A vehicle control and signal optimization method for mixed-flow intersections under vehicle-road collaboration

Abstract

The present invention discloses a mixed-flow intersection vehicle control and signal optimization method under vehicle-road collaboration, which belongs to the field of intelligent transportation. The specific steps of the control optimization method are as follows: (1) constructing a networked signal-controlled intersection scene; (2) constructing a corresponding control model to classify and control vehicles; (3) optimizing the timing of traffic lights through a signal fuzzy control algorithm; (4) collecting information from each group and performing comparative analysis. The present invention can effectively reduce the energy consumption of vehicles at the intersection and improve the traffic efficiency of the intersection, and has strong practical significance and application value.

Description

Technical Field

[0001] The present invention relates to the field of intelligent transportation, and in particular to a mixed-flow intersection vehicle control and signal optimization method under vehicle-road collaboration. Background Art

[0002] In recent years, intelligent transportation and traffic big data have developed rapidly. In the long term, the traffic flow on the road will include both traditional manually driven vehicles and intelligent networked vehicles. That is, the transition from the current homogeneous flow consisting of one type of vehicle to a mixed traffic flow consisting of at least two types of vehicles will be a major trend. At the same time, with the rapid growth of the number of cars, the problem of urban road traffic congestion has become increasingly prominent. As the intersection of roads, road intersections are often the bottleneck of traffic flow in urban road networks. In addition, the control of traffic lights at intersections causes vehicles in the intersection area to start and stop frequently and accelerate and decelerate, resulting in low traffic efficiency and vehicle fuel economy. Vehicle-road collaboration is currently an important technical means to improve traffic safety, improve traffic efficiency, and reduce traffic emissions. In the networked signal intersection environment, the Internet of Vehicles and autonomous driving technology are integrated to obtain the motion status information of surrounding vehicles and the phase and countdown information of signal lights hundreds of meters away through vehicle-to-vehicle communication and vehicle-to-road communication, and then accurately complete the trajectory optimization control to achieve ecological entry and exit of signal intersections;

[0003] Previous studies have mostly focused on controlling vehicle trajectories to achieve minimum fuel consumption or maximum traffic efficiency at signalized intersections, or both, or on achieving the above functions by controlling the signal phase and timing at signalized intersections. They have not considered jointly controlling the vehicle trajectory and the timing and phase of signalized intersections to simultaneously achieve the maximum traffic efficiency or the lowest energy consumption at the intersection; therefore, the present invention proposes a mixed-flow intersection vehicle control and signal optimization method under vehicle-road collaboration. Summary of the invention

[0004] The purpose of the present invention is to solve the defects in the prior art and to propose a mixed-flow intersection vehicle control and signal optimization method under vehicle-road collaboration.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A mixed-flow intersection vehicle control and signal optimization method under vehicle-road collaboration, the control optimization method has the following specific steps:

[0007] (1) Constructing a networked signal-controlled intersection scenario;

[0008] (2) Construct corresponding control models to classify and control vehicles;

[0009] (3) Optimize the timing of traffic lights through signal fuzzy control algorithm;

[0010] (4) Collect information from each group and conduct comparative analysis.

[0011] As a further solution of the present invention, the networked signal-controlled intersection scenario described in step (1) is specifically a two-way four-lane cross-shaped intersection scenario, and the vehicles traveling in the scenario include CAVs and CHVs, both of which have networked communication capabilities, wherein the CAV is a fully autonomous driving vehicle with autonomous intelligent perception, decision-making and control capabilities, as well as the ability to communicate with other vehicles or traffic lights, and the CHV can achieve V2V communication with nearby vehicles. At the same time, the intersection traffic lights in the intersection scenario also have networked communication capabilities and can achieve V2I communication with the vehicles in the scenario.

