Intersection vehicle track and signal collaborative optimization method in hybrid intelligent network connection environment

Abstract

An intersection vehicle track and signal collaborative optimization method in a hybrid intelligent network connection environment belongs to the field of intelligent traffic, and comprises the following steps: defining an intersection optimization area, executing different signal optimization strategies for different defined areas, executing an enhanced adaptive signal optimization strategy for a signal control area, and executing a signal control area; executing a red light phase signal optimization strategy on the conflict identification area; establishing different transverse and longitudinal track optimization control models of the intelligent network connection vehicle and the manual driving vehicle according to different vehicle types and position relations; and iteratively updating the two signal optimization strategies and the vehicle transverse and longitudinal control strategy based on different iteration time intervals. According to the invention, the fuel economy and driving comfort of the vehicle can be improved on the basis of ensuring the maximum passing efficiency of the vehicle.

Description

Technical Field

[0001] The present invention relates to the field of intelligent network technology, and in particular to a method for collaboratively optimizing vehicle trajectories and signals at an intersection in a hybrid intelligent network environment. Background Art

[0002] Intersections are the main bottlenecks and key factors restricting the development of urban transportation systems, and are the main source of various traffic problems. On the one hand, although traditional signal control measures can improve the traffic efficiency of intersections to a certain extent, they often only use macro traffic flow index parameters and cannot give full play to the active coordination role of signal control. On the other hand, although unsignalized intersections also play a regulatory role in traffic operation and can effectively reduce traffic delays when the traffic volume is small, the existence of traffic conflicts makes it very easy to cause safety problems. In general, solving the traffic operation problem at intersections is the core of improving the efficiency and safety of urban road traffic, and it has become an imminent issue.

[0003] The development of intelligent network technology has triggered a revolutionary breakthrough in traditional traffic control systems. The "Intelligent Vehicle Innovation and Development Strategy" jointly issued by multiple national ministries in 2020 clarified the importance of the development of intelligent network technology and identified the research and development of intelligent vehicles as one of the important national development strategies. How to use intelligent network technology to achieve the transformation of intersections from "passive response control" to "active collaborative control" through real-time dynamic interaction between vehicles and roads is an urgent problem to be solved in the development of intelligent transportation. With the gradual increase in the proportion of intelligent networked vehicles on the road, intersections will be in a hybrid intelligent networked traffic environment for a long time. Due to the participation of heterogeneous traffic flows, the complexity of intersection control problems under hybrid intelligent network is relatively high. In the existing related research, the main problems are: First, most of the research on intersection signal optimization and vehicle trajectory collaborative control is highly centralized, and often requires the collection of traffic information of the entire intersection, which has high requirements for vehicle intelligence and real-time communication, making it difficult to meet the control needs of real traffic self-organization scenarios. Secondly, most of the research is on the traffic control problem of hybrid intelligent networked environment at single-entry intersections, ignoring the lateral traffic behavior of vehicles.

[0004] Therefore, in order to make up for the defects in previous studies, this study aims at the problem of intersection signal control in a hybrid intelligent connected environment and proposes a method for coordinated optimization of intersection vehicle trajectories and signals in a hybrid intelligent connected environment. Summary of the invention

[0005] The purpose of the present invention is to make up for the defects of the prior art and provide a method for coordinated optimization of vehicle trajectories and signals at intersections under a hybrid intelligent network environment, which improves vehicle fuel economy and driving comfort while ensuring vehicle traffic efficiency, and is used to solve the problem of lack of coordinated optimization of vehicle trajectories and signals at intersections under a hybrid intelligent network environment in the prior art.

[0006] A method for collaborative optimization of intersection vehicle trajectories and signals in a hybrid intelligent connected environment, the specific steps are:

[0007] Step 1: Define the intersection optimization area. Define the intersection as a two-way four-lane cross intersection with a through lane and a left-turn lane, where lane 1 is the through lane and lane 2 is the left-turn lane. Define the inner area of ​​the intersection as the conflict identification area, the entrance area as the signal control area, and further divide the signal control area into a free driving area and a no lane change area.

