A closed-loop feedback coordinated control method for highway traffic flow under adverse weather conditions

Through simulation modeling and data collection at the entrance of the highway ramp, a vehicle driving behavior parameter table and a cumulative feedback value table are established, and the traffic flow coordination control in unfavorable weather is achieved, which reduces the risk of accidents and the reduction of traffic efficiency is solved. Traffic flow control problems in the prior art.

CN114863674BActive Publication Date: 2024-11-12YUNNAN TRAFFIC PLANNING DESIGN RESEARCH INSTITUTE CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210380883.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-04-12
Publication Date
2024-11-12
Estimated Expiration
2042-04-12

AI Technical Summary

Technical Problem

The existing technology is difficult to effectively control highway traffic flow in unfavorable weather, resulting in an increase in accident risk and a decrease in traffic efficiency, and lacks effective traffic flow control strategies.

Method used

By simulated modeling the entrance of the highway ramp, vehicle driving behavior parameters and meteorological data are collected, vehicle driving behavior parameter tables and cumulative feedback value tables under different weather conditions, and variable limits are used for information boards and ETC control systems. Speed ​​and ramp flow coordinated control.

Benefits of technology

The risk of highway accidents in unfavorable weather has been reduced, and traffic efficiency and adaptability have been improved. Compared with traditional variable speed limit control, the traffic capacity and average speed drop rate under different weather conditions has been significantly improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114863674B_ABST
    Figure CN114863674B_ABST
Patent Text Reader

Abstract

The present invention relates to a closed-loop feedback coordinated control method for highway traffic flow under adverse weather conditions. First, a highway traffic simulation model is established by using virtual simulation technology; second, data is collected on the driving characteristic parameters of highway vehicles under different weather conditions, and a driving behavior parameter table of highway vehicles under different weather conditions is analyzed and formed; third, the driving behavior parameters of the highway traffic simulation model under adverse weather conditions are configured, and traffic simulation of closed-loop feedback coordinated control of highway traffic flow under the weather conditions is performed to obtain a cumulative feedback value table; finally, in highway operation management, the aforementioned driving behavior parameter table and cumulative feedback value table are used to coordinate control of variable speed limit and ramps for the traffic flow of highways under adverse weather conditions. The present invention can effectively improve the traffic efficiency of highways under various weather conditions such as rain and fog and reduce the accident risk at the ramp entrance, and is easy to promote and apply.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of intelligent transportation and traffic safety, and in particular relates to a closed-loop feedback coordinated control method for highway traffic flow under adverse weather conditions. Background Art

[0002] The traffic safety situation in my country is extremely severe due to adverse weather conditions, with frequent traffic accidents, increased traffic interruptions or delays, and other characteristics. Adverse meteorological conditions such as rain, snow, fog, and road icing have led to more facility damage, traffic accidents, traffic delays, environmental pollution, etc., which have seriously affected public travel and also brought huge social and economic losses. Due to the lack of traffic flow control strategies under adverse weather conditions, highways under rain and fog conditions are prone to congestion at the ramp entrance, resulting in reduced traffic efficiency of highways and a sharp increase in accident risks. Intelligent control of highway traffic flow under adverse weather conditions has technical difficulties such as difficulty in collecting vehicle driving behavior, difficulty in calibrating traffic model parameters, and difficulty in feedback evaluation of control strategies. Therefore, based on virtual simulation technology, an active control strategy for highway traffic safety under adverse weather conditions and its microscopic traffic simulation platform are established, and a closed-loop feedback collaborative control method for highway traffic flow under adverse weather conditions is developed to improve the safe operation capability and traffic service level of smart highways under adverse weather conditions from the source, which is of great significance to supporting a strong transportation country and safe transportation under digital transportation. Summary of the invention

[0003] The purpose of the present invention is to address the deficiencies of the prior art and to provide a closed-loop feedback coordinated control method for highway traffic flow under adverse weather conditions. By modeling and simulating the traffic flow environment at the entrance of the highway ramp, a highway vehicle driving behavior parameter table and a cumulative value table under different weather conditions are obtained. Traffic flow control is performed on the main road and the ramp based on these two tables, thereby reducing the risk of accidents on the highway under adverse weather conditions and further improving the traffic efficiency of the section.

[0004] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0005] A closed-loop feedback coordinated control method for highway traffic flow under adverse weather conditions comprises the following steps:

[0006] Step (1), model establishment: simulate and model the highway, the model includes the main roads and ramps of the highway, sensors, camera devices and information boards; the sensors and camera devices are used to collect the speed, acceleration, flow and density of vehicles;

[0007] Step (2), data collection: collecting meteorological data and vehicle driving behavior parameters under different weather conditions on the expressway, and analyzing to obtain a table of vehicle driving behavior parameters on the expressway under different weather conditions;

[0008] Step (3), model simulation: based on the highway vehicle driving behavior parameter table under different weather conditions obtained in step (2), calibrate the driving behavior parameters of the model under different weather conditions, and simulate the model to obtain a cumulative feedback value table under different weather conditions;

[0009] Step (4), traffic flow coordination control: Based on the cumulative feedback value table under different weather conditions obtained in step (3), as well as the data collected by sensors and cameras on the highway and the current meteorological data, the main line traffic flow is coordinated and controlled with variable speed limits and ramp flow through the information board and ETC control system, and the control is published and displayed.

[0010] Further, preferably, the specific method of step (1) is:

[0011] (1.1) The main highway is a one-way three-lane highway with a length of more than 16 km; the ramp is a one-way single lane highway with a length of 1 km; the ramp connects to the main highway at 8 km;

[0012] (1.2) Nodes 1 and 2 for vehicle output are set at the beginning of the main road and ramp respectively; the vehicle driving path plan is built into the model to ensure that the vehicle can realize autonomous lane change behavior; the range of vehicle flow output by node 1 is 0-3000 vehicles per hour, and the range of vehicle flow output by node 2 is 0-400 vehicles per hour, and nodes 1 and 2 output vehicles according to Poisson distribution;

[0013] (1.3) A vehicle data collection module and a road evaluation module are set up on the main road 3 kilometers before and after the ramp entrance, a total of 6 kilometers of road section; the vehicle data collection module is used to simulate the sensor of the expressway, and the road evaluation module is used to simulate the camera device of the expressway to collect the speed, acceleration, flow and density of the vehicle.

