CN104485003B - A kind of intelligent traffic signal control method based on pipeline model - Google Patents

Abstract

The invention provides an intelligent traffic signal control method based on a pipeline model, and the method reduces the running quality of vehicles passing through intersections. The core idea of the intelligent traffic signal control method based on the pipeline model is: first, based on the communication between the vehicle and the infrastructure in the Internet of Vehicles, relying on the roadside unit to establish a pipeline model for accurate detection of vehicle information. Then according to the model, real-time and accurate collection of vehicle information in and out of the pipeline, and finally use this information to rationally allocate the green light passage time of traffic in each direction. The invention can adapt to the dynamic change of the traffic flow, effectively reduce the average waiting time and the average parking times of the vehicles on the premise of ensuring the traffic volume, and improve the driving quality at the intersection.

Description

An Intelligent Traffic Signal Control Method Based on Pipeline Model

technical field

The invention relates to the field of road traffic signal control, in particular to an intelligent city traffic signal control method.

Background technique

In urban traffic environments, the presence of intersections improves the connectivity of the road network. However, the crossing of traffic flows in different directions increases the degree of congestion at the intersection, which is likely to cause the decline of vehicle driving quality. Especially when the time allocated by the traffic signal control system is unreasonable, it will aggravate the congestion at the intersection. The traffic operation status of urban road intersections is closely related to the traffic operation status of the whole city. Solving traffic problems at intersections is the key to alleviating urban road congestion and improving vehicle driving quality.

Traffic signal control is considered to be one of the most economical and effective ways to increase traffic volume at intersections, and its control methods are mainly divided into fixed timing and adaptive timing. According to the historical data of traffic volume, the fixed timing method allocates the appropriate fixed green time for each direction of the intersection. The adaptive timing method feeds back the effect of the current timing scheme through appropriate algorithms or uses vehicle detection to provide real-time traffic information for dynamically adjusting the timing scheme. Both methods have advantages and disadvantages: the fixed timing method is simple and easy to implement, and is widely used in real life, but it cannot adapt to the high dynamics of traffic flow and reduces the driving quality of vehicles passing through intersections. The adaptive timing method can flexibly adapt to the dynamics of traffic flow, but there are problems such as complex implementation and inaccurate vehicle information acquisition. Compared with the fixed timing method, the adaptive timing method is more flexible and effective. Therefore, researchers have proposed and improved a variety of adaptive traffic signal control methods by using various theoretical knowledge or various hardware and software devices. Such as artificial intelligence and machine learning theory, image and video processing technology, wireless sensor network technology, etc. In recent years, Intelligent Traffic System (Intelligent Traffic System, ITS) has played a key role in improving the transportation efficiency and safety of road traffic. Vehicle Ad-hoc Network (VANET) can be regarded as the product of the rapid development of ITS in the past ten years, which provides a more efficient means for the realization of adaptive traffic signal control system solutions.

There are many existing adaptive traffic signal control methods, but there are defects such as complex implementation or difficult to guarantee the accuracy of vehicle-related information. For example, the processing results of images or videos are closely related to the quality of collected samples, especially in the case of bad weather or traffic jams, the effect of such methods is difficult to be guaranteed. Traffic control based on the "green wave" effect is one of the most efficient traffic control strategies currently recognized by realizing that the traffic flow on the arterial road passes through multiple traffic light intersections without stopping. Although the "Green Wave" solution is efficient, it can only improve the driving quality of the main road, and may have adverse effects on the driving of branch roads. At the same time, these methods all ignore the impact of vehicle type on the allocation time.

Contents of the invention

Aiming at the shortcomings of the existing traffic signal control methods, the invention proposes an intelligent traffic signal control method based on a pipeline model.

