signal timing parameters of the downstream intersection based on the vehicle's speed control command and the estimated time for the vehicle to pass the stop line yields the green light duration for each signal phase, including: Based on the green light duration constraints for each signal phase, construct the state space and control variable space of the dynamic programming model; Based on the state space and control variable space, combined with the vehicle speed control command and the estimated time to cross the stop line, the optimal value function and optimal control variable for each state are obtained by traversing each state and control variable combination and calculating the vehicle delay and comfort index. Based on the optimal value function and optimal control variables, the green light duration for each signal phase is obtained through the recursive and backtracking process of dynamic programming.

8. A collaborative decision-making device for high-concurrency traffic scenarios, characterized in that, The collaborative decision-making device for high-concurrency traffic scenarios includes: The dynamic zoning module is used to dynamically divide the functional areas of a road segment based on the monitored traffic volume, and to obtain the lengths of the lane-changing zone and the speed adjustment zone. The collaborative lane-changing module is used to calculate the grouping guidance speed of vehicle queues entering the vehicle lane-changing area based on the length of the vehicle lane-changing area, obtain the target guidance speed of different vehicle queues, and arrange each vehicle queue longitudinally in a staggered manner according to the target guidance speed to generate vehicle lane-changing decisions. The trajectory planning module is used to perform longitudinal trajectory planning for vehicles entering the speed adjustment zone based on the length of the speed adjustment zone and the execution result of the vehicle lane change decision, so as to obtain the vehicle speed control command. The signal optimization module is used to jointly optimize the signal timing parameters of the downstream intersection based on the vehicle speed control command and the estimated time for the vehicle to pass through the stop line, so as to obtain the green light duration for each signal phase.

9. A collaborative decision-making device for high-concurrency traffic scenarios, characterized in that, The collaborative decision-making device for high-concurrency traffic scenarios includes: a memory and at least one processor, wherein the memory stores instructions; The at least one processor invokes the instructions in the memory to cause the collaborative decision-making device for high-concurrency traffic scenarios to perform the steps of the collaborative decision-making method for high-concurrency traffic scenarios as described in any one of claims 1-7.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When the instruction is executed by the processor, it implements the steps of the collaborative decision-making method for high-concurrency traffic scenarios as described in any one of claims 1-7.

Images