[0008] Step 2: Implement different signal optimization strategies for different areas. Implement the enhanced adaptive signal optimization strategy for the signal control area, and implement the red light phase signal optimization strategy for the conflict identification area. The specific method is as follows.

[0009] Step 2.1: Execute the enhanced adaptive signal optimization strategy for the signal control area. The specific steps are as follows.

[0010] Step 2.1.1: Update intersection signal timing and vehicle trajectory information, and determine whether the current green light phase exceeds the minimum green light time , where w represents the current lane number. If no, go to step 2.1.4; otherwise, go to step 2.1.2.

[0011] Step 2.1.2: Calculate the arrival time of vehicles in the lane served by the current green light phase, and determine whether all vehicles can pass through the intersection during the current green light phase. If yes, go to step 2.1.4; otherwise, go to step 2.1.3.

[0012] Step 2.1.3: Calculate whether the maximum green phase can be reached without violating the maximum green phase Under the constraint, the current green light phase time is extended, that is, the judgment ,in is the start time of the current green light phase, t is the current time, and ∆T is the update interval of the enhanced adaptive signal optimization strategy. If yes, the current green light phase time is extended to , go to step 2.1.1; otherwise, let the current green light phase end time , go to step 2.1.4.

[0013] Step 2.1.4: Update the signal timing plan.

[0014] Step 2.2: According to the signal timing plan in step 2.1, the red light phase signal optimization strategy is further implemented in the conflict identification area. The specific steps are as follows.

[0015] Step 2.2.1: Update the intersection vehicle information and signal timing information, and calculate the headway time between any two consecutive vehicles in the current green light phase lane.

[0016] Step 2.2.2: Determine the type of vehicle in the current red light phase lane, and based on this, calculate whether there is a suitable feasible time interval to allow the vehicle to pass through the conflict identification area. The specific calculation method is as follows.

[0017] The intersection conflict identification area is gridded, and the vehicle conflict relationship in the conflict identification area is identified, and the vehicle time relationship is calculated: In the formula, and They represent the time when vehicle i leaves and arrives at the conflict point, represents the time when vehicle l arrives at the conflict point, It indicates the time when vehicle l-1 leaves the conflict zone. Among the three vehicles mentioned above, vehicle i comes from the lane served by the red light phase, and vehicles l and l-1 come from the lane served by the current green light phase, and they are two vehicles that arrive consecutively (vehicle l-1 is the front vehicle).

[0018] Step 2.2.3: Based on the vehicle position relationship and the time relationship calculation formula described in step 2.2.2, find the first feasible time interval for allowing the vehicle to pass through the conflict identification area.

[0019] Step 2.2.4: Switch the red light phase of the lane where the vehicle is located to the green light phase. If the vehicle is an intelligent connected vehicle, notify it to adjust its trajectory accordingly.

[0020] Step 3: Establish the lateral and longitudinal trajectory optimization control model of intelligent networked vehicles and manually driven vehicles. The specific method is as follows.

[0021] Step 3.1: Based on whether the vehicle's driving trajectory is affected by the vehicle in front, four types of vehicle longitudinal trajectory control strategies are established, including the four-stage trajectory control strategy for the leading vehicle of intelligent connected vehicles, the two-stage trajectory control strategy for the leading vehicle degradation of manually driven vehicles, the following vehicle control strategy for intelligent connected vehicles, and the following vehicle control strategy for manually driven vehicles. The calculation formulas are shown below.

[0022] First, the four-stage trajectory control strategy of the leading vehicle of the intelligent connected vehicle is an optimal control problem of multi-stage free terminal time. The model is as follows: .

[0023] Subject to the following constraints: Where a, v, and x are the acceleration, velocity, and position of vehicle i, respectively. and are the time when vehicle i leaves and enters the intersection, and are the positions of vehicle i leaving and entering the intersection, respectively. and are the expected speed of vehicle i and the speed at which it enters the intersection, respectively. The subscripts 1, 2, 3, and 4 represent the four stages of the vehicle trajectory, respectively. The vehicle has a constant acceleration in stages 1 and 3, which belong to the uniform acceleration / deceleration trajectory segment, and has a constant speed in stages 2 and 4, which belong to the uniform speed trajectory segment.