[0012] As a further solution of the present invention, the specific steps of the classification control in step (2) are as follows:

[0013] Step 1: Detect the distance between the controlled vehicle and the vehicle in front of it and record it as d rel , and detect the relative distance to the upcoming signalized intersection and record it as d TL , and construct trajectory control models for CAV vehicles under different conditions;

[0014] Step 2: If the CAV vehicle detects the presence of a vehicle ahead, the CAV vehicle adjusts the vehicle trajectory through the signal-based vehicle trajectory control model to obtain a specific vehicle control strategy, and controls the CAV vehicle and performs SUMO simulation using the obtained control strategy. At the same time, the control strategy is reused to control the next CAV vehicle in the same situation to update the vehicle status of the entire intersection in real time;

[0015] Step 3: If the CAV vehicle detects the presence of a vehicle in front, and at this time the CAV vehicle should follow the vehicle in front, maintain a safe distance from the vehicle in front and achieve a high road resource utilization rate, then the trajectory control model based on the following behavior is used to control the vehicle;

[0016] Step 4: Construct the core functional modules based on the vehicle trajectory control model optimized for real-time traffic signal information and the vehicle trajectory control model for car-following behavior. At the same time, the two sets of models will calculate the optimal control command u at each time step as the input of the CAV vehicle, and obtain the instantaneous output of the CAV in the context of vehicle-road collaboration. The output results are input into the switch logic algorithm for model selection;

[0017] Step 5: The detected d rel With the limit value d limit For comparison, if d rel No more than d limit, then the vehicle trajectory control model based on car-following behavior is selected to control the CAV vehicle to follow the vehicle in front. If d rel Not less than d limit , the signal-based vehicle trajectory control model customizes the variable time interval strategy, and the vehicle trajectory control model based on following behavior is selected to maintain the vehicle spacing for safe, comfortable and efficient use of road resources.

[0018] As a further solution of the present invention, the signal-based vehicle trajectory control model in step 2 is specifically expressed as follows:

[0019]

[0020]

[0021]

[0022] 0≤d TL,i ≤d TL,max +ε 1,i (4)

[0023] v min,i ≤v host,i ≤v max,i (5)

[0024] a min,i ≤a host,i ≤a max,i (6)

[0025] j min,i -ε 2,i ≤a host,i+1 -a host,i ≤j max,i +ε 2,i (7)

[0026] 0≤T m,i ≤T m,max (8)

[0027] In formulas (1)-(8), E(v host,i ,a host,i ,T m,i ) represents the energy consumption of the vehicle, v target represents the target vehicle speed optimized based on signal phase information, as well as Represent weighting factors, ε1 and ε2 represent slack variables, m eq represents the total mass of the vehicle, i g represents a single transmission ratio, η e Represents the total mechanical efficiency of the transmission system, r w Represents the wheel radius, cr represents the rolling resistance coefficient, g represents the acceleration of gravity, θ represents the road slope, ρa represents the air density, and A f Represents the front area of ​​the vehicle, C D represents the air resistance coefficient;

[0028] The specific expression formula of the trajectory control model based on the car-following behavior described in step 3 is as follows:

[0029]

[0030]

[0031]

[0032] 0≤d TL,i ≤d TL,max +ε 1,i (12)

[0033] v min,i ≤v host,i ≤v max,i (13)

[0034] a min,i ≤a host,i ≤a max,i (14)

[0035] j min,i -ε 2,i ≤a host,i+1 -a host,i ≤j max,i -ε 2,i (15)

[0036] 0≤T m,i ≤T m,max (16)

[0037] 0≤ω m,i ≤ω m,max (17)

[0038] v rel,i =v preceding,i -v host,i (18)

[0039]

[0040]

[0041] D safe,i ≤d rel,i ≤D safe,i +ε 3,i (twenty one)

[0042] VTH=τ1·+τ2·v host -τ3·v rel (twenty two)

[0043] v rel =v preceding -v host (twenty three)

[0044]

[0045] In formulas (9)-(24), ε3 represents the slack variable, D safe Represents the dynamic safety distance between the CAV vehicle and the preceding vehicle, d min represents the minimum distance between two vehicles when they are stationary, τ1, τ2, and τ3 represent constant values ​​greater than zero, and v rel Represents the relative speed between the CAV vehicle and the vehicle ahead of it.