The technical solution of the present invention is a pipeline model-based intelligent traffic signal control method, comprising the following steps:

Step 1, establish a pipeline model, the pipeline model includes a roadside unit, a data center server and a traffic control system; the roadside unit is used to collect vehicle-related information, and the data center server is used to process roadside The vehicle information submitted by the unit, the traffic control system is used to allocate reasonable green light passing time for each intersection. Go to step 2;

Step 2. When the vehicle enters the pipeline, it sends an arrival message AM _i to the first roadside unit RSU _1. The content of the arrival message AM _i includes the identifier of the vehicle, the driving lane, the type of the vehicle, the time of arrival at the pipeline, and the priority of the vehicle , i represents the i-th vehicle; when the vehicle leaves the pipeline, it sends a departure message DM _i to the second roadside unit RSU ₂ , and the content of the departure message DM _i contains the identifier of the vehicle; the first roadside unit RSU ₁ receives the arrival message After AM _i , the data center server records the relevant information of the vehicle; after the second roadside unit RSU ₂ receives the leaving message DM _i , the data center server deletes the relevant information of the vehicle. At the same time, the message retransmission strategy and the obsolete information deletion strategy are used to process the information. Go to step 3;

Step 3: Divide the vehicles into three categories: large, medium, and small, and assign influence weights W _x , W _y , and W _z respectively. Small vehicles are the standard impact weights, and the data center server accumulates the weights of various types of vehicles in the pipeline , get the weight value affecting the distribution of green light time at the current moment, record it as Flow_C, and submit it to the traffic control system. Go to step 4;

Step 4, the traffic control system checks whether the lane in the current direction has the right to control the green light time, if yes, go to step 5, otherwise go to step 2;

Step 5, the traffic control system compares the weight value Flow_C and the weight threshold Flow_T of the vehicles in the pipeline that affect the distribution of green light time. If Flow_C>Flow_T, it means that the degree of road congestion is high, then go to step 6, otherwise go to step 8;

Step 6, allocate the green light time for the traffic flow, and continue to compare the weight value Flow_C and the weight threshold Flow_T of the vehicles in the pipeline that affect the green light time allocation. If Flow_C>Flow_T, it means that the road congestion is still at a relatively high level, go to step 7, otherwise go to step 8;

Step 7, the traffic control system judges whether the current green light duration T _G is greater than the longest green light time T _maxG , if yes, go to step 9, otherwise go to step 6;

Step 8, the traffic control system allocates the shortest green light time T _minG to the current lane, and go to step 9;

Step 9, the traffic control system transfers the control right of the green light time of the current lane to the lane in the next direction, and ends the process.

The message retransmission strategy and obsolete information deletion strategy in step 2 are as follows: the vehicle has a backup when sending the arrival message, if it does not receive a response from the first roadside unit RSU ₁ within time γ, then send a backup Assuming that vehicle i sends arrival message AM _i and departure message DM _i to the first roadside unit RSU ₁ and the second roadside unit RSU ₂ respectively when entering and leaving the pipeline, under the premise of adopting the message retransmission strategy, the first The result of the message received by the roadside unit RSU ₁ and the second roadside unit RSU ₂ has the following four situations:

(1) The first roadside unit RSU ₁ receives the arrival message AM _i , and the second roadside unit RSU ₂ receives the departure message DM _i : the pipeline model normally records the entry and exit of the vehicle;

(2) The first roadside unit RSU ₁ does not receive the arrival message AM _i , and the second roadside unit RSU ₂ receives the departure message DM _i : the pipeline model does not record the relevant information of the vehicle, does not include the vehicle value, and does not include Calculation of input weight;

(3) The first roadside unit RSU ₁ receives the arrival message AM _i , but the second roadside unit RSU ₂ does not receive the departure message DM _i : the pipeline model does not record the relevant information of the vehicle, does not include the vehicle value, and does not Bring in the calculation of weight;

(4) The first roadside unit RSU ₁ did not receive the arrival message AM _i , and the second roadside unit RSU ₂ did not receive the departure message DM _i : the pipeline model does not record the relevant information of the vehicle, does not count the vehicle value, and does not into the calculation of weights.

