Self-adaptive maximum pressure signal control method suitable for port traffic

Abstract

The invention discloses a self-adaptive maximum pressure signal control method suitable for port traffic, which is characterized in that a video detector is arranged in each entrance direction of each intersection needing traffic signal control, and a data processor in which a maximum pressure model is pre-stored is arranged at each intersection; a data processor calculates signal lamp control logic of an intersection on the basis of a maximum pressure model according to the queue length and the road occupancy rate of each entrance direction of the intersection and the queue length and the road occupancy rate of the corresponding entrance direction of each intersection adjacent to the intersection, and pressure values of the intersection in each entrance direction are obtained. Determining the direction of the next released entrance of the intersection; the method has the advantages that the method can be suitable for the port complex traffic environment, in the port complex traffic environment, the control process is simple, the control timeliness is high, the signal timing accuracy of the intersection signal lamps is high, and efficient and accurate port traffic management can be achieved.

Description

Technical Field

[0001] The invention relates to a traffic control method, in particular to an adaptive maximum pressure signal control method suitable for port traffic. Background Art

[0002] With the advancement of globalization and the continuous expansion of international trade, the port collection and distribution network, as an important hub connecting ports and inland hinterlands, has a direct impact on the sustainable development of the regional economy and the stability of the international supply chain. However, compared with traditional urban traffic flow, the flow characteristics of port traffic have significant heterogeneity and flow volatility, especially during the peak period of port operations, the flow fluctuation is more severe, which poses a severe challenge to the existing traffic signal control methods.

[0003] Traditional traffic signal control methods are usually based on the assumption of homogeneous traffic flow, that is, it is assumed that all vehicles occupy the same amount of traffic resources. This assumption cannot accurately reflect the complex situation in port traffic. Especially when heavy vehicles account for a high proportion, traditional traffic signal control methods often fail to fully consider the impact of heavy vehicles on the efficiency of intersection traffic, resulting in unreasonable signal timing, which in turn causes low traffic efficiency, traffic congestion and vehicle delays.

[0004] Traditional traffic signal control methods mainly include traffic signal control methods based on fixed timing, traffic signal control methods based on induction control, and traffic signal control methods based on adaptive control.

[0005] The traffic signal control method based on fixed timing controls the intersection lights through a preset periodic signal timing scheme. This method is simple to operate and has high timeliness. However, due to its fixed signal timing scheme, it is difficult to adjust according to the dynamic changes of real-time traffic flow, resulting in limited control accuracy, especially difficult to effectively cope with complex traffic characteristics such as large fluctuations in port traffic flow and a high proportion of heavy vehicles.

[0006] The traffic signal control method based on induction control uses induction equipment to detect vehicle arrival and dynamically adjusts the signal timing to achieve the control of intersection lights. Although this method has the characteristics of flexible response and high control efficiency, its control range is usually limited to a single intersection and lacks the ability to coordinate the overall complex traffic network of the port.

[0007] Traffic signal control methods based on adaptive control (such as SCOOT, SCATS, etc.) achieve optimal control of intersection lights by collecting traffic data in real time and dynamically adjusting signal timing. This type of method usually performs well in urban roads, but because it mostly adopts a centralized architecture, it has high requirements for data transmission and processing capabilities, resulting in slow control response speed and difficulty in meeting real-time control requirements. At the same time, this method has limitations when dealing with port traffic environments, such as being unable to effectively adapt to complex characteristics such as large traffic fluctuations and a high proportion of heavy vehicles.

[0008] In addition, most of the existing traditional traffic signal control methods are based on the assumption of homogeneous traffic flow, that is, it is assumed that all vehicles occupy the same amount of traffic resources, and the differences between heavy vehicles and light vehicles are not fully considered. However, in port traffic, heavy vehicles (such as container trucks) account for a high proportion. Due to their large size, slow speed, and low acceleration, these vehicles occupy more road resources than ordinary vehicles. In the context of increasingly heterogeneous port traffic flows, especially when the proportion of heavy vehicles is high, the above-mentioned traditional signal control methods are difficult to fully consider the actual impact of heavy vehicles on intersection traffic efficiency and cannot accurately reflect traffic pressure. This limitation leads to inaccurate signal timing, which in turn limits the capacity and operating efficiency of the port's collection and distribution network. Summary of the invention

[0009] The technical problem to be solved by the present invention is to provide an adaptive maximum pressure signal control method suitable for port traffic, which can be applicable to the complex traffic environment of the port. In the complex traffic environment of the port, not only the control process is simple and the control timeliness is high, but also the signal timing accuracy of the intersection traffic lights is high, and efficient and accurate port traffic management can be achieved.