[0046] As a further solution of the present invention, the specific steps of optimizing the timing of the signal lights in step (3) are as follows:

[0047] Step I: Give the shortest green light time g to the phase currently allowing traffic min ;

[0048] Step II: Before the green light time of the current green light phase ends, the vehicle queue information and vehicle arrival information of the current green light phase and the next green light phase are input into the observation module, and the traffic intensity of the two phases is obtained through fuzzy reasoning;

[0049] Step III: Input the current green light phase and the next green light phase strength into the decision module, and obtain the green light extension time of the current green light phase through fuzzy reasoning. If the extension time is greater than 8s and the current phase lasts for the green light time g, duration The extended time is less than the maximum green light time g max , then the current green light phase is extended by this time, and the operation of step II is performed;

[0050] Step IV: If the extension time is greater than 8s, and g duration With the extension time greater than or equal to g max , the current phase green light is extended and then switches to the next green light phase, and then the yellow light g yellow After the delay, switch to the next green light phase and perform step Ⅰ. If the extended time is less than or equal to 8s, the yellow light time g yellow After the duration has passed, switch to the next green light phase and proceed to step I.

[0051] As a further solution of the present invention, the specific steps of the comparative analysis in step (4) are as follows:

[0052] Step 1: Use Python to perform secondary development on the SUMO simulation platform, build a simulation environment, then collect the instantaneous speed and acceleration of the CAV vehicle, and build a corresponding energy consumption model through the instantaneous gasoline consumption function;

[0053] Step ②: Collect and analyze the simulation data of each group. When the penetration rate is higher, the energy consumption saved by the joint control of vehicle trajectory and signal is higher than that of vehicle trajectory control. At the same time, as the penetration rate of CAV increases, the average waiting time of vehicles under the joint control of vehicle trajectory and signal is significantly shorter than that under fixed timing and vehicle control only.

[0054] Step ③: Gradually increase the number of groups of vehicles in the simulation environment. As the traffic volume increases, the traffic flow gradually transitions from free flow to a stable state, and the energy consumption of vehicles shows an upward trend. Compared with the vehicle trajectory control scheme and the fixed timing scheme, the vehicle trajectory and signal joint control scheme has a significant effect on reducing the average energy consumption of vehicles. At the same time, at any level of traffic volume, both vehicle control and signal and vehicle joint control have an optimization effect on the average waiting time. Among them, the signal and vehicle joint control method has the best optimization effect on the average waiting time.

[0055] As a further solution of the present invention, the specific formula for constructing the energy consumption model in step ① is as follows:

[0056] F(v,a,t)=[α+β1P(t)+β2m eq a(t) 2 v(t)] a＞0 (25)

[0057]

[0058] Wherein, α represents the constant idle gasoline consumption rate, β1 represents an efficiency parameter related to engine efficiency, β2 represents the gasoline consumption parameter positively related to acceleration, v represents the vehicle speed, a represents the vehicle acceleration, P represents the total power of the vehicle operation, d1 represents the rolling resistance, d2 represents the engine drag coefficient, and d3 represents the air drag coefficient.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] The vehicle control and signal optimization method for mixed intersections under vehicle-road collaboration constructs core functional modules based on the vehicle trajectory control model for real-time traffic signal information optimization and the vehicle trajectory control model for following behavior, obtains the instantaneous output of CAV in the context of vehicle-road collaboration, and inputs the output result into the switch logic algorithm for model selection. At the same time, before the end of the green light time of the current green light phase, the vehicle queuing information and traffic arrival information of the current green light phase and the next green light phase are input into the observation module, and the traffic intensity of the two phases is obtained through fuzzy reasoning. The intensity of the current green light phase and the next green light phase is input into the decision module, and the green light extension time of the current green light phase is obtained through fuzzy reasoning. The intersection signal light is optimized according to the extension time and the continuous green light time of the current phase. Through the double-layer model optimization system of vehicle trajectory and signal joint control, the vehicle energy consumption at the intersection can be effectively reduced and the traffic efficiency of the intersection can be improved, which has strong practical significance and application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0062] Figure 1 A flowchart of a mixed-flow intersection vehicle control and signal optimization method under vehicle-road collaboration proposed by the present invention;

[0063] Figure 2 A schematic diagram of data flow in a double-layer model predictive control system under the switch logic of a mixed-flow intersection vehicle control and signal optimization method under vehicle-road collaboration proposed by the present invention;

[0064] Figure 3 A fuzzy control system block diagram of a mixed-flow intersection vehicle control and signal optimization method under vehicle-road collaboration proposed by the present invention;

[0065] Figure 4 A comparison chart of the average vehicle energy consumption of various schemes under different CAV penetration rates of a mixed-flow intersection vehicle control and signal optimization method under vehicle-road collaboration proposed by the present invention;