[0014] Furthermore, preferably, in step (2), the meteorological data is collected using an anemometer, a rain gauge, a visibility observation instrument, and a thermometer and hygrometer; and the vehicle driving behavior parameters include headway, following vehicle threshold, parking distance, and following vehicle variable.

[0015] Further, preferably, in step (2), the highway vehicle driving behavior parameter table under different weather conditions is as shown in Table 1;

[0016] Table 1

[0017] Weather conditions Headway (s) Following vehicle threshold(s) Parking distance (m) Following car variable (m) sunny 0.7 -24.6 4.2 12.7 Foggy Day 1.0 -21.9 4.2 12.9 light rain 0.9 -23.5 4.6 13.5 moderate rain 0.8 -24.8 4.6 11.7 heavy rain 0.7 -24.4 3.8 13.5 rainstorm 0.7 -21.0 4.4 12.4 .

[0018] Further, preferably, the specific method of step (3) is:

[0019] According to the highway vehicle driving behavior parameter table under different weather conditions obtained in step (2), randomly select a highway vehicle driving behavior parameter under one weather condition; at the same time, randomly set the value of node1; collect the lane, number, speed, position of the vehicle on the road network, and the traffic volume and average density of the road section through the vehicle data collection module and the road evaluation module; calculate the time when each vehicle passes the ramp intrusion line based on these data, where the intrusion line is 100m after the ramp entrance; subtract the time when the front and rear vehicles pass the same intrusion line to obtain the rear intrusion time PET of the rear vehicle; and count the number of vehicles N1 that exceed the PET threshold;

[0020] Then, a speed limit value v is randomly selected and published on the information board, and the ramp flow u is randomly set. The vehicle data of the road section is collected again, the PET of each vehicle is calculated, and the number of vehicles exceeding the PET threshold after the speed limit value v and the ramp flow u are selected is counted. N2. Finally, the feedback value R of selecting this speed limit value v and ramp flow u under the original traffic flow state is determined by comparing the number of times exceeding the PET threshold.

[0021]

[0022] Where R: the feedback value obtained by selecting the speed limit value v and ramp flow u under the traffic flow state;

[0023] N2: the number of times vehicles on the main road exceed the PET threshold after selecting the speed limit value v and ramp flow u;

[0024] N1: the number of times that vehicles on the main road exceed the PET threshold before selecting the speed limit value v and ramp flow u;

[0025] The value is -10, which indicates the penalty coefficient for increasing the number of highway conflicts due to the speed limit value v;

[0026] According to the obtained reward value R, the Q learning algorithm in reinforcement learning is used to calculate the corresponding cumulative reward value Reward of the main road traffic flow d, speed limit value v and ramp flow u and record them in the cumulative reward value table. The initial value of each Reward in the table is 0;

[0027] Reward(d,v,u) t =(1-μ)Reward(d,v,u) t +μ(R t+1 -0.8maxReward(d′,v′,u′))

[0028] Where, d: traffic flow on the main road; v: speed limit; u: ramp flow; R: feedback value; μ: learning rate, which indicates the absorption rate of the feedback value of each learning by the system; R t+1 is the feedback value obtained in the t+1th iteration;

[0029] Reward(d,v,u) t : The cumulative feedback value corresponding to a certain traffic flow d-ramp flow u-speed limit v on the main road in the tth iteration, the initial value is 0;

[0030] maxReward(d',u',v'): after traffic flow u jumps to the next traffic flow u', the maximum cumulative reward value among a series of ramp flows and speed limit values ​​corresponding to the traffic flow u';

[0031] Under certain weather conditions, it is recorded that one iteration is from setting the main road traffic flow d, selecting the speed limit value v, the ramp flow u to obtaining the corresponding feedback value R and updating the corresponding cumulative feedback value. By iterating different main road traffic flow-speed limit value-ramp flow pairs, the cumulative feedback values ​​corresponding to different main road flows, different speed limits and different ramp flows are recorded in the cumulative feedback value table; after multiple iterations, each cumulative feedback value in the cumulative feedback value table converges; at this point, the cumulative feedback value table under this weather condition converges; the method for obtaining the cumulative feedback value table under other weather conditions is the same as the method for obtaining the cumulative feedback value table under this weather condition.

[0032] Further, preferably, in step (3), during simulation, the main road and the ramp are divided into multiple areas, and the driving behavior parameters of each area are dynamically set according to the vehicle driving behavior parameter table under different weather conditions; The value is -10 and the value of μ is 0.8.

[0033] Furthermore, preferably, in step (4), the sensors on the highway are arranged in groups of two, and the interval between the two sensors in each group is set to 5 m. The two sensors in the group are connected to each other, and when the two sensors collect information of the same vehicle, the vehicle information is transmitted; the sensors are laid on the roadside, and the interval between each group of sensors is set to 100 m.

[0034] Further, preferably, in step (4), the camera device adopts optical camera communication, uses a light emitting diode array and an image sensor as a transmitter and a receiver, and uses visible light or infrared light as a data carrier to implement OCC;

[0035] The camera device processes the vehicle's driving trajectory and calibrates the time of passing through the horizontal line where the sensor is located through image acquisition; if the time when the sensor collects the vehicle data differs from the time calibrated by the camera device by no more than 0.01s, and the vehicle's driving trajectory collected by the camera device is consistent with the vehicle position collected by the sensor, then the data collected by the sensor is determined to be available, otherwise the data of the vehicle is ignored.