[0010] The technical solution adopted by the present invention to solve the above technical problems is: an adaptive maximum pressure signal control method suitable for port traffic, by setting a video detector at each entrance direction of each intersection where traffic signal control is required in the port collection and distribution network, and setting a data processor with a pre-stored maximum pressure model at each intersection, the data processor at each intersection is respectively connected to the video detector set at each entrance direction of the intersection, the data processor of the adjacent intersection and the signal light controller of the intersection; the video detector at each entrance direction of each intersection collects the traffic data of the entrance direction in real time according to the preset time interval, and transmits the collected traffic data to the data processor at the intersection, and the data processor at the intersection calculates the intersection according to the received traffic data. The queue length and road occupancy rate of each entrance direction are calculated, and the calculation results are transmitted to the data processor at the intersection adjacent to the intersection; when controlling a certain intersection, the data processor at the intersection calculates the signal light control logic of the intersection based on the maximum pressure model according to the queue length and road occupancy rate of each entrance direction of the intersection, as well as the queue length and road occupancy rate of the corresponding entrance direction of each intersection adjacent to the intersection, obtains the pressure value of each entrance direction of the intersection, and then determines the next entrance direction to be released at the intersection, and outputs the corresponding release instruction to the signal light controller of the intersection. After receiving the release instruction, the signal light controller switches the signal light display state of the intersection to be consistent with the release instruction, thereby realizing periodic dynamic control of the signal light at the intersection.

[0011] Compared with the prior art, the advantages of the present invention are that, firstly, by setting up video detectors to monitor the traffic conditions of each intersection in real time, more accurate traffic data can be provided, so that the control of traffic lights is more in line with actual traffic needs, and can effectively adapt to the complexity of port traffic, taking into account the variable traffic flow and special needs in the port environment, and can adaptively adjust the signal control strategy according to real-time traffic data, avoiding the limitations of traditional fixed cycles or simple logic control methods; secondly, through the maximum pressure model to perform logical calculations of signal control, the traffic pressure in the entrance direction of each intersection can be accurately evaluated, ensuring that the timing of traffic lights is more reasonable and efficient, avoiding the blindness and inefficiency of signal control, and can dynamically adjust the signal timing according to the real-time traffic pressure of each intersection. , achieving higher control timeliness, which is particularly suitable for high-density and complex traffic environments such as ports; the signal control of each intersection is calculated based on the queue length and road occupancy rate information collected in real time. The control process is simple, and the data transmission and processing flow are efficient, which not only improves the timeliness of control, but also greatly reduces the delay problem existing in traditional traffic signal control. Moreover, due to the adaptability of the calculation method, it can be quickly adjusted when different traffic flows and emergencies occur, ensuring the smoothness and efficiency of port traffic; therefore, the present invention can be applied to complex traffic environments in ports. In complex traffic environments in ports, not only the control process is simple, the control timeliness is high, and the signal timing accuracy of intersection traffic lights is high, which can achieve efficient and accurate port traffic management.

[0012] Furthermore, the specific method of setting up video detectors at each entrance direction of each intersection that needs traffic signal control in the port collection and distribution network is as follows: a video detector is installed above or beside the road in each entrance direction of each intersection that needs to be controlled in the port collection and distribution network, and the shooting area of ​​each video detector can cover all vehicles in all lanes in its corresponding entrance direction extending from the stop line to the upstream within a range of 200 meters, and the video detectors set at each entrance direction of each intersection are connected to the data processor set at the intersection, and each video detector is used to collect traffic flow images in its coverage area in real time through a high-definition camera, and each time a traffic flow image is collected, the image processing technology is first used to identify the traffic flow image, and the vehicle model, speed and position of all vehicles in the traffic flow image are obtained, and then the state of the vehicle is judged according to the speed. If the vehicle speed is less than 1m / s, the vehicle is in a parked state, otherwise The vehicle is in a driving state, and the vehicle length is obtained according to the model of each vehicle and a unique identification code is generated for the vehicle. Finally, the identification code, position, length and state of each vehicle are used to form the traffic data of the vehicle, and the obtained traffic data of all vehicles are transmitted to the data processor connected to it; the acquisition frequency of the video detectors at all entrance directions of each intersection is the same, and the time of acquisition is the same, that is, the video detectors at all entrance directions of each intersection synchronously transmit the obtained traffic data of all vehicles to the data processor connected to it; when the data processor of a certain intersection receives the traffic data transmitted by all video detectors connected to it, the data processor of the intersection processes the received traffic data to obtain the queue length and road occupancy rate of all entrance directions of the intersection, and transmits it to the data processor set at the intersection adjacent to the intersection, and the data processor set at each intersection is preset with a data update time interval T s , that is, the data update period, T s <15s, the starting time of data processing for data processors at all intersections is the same, which is recorded as data update 0 time. The data processor at each intersection enters the first data update cycle at data update 0 time, and then every T s , enter the next data update cycle, and the data processor at each intersection updates the data once each time it enters a data update cycle; a certain intersection is recorded as intersection n, and the number of entrance directions at intersection n is recorded as F n , the lth entrance direction at intersection n is called entrance direction l, l = 1, 2, ..., F n , the data update cycle currently entered by the data processor at intersection n is recorded as the tth data update cycle, t = 1, 2, 3, ...; the specific process of the data processor at intersection n performing data update in the tth data update cycle is:

[0013] S1, the data processor at the intersection n determines whether there is a vehicle in each entrance direction of the intersection n at this time according to the traffic data of the tth data update cycle transmitted by the video detector connected to it; if there is no traffic data of the vehicle in the traffic data transmitted to the data processor at the intersection n by the video detector of a certain entrance direction at the intersection n in the tth data update cycle, then there is no vehicle in the entrance direction at the intersection n in the tth data update cycle, that is, the number of vehicles is 0; otherwise, there is a vehicle in the entrance direction at the intersection n in the tth data update cycle;

[0014] S2. The number of vehicles at the entrance direction l of intersection n in the tth data update cycle is recorded as N l,n (t), the queue length of the entrance direction l of intersection n in the tth data update cycle is recorded as x l,n (t), the road occupancy rate of the entrance direction l of the intersection n in the tth data update period is recorded as θ l,n (t), according to N l,n (t) Get x l,n (t) and θ l,n (t), specifically:

[0015] If there is no vehicle in the direction of the lth entrance at intersection n in the tth data update cycle, that is, N l,n (t) = 0, then let x l,n (t) = 0, θ l,n (t) = 0;

[0016] If there is a vehicle at the lth entrance direction at intersection n in the tth data update cycle, the data processor obtains the number of vehicles N according to the identification codes of all vehicles. l,n (t), each vehicle identification code represents a vehicle, and then, according to formulas (1) and (2), we can calculate x l,n (t) and θ l,n (t):

[0017]

[0018]

[0019] Among them, i l,n (t) = 1, 2, ..., N l,n (t); Indicates the i-th intersection n in the direction l of the t-th data update cycle l,n (t) The parking status of the vehicle. If the vehicle is currently parked, otherwise, Indicates the i-th intersection n in the direction l of the t-th data update cycle l,n (t) The length of the vehicle, λ l,nIndicates the number of lanes in the entrance direction l of intersection n;

[0020] S3, the data processor set at intersection n first uses the queue length x of the entrance direction l of intersection n in the tth data update cycle l,n (t) Update the queue length of the intersection n entrance direction l, using the road occupancy rate θ of the intersection n entrance direction l in the tth data update cycle l,n (t) updating the road occupancy rate of the entrance direction l of the intersection n at the location thereof, and then transmitting the queue length and road occupancy rate of the entrance direction l of the intersection n at the location thereof to the data processor of the adjacent intersection connected thereto, and receiving the queue length and occupancy rate data of each entrance direction of the adjacent intersection transmitted by the data processor of the adjacent intersection, and using the received queue length and occupancy rate data of each entrance direction of the adjacent intersection to correspondingly update the queue length and occupancy rate data of each entrance direction of the adjacent intersection;

[0021] Among them, at the data update time 0, the data processor set at intersection n will initialize the number of vehicles, queue length and road occupancy rate of all entrance directions of intersection n, so that the number of vehicles, queue length and road occupancy rate are all 0.