[0066] Figure 5 A comparison chart of the average waiting time of vehicles in a double-layer control model under different CAV penetration rates for a mixed-flow intersection vehicle control and signal optimization method under vehicle-road collaboration proposed by the present invention;

[0067] Figure 6 A comparison chart of the average waiting time of vehicles in various schemes under different CAV penetration rates for a mixed-flow intersection vehicle control and signal optimization method under vehicle-road collaboration proposed by the present invention;

[0068] Figure 7 A comparison chart of average vehicle energy consumption of various schemes under different traffic flows of a mixed-flow intersection vehicle control and signal optimization method under vehicle-road collaboration proposed by the present invention;

[0069] Figure 8 This is a comparison chart of the average waiting time of vehicles under different traffic flows for various schemes of the mixed-flow intersection vehicle control and signal optimization method under vehicle-road collaboration proposed in the present invention. DETAILED DESCRIPTION

[0070] Reference Figure 1-Figure 8 , a mixed-flow intersection vehicle control and signal optimization method under vehicle-road collaboration, the specific steps of the control optimization method are as follows:

[0071] Build a networked signal-controlled intersection scenario.

[0072] It should be further explained that the networked signal-controlled intersection scenario is specifically a two-way four-lane cross-shaped intersection scenario, and the vehicles traveling in the scenario include CAV and CHV. Both types of vehicles have networked communication capabilities. CAV is a fully autonomous driving vehicle with autonomous intelligent perception, decision-making and control capabilities, as well as the ability to communicate with other vehicles or traffic lights. CHV can achieve V2V communication with nearby vehicles. At the same time, the traffic lights at the intersection in this intersection scenario also have networked communication capabilities and can achieve V2I communication with the vehicles in the scenario.

[0073] Construct corresponding control models to perform classified control on vehicles.

[0074] Specifically, refer to Figure 2 It can be seen that the distance between the controlled vehicle and the vehicle in front of it is detected and recorded as d rel , and detect the relative distance to the upcoming signalized intersection and record it as d TLAt the same time, the trajectory control model of CAV vehicles under different situations is constructed. If the CAV vehicle detects the presence of the vehicle in front, the CAV vehicle adjusts the vehicle trajectory through the signal-based vehicle trajectory control model to obtain the specific vehicle control strategy, and controls the CAV vehicle and SUMO simulation through the obtained control strategy. At the same time, the control strategy is reused for the next CAV vehicle in the same situation to control the real-time update of the vehicle status of the entire intersection. If the CAV vehicle detects the presence of the vehicle in front, the CAV vehicle should follow the vehicle in front at this time, and maintain a safe distance from the vehicle in front and achieve a high road resource utilization rate. At this time, the trajectory control model based on following behavior is used for vehicle control. The core functional modules are constructed based on the vehicle trajectory control model optimized for real-time traffic signal information and the vehicle trajectory control model of following behavior. At the same time, the two sets of models will calculate the optimal control command u in each time step as the input of the CAV vehicle, and obtain the instantaneous output of the CAV in the context of vehicle-road collaboration, and input the output result into the switch logic algorithm for model selection, and the detected d rel With the limit value d limit For comparison, if d rel No more than d limit , then the vehicle trajectory control model based on car-following behavior is selected to control the CAV vehicle to follow the vehicle in front. If d rel Not less than d limit , the signal-based vehicle trajectory control model customizes the variable time interval strategy, and the vehicle trajectory control model based on following behavior is selected to maintain the vehicle spacing for safe, comfortable and efficient use of road resources.

[0075] The specific expression formula of the signal-based vehicle trajectory control model is as follows:

[0076]

[0077]

[0078]

[0079] 0≤d TL,i ≤d TL,max +ε 1,i (4)

[0080] v min,i ≤v host,i ≤v max,i (5)

[0081] a min,i ≤a host,i ≤a max,i (6)

[0082] jmin,i -ε 2,i ≤a host,i+1 -a host,i ≤j max,i +ε 2,i (7)

[0083] 0≤T m,i ≤T m,max (8)