In step 3, the method for calculating the weight value Flow_C that ultimately affects the distribution of green light time is: consider the distribution of green light time for traffic flow in one direction, ignore the time distribution of right-turn traffic flow, and assume that the total number of vehicles in the current road pipeline is N, where the left The proportions of turning vehicles, straight-going vehicles and right-turning vehicles are respectively N _a , N _b , N _c , so that the influence weight of a single vehicle is W _i , then:

$\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; flag</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; W</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

Among them, flag _i represents the identification bit of the i-th vehicle's driving direction, W _i represents the influence weight of the i-th vehicle, and the values of flag _i and W _i are shown in formulas (2) and (3):

The data center server receives the vehicle data processed by the first roadside unit RSU ₁ and the second roadside unit RSU ₂ , obtains the weight value Flow_C that affects the distribution of green light time through formulas (1), (2) and (3), and It is handed over to the traffic control system.

The technical effect of the present invention is: an intelligent traffic signal control method based on a pipeline model, firstly based on the communication between the vehicle and the infrastructure in the Internet of Vehicles, and relying on the roadside unit to establish a pipeline model for accurately detecting vehicle information. Then according to the model, real-time and accurate collection of vehicle information in and out of the pipeline, and finally use this information to rationally allocate the green light passage time of traffic in each direction. The invention can adapt to the dynamic change of the traffic flow, effectively reduce the average waiting time and the average parking times of the vehicles on the premise of ensuring the traffic volume, and improve the driving quality at the intersection.

Description of drawings

Fig. 1 is the structural diagram of pipeline model;

Figure 2-1 is a format diagram of vehicle arrival message;

Figure 2-2 is a format diagram of the vehicle leaving message;

Figure 3 is an application scenario diagram based on the pipeline model;

Fig. 4 is a flowchart of an intelligent traffic signal control method based on a pipeline model.

detailed description

The research of the present invention finds that in the current traffic signal control method, there are problems such as complex implementation of the solution, difficulty in ensuring the real-time and accuracy of obtaining vehicle-related information, and the like. Accordingly, the present invention provides a new intelligent traffic signal control method. An intelligent traffic signal control method based on a pipeline model, which is used to reasonably allocate the green time of traffic flows in all directions at intersections. The basic idea of the intelligent traffic signal control method based on the pipeline model is: first, based on the vehicle-to-infrastructure (V2I) communication in VANET, relying on the road side unit (Road Side Unit, RSU) to establish a roadside unit (RSU) for Pipeline model for accurate detection of vehicle information. Then, according to the model, real-time and accurate information of vehicles entering and leaving the pipeline is collected, and finally the information is used to reasonably allocate the green time of traffic flow in each direction.

In order to allocate signal time according to the traffic flow, it is necessary to obtain the traffic density information near the intersection. Although the method of calculating the vehicle density based on the clustering algorithm and using video or image processing technology can estimate the vehicle density, there are the following problems: (1) The calculation of the vehicle density is not accurate enough and is easily interfered by objective factors. For example, when the traffic density is too large or the weather is bad, the results obtained by the above two types of methods are difficult to be guaranteed; (2) The impact of the vehicle type on the allocation time is ignored. Therefore, the present invention first proposes a pipeline model capable of accurately detecting vehicle information based on V2I communication in VANET.

The essence of the pipeline model is to collect and process relevant information of vehicles in the pipeline with the help of roadside units, including vehicle identifiers, driving lanes, vehicle types, arrival time at the pipeline, and vehicle priority. Traffic control systems use this information to assign signal times to intersections. The biggest advantage of the pipeline model is that it can accurately obtain the real-time situation of the traffic flow in the pipeline.