[0022] Furthermore, a time interval of 15 seconds is preset at the data processors set at all intersections that need to perform signal control. The time interval is the length of the signal control cycle. The starting time for controlling all intersections is recorded as the control time 0, where the control time 0 is the same as the data update time 0. The data processor of each intersection enters the first signal control cycle 15 seconds after the control time 0, and then enters the next signal control cycle every 15 seconds. The data processor at each intersection performs signal control once each time it enters a signal control cycle; the number of exit directions corresponding to the entrance direction l of the intersection n is recorded as O. l,n The mth exit direction corresponding to the entrance direction l of intersection n is called the exit direction m of the entrance direction l of intersection n. l,n , m l,n =1,2,…,O l,n ; The signal control cycle currently entered by the data processor at intersection n is recorded as the kth signal control cycle, k = 1, 2, 3, ..., the specific process of the signal control of the data processor at intersection n in the kth signal control cycle is:

[0023] Step 1: The data processor set at intersection n calculates the weight of each entrance direction of intersection n according to the queue length and road occupancy rate of each entrance direction of intersection n and its adjacent signalized intersections pre-stored in the data processor, and the process is as follows:

[0024] S1.1. The average road occupancy rate of intersection n in the kth signal control cycle is recorded as Using formula (3) to calculate

[0025]

[0026] Among them, θ l ' ,n (k) is the latest data of the road occupancy rate of the entrance direction l of the intersection n currently stored in the data processor set at the intersection n;

[0027] S1.2 If Then, the weight of the entrance direction l of the intersection n in the kth signal control cycle is calculated according to formula (4):

[0028]

[0029] Among them, x l ' ,n (k) represents the latest data of the queue length in the entrance direction l of intersection n currently stored in the data processor of intersection n, Indicates that the vehicle is driving from the entrance direction l of intersection n to the m'th direction corresponding to the entrance direction l. l,n The probability of an exit direction, m' l,n =1,2,…,O l,n , O l,n is the number of exit directions corresponding to the entrance direction l of intersection n; Indicates the m'th corresponding to the entrance direction l of the current intersection n l,n The length of the queue in the exit direction is l,n When there is no corresponding intersection requiring traffic signal control in the exit direction, it is equal to 0. l,n When there is a corresponding intersection that needs traffic signal control for the exit direction, it is equal to the m'th entry direction l of intersection n currently saved in the data processor of intersection n. l,n The latest data of the queue length in the entrance direction of the adjacent intersection corresponding to the exit direction; Indicates the m'th corresponding to the entrance direction l of the current intersection n l,n The road occupancy rate of the exit direction is the m'th l,n When there is no corresponding intersection requiring traffic signal control in the exit direction, it is equal to 0. l,nWhen there is a corresponding intersection that needs traffic signal control for the exit direction, it is equal to the m'th entry direction l of intersection n currently saved in the data processor of intersection n. l,n The latest data on the road occupancy rate of the entrance direction of the adjacent intersection corresponding to the exit direction;

[0030] like Then, the weight of the entrance direction l of the intersection n in the kth signal control cycle is calculated according to formula (5):

[0031]

[0032] Step 2: The pressure of the entrance direction l of the intersection n in the kth signal control period is recorded as p l,n (k), and p is calculated according to formula (6) l,n (k):

[0033] p l,n (k) = w l,n (k)×1800 (6)

[0034] Step 3: Get p 1,n (k) to The maximum value among the values ​​is used as the entrance direction of intersection n corresponding to the maximum value, which is recorded as the next entrance direction to be released, and is denoted as entrance direction S n * (k), and generates a corresponding release instruction to transmit to the signal light controller at the intersection n connected to it. When the signal light controller at the intersection n receives the release instruction, if S n * (k) The direction is consistent with the entrance currently being released,

[0035] Then the signal light controller set at intersection n is S at the entrance direction of intersection n. n * (k) Set the green light and the red light for other entrance directions, and re-record the entrance direction S n * (k) Green light duration;

[0036] If S n * (k) If the direction of the entrance is not consistent with the direction of the current entrance, the signal controller at intersection n will first set the yellow light for the direction of the entrance currently being released at intersection n, and the signal lights for other entrance directions will remain unchanged. After 3 seconds, the signal lights for the entrance direction S will be set to yellow. n * (k) Set the green light and set the red light for other entrance directions, and re-record the entrance direction S n * (k) The green light duration. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a flow chart of the adaptive maximum pressure signal control method applicable to port traffic of the present invention;

[0038] Figure 2 A schematic diagram of a simulated road network structure when simulating the adaptive maximum pressure signal control method applicable to port traffic of the present invention;

[0039] Figure 3 A comparison diagram of average vehicle delays when simulating the adaptive maximum pressure signal control method for port traffic of the present invention;

[0040] Figure 4 A total delay comparison diagram when simulating the adaptive maximum pressure signal control method for port traffic of the present invention;

[0041] Figure 5 A comparison diagram of the cumulative total delay when simulating the adaptive maximum pressure signal control method for port traffic of the present invention;

[0042] Figure 6 A comparison chart of the number of vehicles on a road network when simulating the adaptive maximum pressure signal control method for port traffic of the present invention. DETAILED DESCRIPTION

[0043] The present invention is further described in detail below with reference to the accompanying drawings.