[0084] In formulas (1)-(8), E(v host,i ,a host,i ,T m,i ) represents the energy consumption of the vehicle, v target represents the target vehicle speed optimized based on signal phase information, as well as Represent weighting factors, ε1 and ε2 represent slack variables, m eq represents the total mass of the vehicle, i g represents a single transmission ratio, η e Represents the total mechanical efficiency of the transmission system, r w Represents the wheel radius, c r represents the rolling resistance coefficient, g represents the acceleration of gravity, θ represents the road slope, ρa represents the air density, and A f Represents the front area of ​​the vehicle, C D represents the air resistance coefficient;

[0085] The specific expression formula of the trajectory control model based on car-following behavior is as follows:

[0086]

[0087]

[0088]

[0089] 0≤d TL,i ≤d TL,max +ε 1,i (12)

[0090] v min,i ≤v host,i ≤v max,i (13)

[0091] a min,i ≤a host,i ≤a max,i (14)

[0092] j min,i -ε 2,i ≤a host,i+1 -a host,i ≤jmax,i -ε 2,i (15)

[0093] 0≤T m,i ≤T m,max (16)

[0094] 0≤ω m,i ≤ω m,max (17)

[0095] v rel,i =v preceding,i -v host,i (18)

[0096]

[0097]

[0098] D safe,i ≤d rel,i ≤D safe,i +ε 3,i (twenty one)

[0099] VTH=τ1·+τ2·v host -τ3·v rel (twenty two)

[0100] v rel =v preceding -v host (twenty three)

[0101]

[0102] In formulas (9)-(24), ε3 represents the slack variable, D safe Represents the dynamic safety distance between the CAV vehicle and the preceding vehicle, d min represents the minimum distance between two vehicles when they are stationary, τ1, τ2, and τ3 represent constant values ​​greater than zero, and v rel Represents the relative speed between the CAV vehicle and the vehicle ahead of it.

[0103] The signal timing is optimized through signal fuzzy control algorithm.

[0104] Specifically, refer to Figure 3 It can be seen that the shortest green light time g is given to the phase currently allowed to pass minBefore the green light time of the current green light phase ends, the vehicle queue information and traffic arrival information of the current green light phase and the next green light phase are input into the observation module, and the traffic intensity of the two phases is obtained through fuzzy reasoning. The intensity of the current green light phase and the next green light phase is input into the decision module, and the green light extension time of the current green light phase is obtained through fuzzy reasoning. If the extension time is greater than 8s, and the current phase lasts for the green light time g duration The extended time is less than the maximum green light time g max , the current green light phase is extended by this time, and the operation of step II is performed. Before the end of the green light time of the current green light phase, the vehicle queue information and vehicle arrival information of the current green light phase and the next green light phase are input into the observation module, and the traffic intensity of the two phases is obtained through fuzzy reasoning. If the extension time is greater than 8s, and g duration With the extension time greater than or equal to g max , the current phase green light is extended and then switched to the next green light phase, and then the yellow light time g yellow After the duration, switch to the next green light phase and give the current phase that allows passage the shortest green light time g min If the extension time is less than or equal to 8s, the yellow light time g yellow After the duration, switch to the next green light phase and give the current phase that allows passage the shortest green light time g min .

[0105] It should be further explained that the observation module calculates the traffic intensity of the current green light phase and the next green light phase based on the detected real-time traffic flow. The module has two input variables, namely the number of queued vehicles between the upstream and downstream detectors of the evaluated phase and the vehicle arrival rate of the phase, and the output variable is the traffic intensity of the evaluated phase. The vehicle arrival rate is the number of vehicles arriving at the intersection entrance per second, which can be calculated by the number of vehicles arriving in a cycle measured by the lane checker; the decision module determines the green light extension time of the current green light phase based on the traffic intensity of the current phase and the traffic intensity of the next phase. The module has two input variables, namely the traffic intensity of the current phase and the traffic intensity of the next phase, and the output variable is the extension time of the current green light phase.

[0106] Collect information from each group and conduct comparative analysis.