As shown in Figure 1, the basic components of the pipeline model mainly include: roadside unit, data center server and traffic control system. The roadside unit is used to collect relevant information of vehicles, the data center server is used to process the vehicle information submitted by the roadside unit, and the traffic control system is used to allocate reasonable green light passing time for each intersection. The road section between RSU ₁ and RSU ₂ is called a pipeline. When a vehicle enters the pipeline, it sends an arrival message (Arrival Message, AM) to the first roadside unit RSU ₁ , and the AM content includes the vehicle identifier, driving lane, vehicle type, arrival time at the pipeline, and vehicle priority. When the vehicle leaves the pipeline, it sends a departure message (Depart Message, DM) to the second roadside unit RSU ₂ , and the content of the DM only contains the identifier of the vehicle. The formats of the two messages are shown in Figure 2. After receiving the AM, the first roadside unit RSU ₁ records the relevant information of the vehicle; after the second roadside unit RSU ₂ receives the DM, it deletes the relevant information of the vehicle. The two jointly maintain a real-time information database of vehicles in the pipeline. The data center server collects and processes the relevant information of the vehicle, and submits it to the traffic control system to control the distribution of road signal time.

As shown in Fig. 3, it is an embodiment diagram based on the pipeline model. It is the situation before the crossroads where the traffic flow from west to east passes through the intersection. The length of the road is L, the length of the pipeline is D, and the roadside units RSU ₁ and RSU ₂ are the key components of the pipeline model, which are located on both sides of the pipeline to collect the information of vehicles entering and leaving the pipeline.

The pipeline model records the information of vehicles entering and leaving the pipeline in real time. Due to the relatively large traffic flow passing through the intersection during the day, there may be a possibility of message transmission failure during the communication process between vehicles and roadside units, which will cause certain deviations in the recorded results. Although this deviation has little effect on the distribution of signal periods, it may have a more obvious impact if it accumulates for a long time. Therefore, the present invention adopts a message retransmission strategy and an obsolete information deletion strategy to enhance the reliability of the pipeline model.

The core of the message retransmission process is:

The vehicle has a backup when sending the arrival message, if it does not receive a response from RSU ₁ within time γ, it sends a backup message. Assume that vehicle i sends an arrival message AM _i and a departure message DM _i to the first roadside unit RSU ₁ and the second roadside unit RSU ₂ respectively when entering and leaving the pipeline. Under the premise of adopting the message retransmission strategy, the result of RSU ₁ and RSU ₂ receiving the message has the following four situations:

(1) The first roadside unit RSU ₁ receives the arrival message AM _i , and the second roadside unit RSU ₂ receives the DM _i : indicating that the pipeline model has normally recorded the entry and exit of the vehicle;

(2) The first roadside unit RSU ₁ does not receive the arrival message AM _i , and the second roadside unit RSU ₂ receives the DM _i : it means that the pipeline model does not record the relevant information of the vehicle, so no processing is done, that is, it is not counted Vehicle types are divided into large, medium, and small vehicle values, which are not included in the weight calculation.

(3) The first roadside unit RSU ₁ receives the arrival message AM _i , but the second roadside unit RSU ₂ does not receive the departure message DM _i : it means that the pipeline model records the relevant information of the vehicle, but it is deleted when the vehicle leaves the pipeline Vehicle information failed. At this time, the obsolete information deletion strategy is adopted, that is, if the departure message DM _i is still not received within the time period λ after receiving the arrival message AM _i , it is considered that the vehicle has left the pipeline, and the relevant information of the vehicle is automatically deleted. Even if the departure message DM _i is received after the λ time period, no processing will be done, that is, it will not be included in the value of the vehicle type divided into three types of large, medium and small vehicles, and will not be included in the calculation of the weight.

(4) The first roadside unit RSU ₁ did not receive the arrival message AM _i , and the second roadside unit RSU ₂ did not receive the departure message DM _i : it means that the pipeline model has no record of the entry and exit of the vehicle, and no processing is done, that is, no Included in the vehicle type is divided into three categories of large, medium and small vehicle values, not brought into the calculation of the weight.

In real life, the traffic signal control system usually assigns a fixed and equal green light time to traffic in all directions at the intersection. However, the traffic flow in different directions is usually not equal and changes dynamically. The fixed green light time cannot adapt to the dynamics of traffic flow, and the distribution of equal green light time cannot meet the needs of different traffic flows in all directions. Therefore, the present invention proposes an on-demand intelligent traffic signal control method based on the pipeline model, and allocates reasonable green light time for traffic flow in each direction under the premise of satisfying the traffic volume.