[0024] Second, the two-stage trajectory control strategy of the manually driven vehicle with the lead vehicle degradation aims to simulate the driving mode of human drivers. It can be divided into two stages. Stage 1 has a constant acceleration / deceleration, which is regarded as a uniform acceleration / deceleration stage, and stage 2 has a constant speed, which is regarded as a uniform speed stage. The model is as follows: 。

[0025] Third, the intelligent connected vehicle following control strategy adopts the intelligent driver following model, and the formula is as follows: In the formula, is the safe headway of the vehicle, and It is the position difference and speed difference between the front and rear vehicles.

[0026] Fourth, the manually driven vehicle following control strategy adopts an intelligent driver following model, and the formula is the same as the third formula.

[0027] Step 3.2: When the vehicle is traveling in the free driving area, it can perform lane changing operations. Therefore, a lane changing behavior control model is constructed for the vehicle, and the model formula is as follows: , ; In the formula, represents the overall benefit of vehicle lane change, only when , that is, the overall benefit is greater than the target threshold The lane change behavior is only executed when It is the courtesy factor of vehicle lane change, which represents the degree of impact on other vehicles on the road when the subject vehicle changes lanes. , and They represent the acceleration of the main vehicle, the vehicle immediately behind in the target lane, and the vehicle immediately behind in the original lane after the lane change. , and They represent the acceleration of the main vehicle before lane change, the vehicle immediately behind in the target lane, and the vehicle immediately behind in the original lane, respectively.

[0028] Step 4: Iterate and update the above two signal optimization strategies and vehicle lateral and longitudinal control strategies. The enhanced adaptive signal optimization strategy is executed every ∆T. After that, for the vehicles traveling in the lane facing the current red light phase, the red light phase signal optimization strategy is executed every ∆t to check whether there is a feasible time interval between any two consecutive vehicles in the green light phase in the conflict recognition area that allows the vehicle to pass through the intersection without conflict, and different longitudinal trajectory control strategies are implemented for different types of vehicles. The relationship between the two time intervals is calculated as follows: Where t and n are the updating time steps of the red light phase signal optimization strategy and the enhanced adaptive signal optimization strategy, respectively.

[0029] The specific method is as follows.

[0030] Step 4.1: When the vehicle is traveling in the free driving area, the corresponding trajectory control strategy is executed for different vehicle types according to step 3, including lateral lane changing behavior and longitudinal following behavior.

[0031] Step 4.2: When the vehicle enters the lane-changing prohibited area from the free driving area, the lateral lane-changing behavior is prohibited, and only the longitudinal trajectory control strategy for the corresponding vehicle type described in step 3 is executed.

[0032] Step 4.3: When the intersection control center detects a vehicle in the no lane change zone, the enhanced adaptive signal optimization strategy is activated according to step 2.1 to dynamically adjust the green light phase time of the intersection according to real-time traffic information and signal timing.

[0033] Step 4.4: When the vehicle approaches the conflict recognition area, the red light phase signal optimization strategy is activated to determine whether the current vehicle can safely pass through the intersection during the red light phase by calculating the potential conflict relationship of the vehicles in the conflict recognition area. If yes, the red light phase facing the vehicle is switched to the green light phase to allow it to pass through the intersection; otherwise, the red light phase continues to be executed.

[0034] The beneficial effects of the present invention are as follows: the present invention discloses a method for collaborative optimization of vehicle trajectories and signals at intersections under a hybrid intelligent network environment. Aiming at the scenario of intersection control under a hybrid intelligent network environment, the collaborative optimization of vehicle trajectories and signals at signalized intersections under a hybrid intelligent network environment is achieved. By analyzing the conflict relationship between vehicles at intersections, a red light phase optimization strategy is constructed to improve the utilization rate of intersection signals, and an enhanced adaptive signal optimization strategy is used to ensure the operation of intersection signal control. Secondly, corresponding vehicle transverse and longitudinal trajectory optimization strategies are constructed for different types of vehicles. On the basis of ensuring the maximization of vehicle traffic efficiency, the fuel economy and driving comfort of the vehicle are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 The overall flow chart of intersection control described in the present invention;

[0057] Figure 2 This is a schematic diagram of the road area division described in the present invention.