[0107] Specifically, refer to Figure 4-Figure 8It can be seen that the SUMO simulation platform is secondary developed through Python to build a simulation environment. Then the instantaneous speed and acceleration of CAV vehicles are collected, and the corresponding energy consumption model is constructed through the instantaneous gasoline consumption function. The simulation data of each group are collected and analyzed. When the penetration rate is higher, the energy consumption saved by the joint control of vehicle trajectory and signal is improved compared with the vehicle trajectory control. At the same time, with the increase of CAV penetration rate, the average waiting time of vehicles under the joint control of vehicle trajectory and signal is significantly shorter than the average waiting time when fixed timing and only controlling vehicles are shorter. Then, each group of vehicles in the simulation environment is gradually increased. With the increase of traffic volume, the traffic flow gradually transitions from free flow to a stable state, and the energy consumption of vehicles shows an upward trend. Compared with the vehicle trajectory control scheme and the fixed timing scheme, the joint control scheme of vehicle trajectory and signal has a significant effect on the reduction of the average energy consumption of vehicles. At the same time, under any traffic volume, vehicle control and signal and vehicle joint control have optimization effects on the average waiting time. Among them, the signal and vehicle joint control method has the best optimization effect on the average waiting time.

[0108] In addition, it needs to be further explained that the specific construction formula of the energy consumption model is as follows:

[0109] F(v,a,t)=[α+β1P(t)+β2m eq a(t) 2 v(t)] a＞0 (25)

[0110]

[0111] Wherein, α represents the constant idle gasoline consumption rate, β1 represents an efficiency parameter related to engine efficiency, β2 represents the gasoline consumption parameter positively related to acceleration, v represents the vehicle speed, a represents the vehicle acceleration, P represents the total power of the vehicle operation, d1 represents the rolling resistance, d2 represents the engine drag coefficient, and d3 represents the air drag coefficient.

Claims

1. A mixed-flow intersection vehicle control and signal optimization method under vehicle-road collaboration, characterized in that: The specific steps of the vehicle control and signal optimization method are as follows: (1) Constructing a networked signal-controlled intersection scenario; (2) Construct corresponding control models to classify and control vehicles; (3) Optimize the timing of traffic lights through signal fuzzy control algorithm; (4) Collect information from each group and conduct comparative analysis; The specific steps of classification control described in step (2) are as follows: Step 1: Detect the distance between the controlled vehicle and the vehicle in front of it and record it as , while detecting the relative distance of the upcoming signalized intersection and recording it as , and construct trajectory control models for CAV vehicles under different conditions; Step 2: If the CAV vehicle detects the presence of a vehicle ahead, the CAV vehicle adjusts the vehicle trajectory through the signal-based vehicle trajectory control model to obtain a specific vehicle control strategy, and controls the CAV vehicle and performs SUMO simulation using the obtained control strategy. At the same time, the control strategy is reused to control the next CAV vehicle in the same situation to update the vehicle status of the entire intersection in real time; Step 3: If the CAV vehicle detects the presence of a vehicle in front, and at this time the CAV vehicle should follow the vehicle in front, maintain a safe distance from the vehicle in front and achieve a high road resource utilization rate, then the trajectory control model based on the following behavior is used to control the vehicle; Step 4: Construct the core functional modules based on the vehicle trajectory control model optimized for real-time traffic signal information and the vehicle trajectory control model for car-following behavior. At the same time, the two sets of models will calculate the optimal control command u at each time step as the input of the CAV vehicle, and obtain the instantaneous output of the CAV in the context of vehicle-road collaboration. The output results are input into the switch logic algorithm for model selection; Step 5: With limit value For comparison, if No more than , then the vehicle trajectory control model based on car-following behavior is selected to control the CAV vehicle to follow the vehicle in front. Not less than , the signal-based vehicle trajectory control model customizes the variable time interval strategy, and the vehicle trajectory control model based on following behavior is selected to maintain the vehicle spacing for safe, comfortable and efficient use of road resources.

2. The method for vehicle control and signal optimization at a mixed-flow intersection under vehicle-road collaboration according to claim 1 is characterized in that: The networked signal-controlled intersection scenario described in step (1) is specifically a two-way four-lane cross-shaped intersection scenario, and the vehicles traveling in the scenario include CAVs and CHVs, both of which have networked communication capabilities, wherein the CAV is a fully autonomous driving vehicle with autonomous intelligent perception, decision-making and control capabilities, as well as the ability to communicate with other vehicles or traffic lights, and the CHV can achieve V2V communication with nearby vehicles. At the same time, the traffic lights at the intersection in the intersection scenario also have networked communication capabilities and can achieve V2I communication with the vehicles in the scenario.