[0036] Further, preferably, in step (4), the information board is arranged every 200m on a road section of 6 kilometers, 3 kilometers before and after the ramp entrance; and every 500m on the road section and ramp outside the ramp entrance; the information board is used to display vehicle behavior suggestions under current weather conditions, that is, the optimal speed limit value based on the current traffic flow and weather conditions. Further, preferably, in step (4), when traffic flow coordination control is performed, the current weather conditions are determined by obtaining the current meteorological data, and the data collected by the sensor is compared with the cumulative feedback value table of the corresponding weather, and the optimal speed limit value and the optimal ramp flow under the current traffic flow state are selected, that is, the speed limit value and ramp flow with the maximum cumulative feedback value under the current traffic flow; the optimal speed is published in the speed limit area through the information board; and the optimal ramp flow is sent to the ETC control system to control the traffic flow of the ramp.

[0037] In the present invention, node1 is used to simulate the traffic flow state of the main road, and node2 simulates the vehicle input of the ramp and the control of the ramp flow by the ETC system;

[0038] In the present invention, each vehicle data collection module corresponds to a group of sensors on the expressway, and each road evaluation module corresponds to a camera device on the expressway.

[0039] In the present invention, the vehicle following variable is calibrated by the average value of the difference between the maximum following distance and the minimum following distance of all vehicles in the following process.

[0040] The vehicle driving behavior parameter table described in Table 1 is calibrated according to the Wiedemann99 vehicle following model parameter calibration method using vehicle data on mountain highways under different weather conditions.

[0041] In the present invention, the PET threshold is preferably 2.08s. The preferred value is -10.

[0042] When the method of the present invention is used to control the traffic flow of highways under adverse weather conditions, it should be ensured that the highway section does not form congestion in front of the confluence due to aging and damage of the road surface. For mountain highways, there are 2m1 kilometers before and after the ramp confluence of the highway, and the section is not equipped with sensors and cameras that can obtain the speed, flow and density of vehicles on the lane in real time, and the information board that can change the release form according to different weather conditions. They should be installed to achieve the control of the present invention. The sensor has a more uniform signal compared to the sensor laid in the middle of the road. The sensor is arranged on the side of the road, so its original signal level is low. The signal amplifier integrated in the detector can amplify the original level to obtain a better signal-to-noise ratio, and the cutoff frequency of the signal is set to 100Hz.

[0043] The camera device is an auxiliary equipment. The laying interval is the same as that of the sensor. The collected data is compared with the data collected by the sensor. Compared with other data collection equipment, this set of equipment can significantly improve the accuracy of data collection.

[0044] The camera device uses optical camera communication, that is, using light-emitting diode arrays and image sensors as transmitters and receivers, and visible light or infrared rays as data carriers to achieve OCC. The camera device has a high degree of security and mobile stability. Compared with previous radio frequency-based technologies, the OCC technology has high energy efficiency and high image anti-interference ability. At the same time, the speed acquisition rate and accuracy are faster and better than previous technologies. The camera device can obtain and transmit the accuracy, dimension and safety information of the vehicle, as well as various driving assistance information such as accidents, road construction, maintenance projects and traffic flow through light-emitting diodes and image sensors. It can also automatically determine distortion compensation and realize the calibration of internal camera parameters and external camera device parameters.

[0045] The information board can not only publish vehicle driving behavior suggestions, but also publish real-time speed limit values. The road sections and ramps outside the ramp entrance are arranged every d3 kilometers, and the information published is only vehicle driving behavior suggestions. In particular, the layout interval d3 outside the ramp entrance is greater than the layout interval d2 near the ramp entrance. Preferably, the information board is activated when the sensor and camera device detect that a vehicle has arrived in the area, and is in a dormant state when the sensor and camera device do not detect a vehicle.

[0046] The information board automatically sets the format and color of the information according to different weather conditions. On foggy days, the information board uses a picture with white text on a blue background to display single or multiple lines of speed information and driving behavior advice; under normal circumstances, the information board displays information about congestion, speed, and driving behavior advice in a plain text format; in the case of congestion, the cause of the congestion and driving behavior advice will be displayed in the form of a graphic with black text on a red background plus text. Information displayed in plain text form is generally black text on a red background, and is displayed as a two-line message. The first line of information mainly displays the speed recommendation value, and the second line displays different driving behavior advice according to the current weather conditions. The graphic format is yellow text on a black background in foggy weather, and speed advice and driving behavior advice are mainly displayed through graphics.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] The present invention introduces driving behavior parameters and ramp flow control on the basis of traditional variable speed limit control, which not only enriches the strategic selection of traffic flow control, but also reduces the accident risk at the ramp entrance under various weather conditions. Compared with the traditional variable speed limit control system, the present invention can issue driving behavior suggestions for vehicles according to different weather conditions, making it more adaptable in highway sections with changeable weather conditions. Under the condition of applying traditional variable speed limit control, compared with sunny conditions, the highway capacity under light rain, moderate rain, heavy rain, rainstorm and foggy conditions decreased by 10.5%, 17.4%, 27.1%, 39.7% and 13.5% respectively, and the average speed decreased by 4.7%, 9.8%, 16.1%, 28.2% and 12.1% respectively. The present invention effectively alleviates the decline in capacity and average speed. The corresponding capacity decreases were 5.6%, 9.6%, 22.1%, 29.7% and 10.3%, and the average speed decreases were 3.9%, 8.5%, 13.4%, 26.7% and 11.1%. In summary, this method has a high application value in the field of highway speed limit control and accident prevention under adverse weather conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 is a flow chart of the traffic flow control method of the present invention;

[0050] Figure 2 It is a schematic diagram of the model of the expressway and its pavement facilities of the present invention; wherein a) is a model diagram of the expressway model at the ramp entrance connected to the tunnel; b) is a model diagram of the expressway model at the ramp entrance connected to the bridge; c) indicates the specific meanings of various icons in the models a) and b);

[0051] Figure 3: is a comparison diagram of the model of the present invention before and after control in a sunny scene; wherein a) driving behavior parameters, main road traffic flow, ramp flow and speed value; b) simulation diagram before speed and flow limitation by the method of the present invention; c) simulation diagram after speed and flow limitation by the method of the present invention;