Claims

1. A method for coordinated optimization of intersection vehicle trajectories and signals in a hybrid intelligent network environment, characterized in that: This is accomplished by following these steps: Step 1: Define the intersection optimization area. Define the intersection as a two-way four-lane cross intersection with a through lane and a left-turn lane, where lane 1 is the through lane and lane 2 is the left-turn lane. Define the inner area of ​​the intersection as the conflict identification area, the entrance area as the signal control area, and further divide the signal control area into a free driving area and a no lane change area; Step 2: Implement different signal optimization strategies for different areas. Implement enhanced adaptive signal optimization strategy for signal control area, and implement red light phase signal optimization strategy for conflict identification area. The specific method is: Step 2.1: Implement enhanced adaptive signal optimization strategy for the signal control area. The specific steps are as follows: Step 2.1.1: Update intersection signal timing and vehicle trajectory information, and determine whether the current green light phase exceeds the minimum green light time , where w represents the current lane number. If no, go to step 2.1.4; Otherwise, go to step 2.1.2; Step 2.1.2: Calculate the arrival time of vehicles in the lane served by the current green light phase, and determine whether all vehicles can pass through the intersection during the current green light phase. If yes, go to step 2.1.4; otherwise, go to step 2.1.3; Step 2.1.3: Calculate whether the maximum green phase can be reached without violating the maximum green phase Under the constraint, the current green light phase time is extended, that is, the judgment ,in is the start time of the current green light phase, t is the current time, and ∆T is the update interval of the enhanced adaptive signal optimization strategy. If yes, the current green light phase time is extended to , go to step 2.1.1; otherwise, let the current green light phase end time , go to step 2.1.4; Step 2.1.4: Update the signal timing plan; Step 2.2: According to the signal timing plan in step 2.1, the red light phase signal optimization strategy is further implemented in the conflict identification area. The specific steps are as follows: Step 2.2.1: Update the intersection vehicle information and signal timing information, and calculate the headway time between any two consecutive vehicles in the current green light phase lane; Step 2.2.2: Determine the type of vehicle in the current red light phase lane, and calculate whether there is a suitable feasible time interval to allow the vehicle to pass through the conflict identification area. The specific calculation method is as follows: The intersection conflict identification area is gridded, and the vehicle conflict relationship in the conflict identification area is identified, and the vehicle time relationship is calculated: In the formula, and They represent the time when vehicle i leaves and arrives at the conflict point, represents the time when vehicle l arrives at the conflict point, It indicates the time when vehicle l-1 leaves the conflict zone. Among the three vehicles mentioned above, vehicle i comes from the lane served by the red light phase, and vehicles l and l-1 come from the lane served by the current green light phase, and they are two vehicles that arrive consecutively (vehicle l-1 is the front vehicle); Step 2.2.3: Find the first feasible time interval for allowing the vehicle to pass through the conflict identification area based on the vehicle position relationship and the time relationship calculation formula described in step 2.2.2; Step 2.2.4: Switch the red light phase of the lane where the vehicle is located to the green light phase. If the vehicle is an intelligent connected vehicle, notify it to adjust its trajectory accordingly; Step 3: Establish the lateral and longitudinal trajectory optimization control model of intelligent connected vehicles and manually driven vehicles. The specific method is as follows: Step 3.1: According to whether the vehicle's driving trajectory is affected by the vehicle in front, four types of vehicle longitudinal trajectory control strategies are established, including the four-stage trajectory control strategy of the leading vehicle of the intelligent networked vehicle, the two-stage trajectory control strategy of the leading vehicle degradation of the manually driven vehicle, the following vehicle control strategy of the intelligent networked vehicle, and the following vehicle control strategy of the manually driven vehicle. The calculation formulas are as follows: First, the four-stage trajectory control strategy of the leading vehicle of the intelligent connected vehicle is an optimal control problem of multi-stage free terminal