For a vehicle, the driving quality passing through an intersection is closely related to the stop waiting time and the number of stops. If the waiting time is too long, the traffic volume at the intersection will be reduced; if there are too many stops, the service life of the vehicle will be reduced and the exhaust emission will be increased. Therefore, the goal of a good traffic signal control method is to reduce the average stop waiting time and the average number of stops of the traffic flow as much as possible under the premise of ensuring the traffic volume. The pipeline model records the real-time information of vehicles entering and leaving the pipeline, and assigns a reasonable green light transit time to them based on the traffic flow in the pipeline. When the traffic flow is small, a shorter green light time should be allocated to reduce the waiting time for vehicles to stop; when the traffic flow is large, a longer green light time should be allocated to reduce the number of vehicles stopping.

The allocation of green light time is actually the process of transferring control over green light time. When a road in a certain direction obtains the right to control the green light time, a reasonable allocation of the green light time will be carried out according to the current traffic flow of the road. the way.

Assuming that the weight threshold is Flow_T, the process of allocating green light passing time is shown in Figure 4, and the specific steps are as follows:

Step 1, establish a pipeline model, the pipeline model includes a roadside unit, a data center server and a traffic control system; the roadside unit is used to collect vehicle-related information, and the data center server is used to process roadside The vehicle information submitted by the unit, the traffic control system is used to allocate reasonable green light passing time for each intersection. Go to step 2;

Step 2. When the vehicle enters the pipeline, it sends an arrival message AM _i to the first roadside unit RSU _1. The content of the arrival message AM _i includes the identifier of the vehicle, the driving lane, the type of the vehicle, the time of arrival at the pipeline, and the priority of the vehicle . When the vehicle leaves the pipeline, it sends a departure message DM _i to the second roadside unit RSU ₂ , and the content of the departure message DM _i contains the identifier of the vehicle; after the first roadside unit RSU ₁ receives the arrival message AM _i , the data center server records the vehicle The relevant information of the vehicle; after the second roadside unit RSU ₂ receives the leaving message DM _i , the data center server deletes the relevant information of the vehicle. At the same time, the message retransmission strategy and the obsolete information deletion strategy are used to process the information. Go to step 3;

The message retransmission strategy and outdated information deletion strategy are as follows: the vehicle has a backup when sending the arrival message, and if no response is received from RSU ₁ within time γ, the backup message is sent. Assume that vehicle i sends an arrival message AM _i and a departure message DM _i to the first roadside unit RSU ₁ and the second roadside unit RSU ₂ when entering and leaving the pipeline, respectively. Under the premise of adopting the message retransmission strategy, the result of RSU ₁ and RSU ₂ receiving the message has the following four situations:

(1) The first roadside unit RSU ₁ receives the arrival message AM _i , and the second roadside unit RSU ₂ receives the DM _i : indicating that the pipeline model has normally recorded the entry and exit of the vehicle;

(2) The first roadside unit RSU ₁ does not receive the arrival message AM _i , and the second roadside unit RSU ₂ receives the DM _i : it means that the pipeline model does not record the relevant information of the vehicle, so no processing is done, that is, it is not counted Vehicle types are divided into large, medium, and small vehicle values, which are not included in the weight calculation.

(3) The first roadside unit RSU ₁ receives the arrival message AM _i , but the second roadside unit RSU ₂ does not receive the departure message DM _i : it means that the pipeline model records the relevant information of the vehicle, but it is deleted when the vehicle leaves the pipeline Vehicle information failed. At this time, the obsolete information deletion strategy is adopted, that is, if the departure message DM _i is still not received within the time period λ after receiving the arrival message AM _i , it is considered that the vehicle has left the pipeline, and the relevant information of the vehicle is automatically deleted. Even if the departure message DM _i is received after the λ time period, no processing will be done, that is, it will not be included in the value of the vehicle type divided into three types of large, medium and small vehicles, and will not be included in the calculation of the weight.