[0052] Figure 4 It is a comparison diagram before and after the control of the model of the present invention in a foggy scene; wherein a) driving behavior parameters, main road traffic flow, ramp flow and speed value; b) simulation diagram before speed and flow limitation by the method of the present invention; c) simulation diagram after speed and flow limitation by the method of the present invention;

[0053] Figure 5 : is a comparison diagram of the model of the present invention before and after control in a light rain scene; wherein a) driving behavior parameters, main road traffic flow, ramp flow and speed value; b) simulation diagram before speed and flow limitation by the method of the present invention; c) simulation diagram after speed and flow limitation by the method of the present invention;

[0054] Figure 6 : is a comparison diagram before and after the control of the model of the present invention in a moderate rain scene; wherein a) driving behavior parameters, main road traffic flow, ramp flow and speed value; b) simulation diagram before speed and flow limitation by the method of the present invention; c) simulation diagram after speed and flow limitation by the method of the present invention;

[0055] Figure 7 : is a comparison diagram of the model of the present invention before and after control in a heavy rain scene; wherein a) driving behavior parameters, main road traffic flow, ramp flow and speed value; b) simulation diagram before speed and flow limitation by the method of the present invention; c) simulation diagram after speed and flow limitation by the method of the present invention;

[0056] Figure 8 It is a comparison diagram before and after the control of the model of the present invention in a rainstorm scene; wherein, a) driving behavior parameters, main road traffic flow, ramp flow and speed value; b) simulation diagram before speed and flow limitation by the method of the present invention; c) simulation diagram after speed and flow limitation by the method of the present invention. DETAILED DESCRIPTION

[0057] The present invention is further described in detail below in conjunction with embodiments.

[0058] Those skilled in the art will appreciate that the following examples are only used to illustrate the present invention and should not be considered to limit the scope of the present invention. If no specific techniques or conditions are specified in the examples, the techniques or conditions described in the literature in the art or the product specifications are used. If the manufacturer of the materials or equipment used is not specified, they are all conventional products that can be purchased.

[0059] A closed-loop feedback coordinated control method for highway traffic flow under adverse weather conditions comprises the following steps:

[0060] Step (1), model establishment: simulate and model the highway, the model includes the main roads and ramps of the highway, sensors, camera devices and information boards; the sensors and camera devices are used to collect the speed, acceleration, flow and density of vehicles;

[0061] Step (2), data collection: collecting meteorological data and vehicle driving behavior parameters under different weather conditions on the expressway, and analyzing to obtain a table of vehicle driving behavior parameters on the expressway under different weather conditions;

[0062] Step (3), model simulation: based on the highway vehicle driving behavior parameter table under different weather conditions obtained in step (2), calibrate the driving behavior parameters of the model under different weather conditions, and simulate the model to obtain a cumulative feedback value table under different weather conditions;

[0063] Step (4), traffic flow coordination control: Based on the cumulative feedback value table under different weather conditions obtained in step (3), as well as the data collected by sensors and cameras on the highway and the current meteorological data, the main line traffic flow is coordinated and controlled with variable speed limits and ramp flow through the information board and ETC control system, and the control is published and displayed.

[0064] Specifically preferred embodiment, the specific method of step (1) is:

[0065] (1.1) The main highway is a one-way three-lane highway with a length of more than 16 km; the ramp is a one-way single lane highway with a length of 1 km; the ramp connects to the main highway at 8 km;

[0066] (1.2) Nodes 1 and 2 for vehicle output are set at the beginning of the main road and ramp respectively; the vehicle driving path plan is built into the model to ensure that the vehicle can realize autonomous lane change behavior; the range of vehicle flow output by node 1 is 0-3000 vehicles per hour, and the range of vehicle flow output by node 2 is 0-400 vehicles per hour, and nodes 1 and 2 output vehicles according to Poisson distribution;

[0067] (1.3) A vehicle data collection module and a road evaluation module are set up on the main road 3 kilometers before and after the ramp entrance, a total of 6 kilometers of road section; the vehicle data collection module is used to simulate the sensor of the expressway, and the road evaluation module is used to simulate the camera device of the expressway to collect the speed, acceleration, flow and density of the vehicle.

[0068] Specifically, in step (2), the meteorological data is collected using an anemometer, a rain gauge, a visibility observation instrument, and a thermometer and hygrometer; the vehicle driving behavior parameters include headway, following vehicle threshold, parking distance, and following vehicle variable.

[0069] Specifically, in step (2), the table of highway vehicle driving behavior parameters under different weather conditions is shown in Table 1.

[0070] Specifically preferred solution, the specific method of step (3) is:

[0071] According to the highway vehicle driving behavior parameter table under different weather conditions obtained in step (2), randomly select a highway vehicle driving behavior parameter under one weather condition; at the same time, randomly set the value of node1; collect the lane, number, speed, position of the vehicle on the road network, and the traffic volume and average density of the road section through the vehicle data collection module and the road evaluation module; calculate the time when each vehicle passes the ramp intrusion line based on these data, where the intrusion line is 100m after the ramp entrance; subtract the time when the front and rear vehicles pass the same intrusion line to obtain the rear intrusion time PET of the rear vehicle; and count the number of vehicles N1 that exceed the PET threshold;

[0072] Then, a speed limit value v is randomly selected and published on the information board, and the ramp flow u is randomly set. The vehicle data of the road section is collected again, the PET of each vehicle is calculated, and the number of vehicles exceeding the PET threshold after the speed limit value v and the ramp flow u are selected is counted. N2. Finally, the feedback value R of selecting this speed limit value v and ramp flow u under the original traffic flow state is determined by comparing the number of times exceeding the PET threshold.