time. The model is as follows: Subject to the following constraints: Where a, v, and x are the acceleration, velocity, and position of vehicle i, respectively. and are the time when vehicle i leaves and enters the intersection, and are the positions of vehicle i leaving and entering the intersection, respectively. and are the expected speed of vehicle i and the speed entering the intersection, respectively. Subscripts 1, 2, 3, and 4 represent the four stages of the vehicle trajectory, respectively. The vehicle has a constant acceleration in stages 1 and 3, belonging to the uniform acceleration / deceleration trajectory segment, while it has a constant speed in stages 2 and 4, belonging to the uniform speed trajectory segment; Second, the two-stage trajectory control strategy of the manually driven vehicle with the lead vehicle degradation aims to simulate the driving mode of human drivers. It can be divided into two stages. Stage 1 has a constant acceleration / deceleration, which is regarded as a uniform acceleration / deceleration stage, and stage 2 has a constant speed, which is regarded as a uniform speed stage. The model is as follows: Third, the intelligent connected vehicle following control strategy adopts the intelligent driver following model, and the formula is as follows: In the formula, is the safe headway of the vehicle, and is the position difference and speed difference of the front and rear vehicles; Fourth, the control strategy for manually driven vehicles following cars adopts an intelligent driver following model, and the formula is the same as the formula described in the third; Step 3.2: When the vehicle is traveling in the free driving area, it can perform lane changing operations. Therefore, a lane changing behavior control model is constructed for the vehicle, and the model formula is as follows: ， ; In the formula, represents the overall benefit of the vehicle changing lanes. , that is, the overall benefit is greater than the target threshold The lane change behavior is only executed when It is the courtesy factor of vehicle lane change, which represents the degree of impact on other vehicles on the road when the subject vehicle changes lanes. , and They represent the acceleration of the main vehicle, the vehicle immediately behind in the target lane, and the vehicle immediately behind in the original lane after the lane change. , and They represent the acceleration of the main vehicle before lane change, the vehicle immediately behind the target lane, and the vehicle immediately behind the original lane respectively; Step 4: Iterate and update the above two signal optimization strategies and vehicle lateral and longitudinal control strategies. The enhanced adaptive signal optimization strategy is executed every ∆T. After that, for the vehicles traveling in the lane facing the current red light phase, the red light phase signal optimization strategy is executed every ∆t to check whether there is a feasible time interval between any two consecutive vehicles in the green light phase in the conflict recognition area that allows the vehicle to pass through the intersection without conflict, and different longitudinal trajectory control strategies are implemented for different types of vehicles. The relationship between the two time intervals is calculated as follows: Where t and n are the time steps for updating the red light phase signal optimization strategy and the enhanced adaptive signal optimization strategy, respectively; The specific method is as follows: Step 4.1: When the vehicle is traveling in the free driving area, the corresponding trajectory control strategy is executed for different vehicle types according to step 3, including lateral lane changing behavior and longitudinal following behavior; Step 4.2: When the vehicle enters the lane-changing prohibited area from the free driving area, the lateral lane-changing behavior is prohibited, and only the longitudinal trajectory control strategy for the corresponding vehicle type described in step 3 is executed; Step 4.3: When the intersection control center detects that there is a vehicle in the no-lane-change zone, the enhanced adaptive signal optimization strategy is activated according to step 2.1 to dynamically adjust the green light phase time of the intersection according to real-time traffic information and signal timing; Step 4.4: When the vehicle approaches the conflict recognition area, the red light phase signal optimization strategy is activated to determine whether the current vehicle can safely pass through the intersection during the red light phase by calculating the potential conflict relationship of the vehicles in the conflict recognition area. If yes, the red light phase facing the vehicle is switched to the green light phase to allow it to pass through the intersection; otherwise, the red light phase continues to be executed.

Images