3. The method for vehicle control and signal optimization at a mixed-flow intersection under vehicle-road collaboration according to claim 1, characterized in that: The specific expression formula of the signal-based vehicle trajectory control model in step 2 is as follows: （4） （5） （6） （7） （8） In formulas (1)-(8), E(v host,i ,a host,i ,T m,i ) represents the energy consumption of the vehicle, represents the target vehicle speed optimized based on signal phase information, , as well as Represent the weighting factors, as well as represent slack variables, represents the total mass of the vehicle, Represents a single gear ratio, represents the total mechanical efficiency of the transmission system, represents the wheel radius, represents the rolling resistance coefficient, represents the acceleration due to gravity, Represents the road slope, represents the air density, Represents the front area of ​​the vehicle, represents the air resistance coefficient; The specific expression formula of the trajectory control model based on the car-following behavior described in step 3 is as follows: （12） （13） （14） （15） （16） （17） （18） （19） （20） （21） （22） （23） （24） In formulas (9)-(24), represents the slack variable, represents the dynamic safety distance between the CAV vehicle and the preceding vehicle, Represents the minimum distance between two vehicles when they are stationary. , as well as represents a constant value greater than zero, Represents the relative speed between the CAV vehicle and the vehicle ahead of it.

4. The method for vehicle control and signal optimization at a mixed-flow intersection under vehicle-road collaboration according to claim 1, characterized in that: The specific steps for optimizing the timing of the signal lights in step (3) are as follows: Step I: Give the shortest green light time to the phase that currently allows traffic ; Step II: Before the green light time of the current green light phase ends, the vehicle queue information and vehicle arrival information of the current green light phase and the next green light phase are input into the observation module, and the traffic intensity of the two phases is obtained through fuzzy reasoning; Step III: Input the current green light phase and the next green light phase strength into the decision module, and obtain the green light extension time of the current green light phase through fuzzy reasoning. If the extension time is greater than 8s and the current phase lasts for the green light time The extended time is less than the maximum green time , then the current green light phase is extended by this time, and the operation of step II is performed; Step IV: If the extension time is greater than 8s, and With extension time greater than or equal to , the current phase green light is extended and then switches to the next green light phase, and then the yellow light After the delay, switch to the next green light phase and perform step Ⅰ. If the extended time is less than or equal to 8s, the yellow light time After the duration has passed, switch to the next green light phase and proceed to step I.

5. The method for vehicle control and signal optimization at a mixed-flow intersection under vehicle-road collaboration according to claim 1, characterized in that: The specific steps of the comparative analysis described in step (4) are as follows: Step 1: Use Python to conduct secondary development of the SUMO simulation platform, build a simulation environment, then collect the instantaneous speed and acceleration of the CAV vehicle, and build a corresponding energy consumption model through the instantaneous gasoline consumption function; Step ②: Collect and analyze the simulation data of each group. When the penetration rate is higher, the energy consumption saved by the joint control of vehicle trajectory and signal is higher than that of vehicle trajectory control. At the same time, as the penetration rate of CAV increases, the average waiting time of vehicles under the joint control of vehicle trajectory and signal is significantly shorter than that under fixed timing and vehicle control only. Step ③: Gradually increase the number of groups of vehicles in the simulation environment. As the traffic volume increases, the traffic flow gradually transitions from free flow to a stable state, and the energy consumption of vehicles shows an upward trend. Compared with the vehicle trajectory control scheme and the fixed timing scheme, the vehicle trajectory and signal joint control scheme has a significant effect on reducing the average energy consumption of vehicles. At the same time, at any level of traffic volume, both vehicle control and signal and vehicle joint control have an optimization effect on the average waiting time. Among them, the signal and vehicle joint control method has the best optimization effect on the average waiting time.

6. The method for vehicle control and signal optimization at a mixed-flow intersection under vehicle-road collaboration according to claim 5 is characterized in that: The specific formula for constructing the energy consumption model in step ① is as follows: （25） （26） In the formula, represents a constant idle gasoline consumption rate, Represents an efficiency parameter and engine efficiency, represents the positive acceleration related to gasoline consumption parameters, represents the vehicle speed, represents the vehicle acceleration, Represents the total power of the vehicle. represents rolling resistance, represents the engine drag coefficient, Represents the coefficient of air resistance.

Images