[0073]

[0074] Where R: the feedback value obtained by selecting the speed limit value v and ramp flow u under the traffic flow state;

[0075] N2: the number of times vehicles on the main road exceed the PET threshold after selecting the speed limit value v and ramp flow u;

[0076] N1: the number of times that vehicles on the main road exceed the PET threshold before selecting the speed limit value v and ramp flow u;

[0077] represents the penalty coefficient for increasing the number of highway conflicts due to the speed limit value v;

[0078] According to the obtained reward value R, the Q learning algorithm in reinforcement learning is used to calculate the corresponding cumulative reward value Reward of the main road traffic flow d, speed limit value v and ramp flow u and record them in the cumulative reward value table. The initial value of each Reward in the table is 0;

[0079] Reward(d,v,u) t =(1-μ)Reward(d,v,u) t +μ(R t+1 -0.8maxReward(d′,v′,u′))

[0080] Where, d: traffic flow on the main road; v: speed limit; u: ramp flow; R: feedback value; μ: learning rate, which indicates the absorption rate of the feedback value of each learning by the system; R t+1 is the reward value obtained in the t+1th iteration; Reward(d,v,u) t : The cumulative reward value corresponding to a traffic flow d-ramp flow u-speed limit v on the main road in the tth iteration, with an initial value of 0; maxReward(d',u',v'): After the traffic flow u jumps to the next traffic flow u', the maximum cumulative reward value among a series of ramp flows and speed limit values ​​corresponding to the traffic flow u';

[0081] Under certain weather conditions, it is recorded that one iteration is from setting the main road traffic flow d, selecting the speed limit value v, the ramp flow u to obtaining the corresponding feedback value R and updating the corresponding cumulative feedback value. By iterating different main road traffic flow-speed limit value-ramp flow pairs, the cumulative feedback values ​​corresponding to different main road flows, different speed limits and different ramp flows are recorded in the cumulative feedback value table; after multiple iterations, each cumulative feedback value in the cumulative feedback value table converges; at this point, the cumulative feedback value table under this weather condition converges; the method for obtaining the cumulative feedback value table under other weather conditions is the same as the method for obtaining the cumulative feedback value table under this weather condition.

[0082] In a specific preferred solution, in step (3), during simulation, the main road and the ramp are divided into multiple areas, and the driving behavior parameters of each area are dynamically set according to the vehicle driving behavior parameter table under different weather conditions; The value is -10 and the value of μ is 0.8.

[0083] In a specific preferred embodiment, in step (4), the sensors on the highway are arranged in groups of two, and the interval between the two sensors in each group is set to 5 m. The two sensors in the group are connected to each other. When the two sensors collect information about the same vehicle, the vehicle information is transmitted. The sensors are laid on the roadside, and the interval between each group of sensors is set to 100 m.

[0084] Specifically, in step (4), the camera device uses optical camera communication, uses a light-emitting diode array and an image sensor as a transmitter and a receiver, and uses visible light or infrared light as a data carrier to implement OCC;

[0085] The camera device processes the vehicle's driving trajectory and calibrates the time of passing through the horizontal line where the sensor is located through image acquisition; if the time when the sensor collects the vehicle data differs from the time calibrated by the camera device by no more than 0.01s, and the vehicle's driving trajectory collected by the camera device is consistent with the vehicle position collected by the sensor, then the data collected by the sensor is determined to be available, otherwise the data of the vehicle is ignored.

[0086] According to a specific preferred solution, in step (4), information boards are arranged every 200 m on a road section of 6 km, 3 km before and after the ramp entrance; and every 500 m on road sections and ramps outside the ramp entrance; the information boards are used to display vehicle behavior recommendations under current weather conditions, i.e., the optimal speed limit value based on current traffic flow and weather conditions.

[0087] Specifically, in step (4), when traffic flow is coordinated and controlled,

[0088] The current weather conditions are determined by obtaining the current meteorological data, and the data collected by the sensors are compared with the cumulative feedback value table of the corresponding weather. The optimal speed limit and ramp flow under the current traffic flow state are selected, that is, the speed limit and ramp flow with the largest cumulative feedback value under the current traffic flow; the optimal speed is announced in the speed limit area through the information board; the optimal ramp flow is sent to the ETC control system to control the traffic flow on the ramp.

[0089] Application Examples

[0090] Step 1: Select a section of highway that meets the requirements. The selected highway section does not have congestion in front of the confluence entrance due to aging and damage of the road surface, and the meteorological conditions in the area are poor, with frequent unfavorable weather. Customized sensors, cameras and information boards are installed 3 kilometers before and after the ramp confluence entrance of the highway, a total of 6 kilometers.

[0091] The sensors are arranged in groups of two. When laid, the interval between the two sensors in each group is 5m. The sensor will flash green after the vehicle passes. The two sensors in the group are connected to each other. When both sensors flash, they transmit the current vehicle information to the system. The sensors are set on the roadside. The original level can be amplified by the signal amplifier integrated in the detector to obtain a better signal-to-noise ratio, thereby obtaining accurate data, and the cutoff frequency of the signal is set to 100Hz. Each group of sensors is laid every 100m.

[0092] The camera device adopts optical camera communication. It acquires and transmits the accuracy, dimension and safety information of the vehicle, as well as various driving assistance information, such as data on accidents, road construction, maintenance works and traffic flow, through light-emitting diodes and image sensors. In addition, the camera device can automatically determine the distortion compensation device and realize the calibration of the parameters of the internal camera and the external camera device.

[0093] The information board can publish vehicle driving behavior suggestions and real-time speed limit values. The information board is arranged every 200m on a total of 6 kilometers of road sections 3 kilometers before and after the ramp entrance; it is arranged every 500m on sections and ramps outside the ramp entrance; the information that can be published is only vehicle driving behavior suggestions. The information board automatically sets the format and color of the information according to different weather conditions. Under foggy weather, the information board uses a picture with white text on a blue background to display single or multiple lines of speed information and driving behavior suggestions; under normal circumstances, the information board displays information about congestion, speed and driving behavior suggestions in a plain text format; in the case of congestion, the cause of congestion and driving behavior suggestions will be displayed in the form of a graphic with black text on a red background plus text. The information displayed in plain text form is generally black text on a red background, and is displayed in a two-line message. The first line of information mainly displays the speed suggestion value, and the second line displays different driving behavior suggestions according to the current weather conditions. The graphic format is yellow text on a black background in foggy weather, and speed suggestions and driving behavior suggestions are mainly displayed through graphics.