(4) The first roadside unit RSU ₁ did not receive the arrival message AM _i , and the second roadside unit RSU ₂ did not receive the departure message DM _i : it means that the pipeline model has no record of the entry and exit of the vehicle, and no processing is done, that is, no Included in the vehicle type is divided into three categories of large, medium and small vehicle values, not brought into the calculation of the weight.

Step 3: Divide the vehicles into three categories: large, medium, and small, and assign influence weights W _x , W _y , and W _z respectively. Small vehicles are the standard impact weights, and the data center server accumulates the weights of various types of vehicles in the pipeline , get the weight value affecting the green light time distribution at the current moment, denoted as Flow_C;

Consider the distribution of green light time for traffic flow in one direction, ignoring the time distribution of right-turn traffic flow, assuming that the total number of vehicles in the current road pipeline is N, and the proportions of left-turning vehicles, straight-going vehicles and right-turning vehicles are N _a , N _b , N _c , let the influence weight of a single vehicle be W _i , then:

$\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; flag</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; W</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

Among them, flag _i represents the identification bit of the i-th vehicle's driving direction, W _i represents the influence weight of the i-th vehicle, and the values of flag _i and W _i are shown in formulas (2) and (3):

The data center server receives the vehicle data processed by the first roadside unit RSU ₁ and the second roadside unit RSU ₂ , obtains the weight value Flow_C that affects the distribution of green light time through formulas (1), (2) and (3), and It is handed over to the traffic control system. Go to step 4;

Step 4, the traffic control system checks whether the lane in the current direction has the right to control the green light time, if yes, go to step 5, otherwise go to step 2;

Step 5, the traffic control system compares the weight value Flow_C and the weight threshold Flow_T of the vehicles in the pipeline that affect the distribution of green light time. If Flow_C>Flow_T, it means that the degree of road congestion is high, then go to step 6, otherwise go to step 8;

Step 6, allocate the green light time to the traffic flow, and continue to compare the weight value Flow_C and the weight threshold Flow_T of the vehicles in the pipeline that affect the green light time allocation. If Flow_C>Flow_T, it means that the degree of road congestion is still at a relatively high level, go to step 7, otherwise go to step 8;

Step 7, the traffic control system judges whether the current green light duration T _G is greater than the longest green light time T _maxG , if yes, go to step 9, otherwise go to step 6;

Step 8, the traffic control system allocates the shortest green light time T _minG to the current lane, and go to step 9;

Step 9, the traffic control system transfers the control right of the green light time of the current lane to the lane in the next direction, and ends the process.

The specific embodiments described in the present invention are only to illustrate the spirit of the present invention. Those skilled in the art to which the present invention belongs can make various modifications or supplements to the described specific embodiments or adopt similar methods to replace them, but they will not deviate from the spirit of the present invention or go beyond the definition of the appended claims range.

Claims (3)