[0094] Step 2: Establish a parameter table of highway vehicle driving behavior under different weather conditions. The parameters of vehicle driving behavior under different weather conditions, namely parking distance, following vehicle threshold, headway, and following vehicle variables, are determined by analyzing local environmental data to form a parameter table of highway vehicle driving behavior under weather conditions, as shown in Table 1. Anemometers, rain gauges, and fog visibility observation instruments are set up on the highway to collect rainfall and wind speed environmental data under different weather conditions to determine the weather conditions of the highway. The driving behavior of the vehicle is determined through the highway vehicle driving behavior parameter table, and vehicle behavior suggestions under current weather conditions are published using information boards.

[0095] Step 3: Build a model of the highway in a virtual environment, and model sensors, cameras, and information boards.

[0096] 1) The main road of the model is a one-way three-lane road with a length of more than 16 kilometers. The ramp is a one-way single lane road with a length of 1 kilometer. The ramp connects to the main road at 8 kilometers.

[0097] 2) Nodes node1 and node2 for vehicle output are set at the beginning of the main road and ramp. Node1 is used to simulate the traffic flow state of the main road, and node2 controls the vehicle input of the ramp and the control of the ramp flow by the ETC system. At the same time, the model sets the vehicle driving path decision to ensure that the vehicle can realize autonomous lane change behavior. The range of vehicle flow output by node node1 is 0-3000 vehicles per hour, and the range of vehicle flow output by node node2 should be 0-400 vehicles per hour. The output interval of the vehicle is 50 vehicles per hour. And nodes node1 and node2 output vehicles according to Poisson distribution.

[0098] 3) The main roads and ramps of the model are divided into different areas, and the driving behavior parameters of each area can be dynamically set according to the vehicle driving behavior parameter table under different weather conditions to simulate adverse weather conditions. The area near the ramp entrance and the ramp can also control the speed of the vehicle in the area near the entrance and the flow of the ramp in real time through the cumulative feedback value table.

[0099] 4) On the main road of the model, 3 kilometers before and after the ramp entrance, a total of 6 kilometers of road sections are set up with vehicle data collection modules and road evaluation modules to simulate the customized sensors and camera devices of the expressway, which are used to collect information on the speed, acceleration, and flow and density of vehicles on the main road in the simulation. Each vehicle data collection module corresponds to a group of sensors on the expressway, and each road evaluation module corresponds to a customized camera device on the expressway.

[0100] The model is Figure 2 As shown, the figure includes three parts: a), b), and c). Among them, a) and b) are model diagrams of the model at the ramp entrance connecting the tunnel and the ramp entrance connecting the bridge, which mainly show the laying positions of sensors, information boards and cameras near the ramp entrance, as well as the setting positions of data collection points in the corresponding simulation model. c) represents the specific meanings of various icons in the a) and b) models. Since the collection areas of the bridge and the tunnel are the same, the present invention combines the two in subsequent simulations without making any distinction.

[0101] Step 4: Simulate the expressway in the virtual environment, with a single simulation time of 2000 seconds.

[0102] At the beginning of the simulation, the driving behavior parameters of the vehicle under one weather condition are randomly selected according to the table of driving behavior parameters of highway vehicles under different weather conditions.

[0103] When the simulation is running for 50 seconds, a random output value is set for the vehicle output on the main road, ranging from 0 to 3000 vehicles per hour, so that the traffic flow on the highway changes with the change of output volume.

[0104] During the period from 200 seconds to 700 seconds, the vehicle data collection module and road evaluation module on the road network collect the lane, number, speed, location, traffic volume and average density of the vehicle on the road network. The PET of each vehicle is calculated based on these data, and the number of vehicles exceeding the PET threshold N1 is counted.

[0105] When the simulation reaches 800 seconds, a speed limit value v is randomly selected and published on the variable speed limit information board. The optional speed limits are: 30km / h, 40km / h, 50km / h, 60km / h, 70km / h, 80km / h, 90km / h, 100km / h, 110km / h, 120km / h. At the same time, the ramp flow u is randomly set, ranging from 0 to 400 vehicles per hour.

[0106] During the period from 900 seconds to 1400 seconds, the vehicle data collection module and road evaluation module on the road network collect data on the lane, number, speed, location, traffic flow and average density of the vehicle on the road network. Based on these data, the PET of each vehicle is calculated, and the number of vehicles N2 exceeding the PET threshold after selecting the speed limit value v and ramp flow u is counted.

[0107] By comparing the number of times the PET threshold is exceeded, the feedback value of selecting the speed limit value v and ramp flow u under the original traffic flow state is determined.

[0108]

[0109] Among them, R represents the feedback value obtained by selecting the speed limit value v and ramp flow u under the traffic flow state; N2 represents the number of vehicles on the main road exceeding the PET threshold after the speed limit value v and ramp flow u are selected; N1 represents the number of vehicles on the main road exceeding the PET threshold before the speed limit value v and ramp flow u are selected The value is -10.

[0110] Then use v, u and R to iterate the corresponding cumulative feedback value until convergence.

[0111] Reward(d,v,u) t =(1-μ)Reward(d,v,u) t +μ(R t+1 -0.8maxReward(d′,v′,u′))

[0112] Among them, d represents the traffic flow of the main road; v represents the speed limit; u represents the ramp flow; R represents the feedback value; μ represents the learning rate, which is 0.8, that is, the absorption rate of the feedback value learned by the system each time; Reward represents the cumulative feedback value, and its initial value is 0. t+1is the reward value obtained in the t+1th iteration; Reward(d,v,u) t : The cumulative reward value corresponding to a traffic flow d-ramp flow u-speed limit v on the main road in the tth iteration, with an initial value of 0; maxReward(d',u',v'): After the traffic flow u jumps to the next traffic flow u', the maximum cumulative reward value among a series of ramp flows and speed limit values ​​corresponding to the traffic flow u';

[0113] In the cumulative feedback value table, each traffic flow d corresponds to multiple speed limit values ​​and ramp flow rates, among which the speed limit value and ramp flow rate corresponding to the traffic flow with the largest cumulative feedback value are the optimal speed limit value and optimal ramp flow rate under the traffic flow.

[0114] Table 2

[0115]

[0116]

[0117] Since there are many parameters in the cumulative feedback value table, only part of the data in the cumulative feedback value table is shown here, and the composition of the rest of the cumulative feedback value table is consistent with this table. Table 2 is a partial cumulative feedback value table under moderate rainy weather, in which the first column is the different combinations of speed limit value-ramp flow, and the first row is the flow of the main road. The remaining values ​​are the cumulative feedback values ​​of speed limit value-ramp flow and main roads. The speed limit value-ramp flow combination with the highest cumulative feedback value corresponding to each main road traffic flow is the optimal speed limit value and optimal ramp flow for the flow. Note: The cumulative feedback value table obtained according to different highway models will be different, and it is necessary to iterate according to the method of the present invention to obtain the corresponding cumulative feedback value table.

[0118] Figure 3 to Figure 8 The following are the simulation diagrams of the model under different weather conditions, including a) driving behavior parameters, main road traffic flow, ramp flow and speed values, b) simulation diagram before speed limit and flow limit, and c) simulation diagram after speed limit and flow limit. Among them, a) includes the driving behavior parameters set according to the driving behavior parameter table under the current weather conditions, and the table includes the main road traffic flow, average speed value and ramp flow of vehicles in the current model, as well as the ramp flow and speed limit value after speed limit and flow limit.

[0119] Step 5: For each weather condition, repeat the iteration until the accumulated values ​​of different traffic flows d and their corresponding speed limits v and ramp flows u under each weather condition converge, thus forming the accumulated feedback value table of the highway.

[0120] The system determines the current weather conditions through the anemometer, rain gauge and fog area visibility observation instrument, and then determines the current vehicle driving behavior parameters by comparing the current weather conditions with the highway vehicle driving behavior parameter table under different weather conditions, and then uses the information board to publish the driving behavior suggestions in a set form.

[0121] The system obtains the average vehicle speed, density and traffic volume of the main road near the ramp through the sensor and camera device, selects the optimal speed and ramp flow of the current traffic flow state by comparing with the cumulative feedback value table, and then publishes the speed in a set form through the information board. The optimal ramp flow is sent to the ETC control system, which controls the flow of the ramp while controlling the main road.

[0122] The present invention introduces driving behavior parameters and ramp flow control on the basis of traditional variable speed limit control, which not only enriches the strategy selection of traffic flow control, but also reduces the accident risk at the ramp entrance under various weather conditions. In summary, the present method has high application value in the field of high-speed speed limit control and accident prevention.

[0123] In moderate rainy weather, Figure 6 a), referring to the driving behavior parameter table, it can be seen that the driving behavior of the vehicle in the current model should be set to the headway time of 0.8s, the following threshold of -24.8s, the parking distance of 4.6m, and the following variable of 11.7m. The main road traffic flow of the current road is collected and extracted by sensors and cameras. According to the partial cumulative feedback value table obtained by simulation under this weather condition, as shown in Table 3. The optimal speed limit value under the current traffic flow is the speed limit value corresponding to the maximum cumulative feedback value - ramp flow, that is, the speed limit value is set to 100km / h, and the ramp flow is set to 350veh / h. Finally, the information board on the highway is used to publish the driving behavior suggestions and speed limit values ​​to the driver, and the corresponding ramp flow is sent to the ETC control system. The parameters are imported into the model for synchronous simulation, and the Figure 6 From b) and c), we can see that there was obvious congestion in the ramp merging area before the speed and flow restrictions. After the speed and flow restrictions, the congestion in the merging area was effectively alleviated.

[0124] The model simulation diagrams under other weather conditions are as follows: Figure 3-5 , Figure 7-8 shown.

[0125] Table 3

[0126]