1. an intelligent traffic signal control method based on pipeline model, is characterized in that, comprises the steps: Step 1, establish a pipeline model, the pipeline model includes a roadside unit, a data center server and a traffic control system; the roadside unit is used to collect vehicle-related information, and the data center server is used to process roadside The vehicle information submitted by the unit, the traffic control system is used to allocate reasonable green light passing time for each intersection; Step 2. When the vehicle enters the pipeline, it sends an arrival message AM _i to the first roadside unit RSU _1. The content of the arrival message AM _i includes the identifier of the vehicle, the driving lane, the type of the vehicle, the time of arrival at the pipeline, and the priority of the vehicle , i represents the i-th vehicle; when the vehicle leaves the pipeline, it sends a departure message DM _i to the second roadside unit RSU ₂ , and the content of the departure message DM _i contains the identifier of the vehicle; the first roadside unit RSU ₁ receives the arrival message After AM _i , the data center server records the relevant information of the vehicle; after the second roadside unit RSU ₂ receives the departure message DM _i , the data center server deletes the relevant information of the vehicle; at the same time, adopts the message retransmission strategy and deletes outdated information policy processing information; Step 3: Divide the vehicles into three categories: large, medium, and small, and assign influence weights W _x , W _y , and W _z respectively. Small vehicles are the standard impact weights, and the data center server accumulates the weights of various types of vehicles in the pipeline , get the weight value affecting the distribution of green light time at the current moment, record it as Flow_C, and hand it over to the traffic control system; Step 4, the traffic control system checks whether the lane in the current direction has the right to control the green light time, if yes, go to step 5, otherwise go to step 2; Step 5, the traffic control system compares the weight value Flow_C and the weight threshold Flow_T of the vehicles in the pipeline that affect the distribution of green light time. If Flow_C>Flow_T, it means that the degree of road congestion is high, then go to step 6, otherwise go to step 8; Step 6, allocate green light time for traffic flow, and continue to compare the weight value Flow_C and the weight threshold Flow_T of vehicles in the pipeline that affect the green light time allocation; if Flow_C>Flow_T, the road congestion is still at a high level, go to step 7, Otherwise go to step 8; Step 7, the traffic control system judges whether the current green light duration T _G is greater than the longest green light time T _maxG , if yes, go to step 9, otherwise go to step 6; Step 8, the traffic control system allocates the shortest green light time T _minG to the current lane, and go to step 9; Step 9, the traffic control system transfers the green light time control right of the current lane to the lane in the next direction, and ends the process. 2. A kind of intelligent traffic signal control method based on pipeline model according to claim 1, it is characterized in that: the message retransmission strategy and the obsolete information deletion strategy in the described step 2 are: when the vehicle sends the arrival message, leave There is a backup, if no response is received from the first roadside unit RSU ₁ within time γ, a backup message is sent, assuming that vehicle i sends a message to the first roadside unit RSU ₁ and the second roadside unit respectively when entering and leaving the pipeline The unit RSU ₂ sends the arrival message AM _i and the departure message DM _i , and under the premise of adopting the message retransmission strategy, the result of receiving the message by the first road side unit RSU ₁ and the second road side unit RSU ₂ has the following four situations: (1) The first roadside unit RSU ₁ receives the arrival message AM _i , and the second roadside unit RSU ₂ receives the departure message DM _i : the pipeline model normally records the entry and exit of the vehicle; (2) The first roadside unit RSU ₁ does not receive the arrival message AM _i , and the second roadside unit RSU ₂ receives the departure message DM _i : the pipeline model does not record the relevant information of the vehicle, does not include the vehicle value, and does not include Calculation of input weight; (3) The first roadside unit RSU ₁ receives the arrival message AM _i , but the second roadside unit RSU ₂ does not receive the departure message DM _i : the pipeline model does not record the relevant information of the vehicle, does not include the vehicle value, and does not Bring in the calculation of weight; (4) The first roadside unit RSU ₁ did not receive the arrival message AM _i , and the second roadside unit RSU ₂ did not receive the departure message DM _i : the pipeline model does not record the relevant information of the vehicle, does not count the vehicle value, and does not into the calculation of weights. 3. A kind of intelligent traffic signal control method based on pipeline model according to claim 1, it is characterized in that, in described step 3, the method for calculating the weight value Flow_C that finally affects the distribution of green light time is: consider a direction traffic flow The distribution of green light time, ignoring the time distribution of right-turning traffic flow, assuming that the total number of vehicles in the current road pipeline is N, and the proportions of left-turning vehicles, straight-going vehicles and right-turning vehicles are respectively N _a , N _b , N _c , so that a single The influence weight of the vehicle is W _i , then: $\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; flag</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; W</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$ Among them, flag _i represents the identification bit of the i-th vehicle's driving direction, W _i represents the influence weight of the i-th vehicle, and the values of flag _i and W _i are shown in formulas (2) and (3): The data center server receives the vehicle data processed by the first roadside unit RSU ₁ and the second roadside unit RSU ₂ , obtains the weight value Flow_C that affects the distribution of green light time through formulas (1), (2) and (3), and It is handed over to the traffic control system.