[0127] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A closed-loop feedback coordinated control method for highway traffic flow under adverse weather conditions, characterized in that: The steps include: Step (1), model establishment: simulate and model the highway, the model includes the main roads and ramps of the highway, sensors, camera devices and information boards; the sensors and camera devices are used to collect the speed, acceleration, flow and density of vehicles; Step (2), data collection: collecting meteorological data and vehicle driving behavior parameters under different weather conditions on the expressway, and analyzing to obtain a table of vehicle driving behavior parameters on the expressway under different weather conditions; Step (3), model simulation: based on the highway vehicle driving behavior parameter table under different weather conditions obtained in step (2), calibrate the driving behavior parameters of the model under different weather conditions, and simulate the model to obtain a cumulative feedback value table under different weather conditions; Step (4), traffic flow coordination control: Based on the cumulative feedback value table under different weather conditions obtained in step (3), as well as the data collected by sensors and cameras on the highway and the current meteorological data, the main line traffic flow is coordinated and controlled with variable speed limits and ramp flow through the information board and ETC control system, and the control is published and displayed; The specific method of step (1) is: (1.1) The main highway is a one-way three-lane highway with a length of more than 16 km; the ramp is a one-way single lane highway with a length of 1 km; the ramp is connected to the main highway at 8 km; (1.2) Nodes 1 and 2 for vehicle output are set at the beginning of the main road and ramp respectively; the vehicle driving path plan is built into the model to ensure that the vehicle can realize autonomous lane change behavior; the range of vehicle flow output by node 1 is 0-3000 vehicles per hour, and the range of vehicle flow output by node 2 is 0-400 vehicles per hour, and nodes 1 and 2 output vehicles according to Poisson distribution; (1.3) A vehicle data collection module and a road evaluation module are set up on the main road 3 kilometers before and after the ramp entrance, a total of 6 kilometers of road section; the vehicle data collection module is used to simulate the sensor of the expressway, and the road evaluation module is used to simulate the camera device of the expressway to collect the speed, acceleration, flow and density of the vehicle; The specific method of step (3) is: According to the highway vehicle driving behavior parameter table under different weather conditions obtained in step (2), randomly select a highway vehicle driving behavior parameter under one weather condition; and randomly set the value of node1; collect the lane, number, speed, position of the vehicle on the road network, and the traffic volume and average density of the road section through the vehicle data collection module and the road evaluation module; Based on these data, the time when each vehicle passes through the ramp intrusion line is calculated, where the intrusion line is 100 meters after the ramp entrance; the time when the front and rear vehicles pass the same intrusion line is subtracted to obtain the rear intrusion time PET of the rear vehicle; the number of vehicles exceeding the PET threshold N1 is counted; Then, a speed limit value v is randomly selected and published on the information board, and the ramp flow u is randomly set. The vehicle data of the road section is collected again, the PET of each vehicle is calculated, and the number of vehicles exceeding the PET threshold after the speed limit value v and the ramp flow u are selected is counted. N2. Finally, the feedback value R of selecting this speed limit value v and ramp flow u under the original traffic flow state is determined by comparing the number of times exceeding the PET threshold. , Where R: the feedback value obtained by selecting the speed limit value v and ramp flow u under the traffic flow state; N2: the number of times vehicles on the main road exceed the PET threshold after selecting the speed limit value v and ramp flow 𝑢; N1: the number of times that vehicles on the main road exceed the PET threshold before selecting the speed limit value v and the ramp flow 𝑢; : represents the penalty coefficient for the increase in the number of highway conflicts caused by the speed limit value v; According to the obtained reward value R, the corresponding cumulative reward value Reward of the main road traffic flow d, speed limit value v and ramp flow u is calculated and recorded in the cumulative reward value table. The initial value of each Reward in the table is 0; , Where, d: traffic flow on the main road; v: speed limit; u: ramp flow; R: feedback value; μ: learning rate, which indicates the system's absorption rate of the feedback value of each learning; is the feedback value obtained in the t+1th iteration; : The cumulative reward value corresponding to a certain traffic flow d-ramp flow u-speed limit v on the main road in the tth iteration, the initial value is 0; maxReward(d',u',v'): traffic flow u Jump to next traffic flow u’ After that, the traffic flow u’ The maximum cumulative feedback value among the corresponding series of ramp flow and speed limit values; Under certain weather conditions, it is recorded that one iteration is from setting the main road traffic flow d, selecting the speed limit value v, the ramp flow u to obtaining the corresponding feedback value R and updating the corresponding cumulative feedback value. By iterating different main road traffic flow-speed limit value-ramp flow pairs, the cumulative feedback values ​​corresponding to different main road flows, different speed limits and different ramp flows are recorded in the cumulative feedback value table; after multiple iterations, each cumulative feedback value in the cumulative feedback value table converges; at this point, the cumulative feedback value table under this weather condition converges; the method for obtaining the cumulative feedback value table under other weather conditions is the same as the method for obtaining the cumulative feedback value table under this weather condition.

2. The closed-loop feedback coordinated control method for highway traffic flow under adverse weather conditions according to claim 1 is characterized in that: In step (2), the meteorological data is collected using an anemometer, a rain gauge, a visibility observation instrument, and a thermometer and hygrometer; the vehicle driving behavior parameters include headway, following vehicle threshold, parking distance, and following vehicle variable.

3. The closed-loop feedback coordinated control method for highway traffic flow under adverse weather conditions according to claim 1 is characterized in that: In step (2), the highway vehicle driving behavior parameter table under different weather conditions is shown in Table 1; 。 4. The closed-loop feedback coordinated control method for highway traffic flow under adverse weather conditions according to claim 1 is characterized in that: In step (3), during simulation, the main road and the ramp are divided into multiple areas, and the driving behavior parameters of each area are dynamically set according to the vehicle driving behavior parameter table under different weather conditions; The value is -10 and the value of μ is 0.

8.

5. The closed-loop feedback coordinated control method for highway traffic flow under adverse weather conditions according to claim 1 is characterized in that: In step (4), the sensors on the highway are arranged in groups of two, and the interval between the two sensors in each group is set to 5 m. The two sensors in the group are connected to each other. When the two sensors collect information about the same vehicle, the vehicle information is transmitted. The sensors are laid on the roadside, and the interval between each group of sensors is set to 100 m.

6. The closed-loop feedback coordinated control method for highway traffic flow under adverse weather conditions according to claim 1 is characterized in that: In step (4), the camera device adopts optical camera communication, uses a light emitting diode array and an image sensor as a transmitter and a receiver, and uses visible light or infrared light as a data carrier to implement OCC; The camera device processes the vehicle's driving trajectory and calibrates the time of passing through the horizontal line where the sensor is located through image acquisition; if the time when the sensor collects the vehicle data differs from the time calibrated by the camera device by no more than 0.01s, and the vehicle's driving trajectory collected by the camera device is consistent with the vehicle position collected by the sensor, then the data collected by the sensor is determined to be available, otherwise the data of the vehicle is ignored.

7. The closed-loop feedback coordinated control method for highway traffic flow under adverse weather conditions according to claim 1 is characterized in that: In step (4), the information boards are arranged every 200 m on a road section of 6 km, 3 km before and after the ramp entrance; and every 500 m on the road section and ramps outside the ramp entrance; The information board is used to display vehicle behavior recommendations under current weather conditions, that is, the optimal speed limit value based on current traffic flow and weather conditions.

8. The closed-loop feedback coordinated control method for highway traffic flow under adverse weather conditions according to claim 1 is characterized in that: In step (4), when traffic flow is coordinated and controlled, the current weather conditions are determined by obtaining the current meteorological data, and the data collected by the sensors are compared with the cumulative feedback value table of the corresponding weather, and the best speed limit value and the best ramp flow rate under the current traffic flow state are selected, that is, the speed limit value and ramp flow rate with the maximum cumulative feedback value under the current traffic flow; The optimal speed is announced in the speed limit area through the information board; the optimal ramp flow is sent to the ETC control system to control the traffic flow on the ramp.