[0044] Embodiment 1: An adaptive maximum pressure signal control method suitable for port traffic, characterized in that a video detector is set at each entrance direction of each intersection where traffic signal control is required in the port collection and distribution network, and a data processor with a pre-stored maximum pressure model is set at each intersection, and the data processor at each intersection is respectively connected to the video detectors set at each entrance direction of the intersection, the data processors of adjacent intersections and the signal light controller of the intersection; the video detectors at each entrance direction of each intersection collect traffic data of the entrance direction in real time according to a preset time interval, and transmit the collected traffic data to the data processor at the intersection, and the data processor at the intersection calculates the maximum pressure model of each entrance direction of the intersection according to the received traffic data. The data processor at the intersection calculates the queue length and road occupancy rate in each entrance direction of the intersection and transmits the calculation result to the data processor at the intersection adjacent to the intersection; when controlling a certain intersection, the data processor at the intersection calculates the signal light control logic of the intersection based on the maximum pressure model according to the queue length and road occupancy rate in each entrance direction of the intersection, as well as the queue length and road occupancy rate in the corresponding entrance direction of each intersection adjacent to the intersection, obtains the pressure value of each entrance direction of the intersection, and then determines the next entrance direction to be released at the intersection, and outputs the corresponding release instruction to the signal light controller of the intersection. After receiving the release instruction, the signal light controller switches the display state of the signal light of the intersection to be consistent with the release instruction, thereby realizing periodic dynamic control of the signal light of the intersection.

[0045] In this embodiment, by setting a video detector and a data processor with a pre-stored maximum pressure model at the entrance direction of each intersection in the port traffic network, the queue length and road occupancy rate of each entrance direction can be monitored in real time and accurately calculated, thereby providing accurate traffic data support for signal control; at the same time, the calculation of the signal light control logic based on the maximum pressure model can comprehensively consider the traffic pressure of each intersection and adjacent intersections, ensure the optimization of the release order of the signal lights, and avoid the congestion and delay problems that may occur in traditional signal control; in addition, the interconnection between data processors makes information transmission more efficient and real-time, reduces the common delays and coordination problems in traditional signal control methods, and further improves the management efficiency of port traffic. Therefore, compared with the traditional signal control method based on fixed cycles or empirical rules, the present invention can respond to the dynamic changes of port traffic flow more flexibly and accurately, and improve the accuracy of signal timing and the timeliness of control.

[0046] Embodiment 2: This embodiment is basically the same as Embodiment 1, except that: in this embodiment, Figure 1As shown, the specific method of setting up video detectors at each entrance direction of each intersection that needs to be controlled by traffic signals in the port collection and distribution network is as follows: a video detector is installed above or beside the road in each entrance direction of each intersection that needs to be controlled in the port collection and distribution network, and the shooting area of ​​each video detector can cover all vehicles within a range of 200 meters extending upstream from the stop line in all lanes of its corresponding entrance direction. The video detectors set at each entrance direction of each intersection are connected to the data processor set at the intersection. Each video detector is used to collect traffic flow images in its coverage area in real time through a high-definition camera, and each time a traffic flow image is collected, the image processing technology is first used to identify the traffic flow image, and the vehicle model, speed and position of all vehicles in the traffic flow image are obtained, and then the state of the vehicle is judged according to the speed. If the vehicle speed is less than 1m / s, the vehicle is in a parked state, otherwise The video detectors at all entrance directions of each intersection have the same acquisition frequency and the same acquisition time, that is, the video detectors at all entrance directions of each intersection synchronously transmit the traffic data of all vehicles obtained to the data processor connected to it; when the data processor at a certain intersection receives the traffic data transmitted by all video detectors connected to it, the data processor at the intersection processes the received traffic data to obtain the queue length and road occupancy rate of all entrance directions of the intersection, and transmits them to the data processors set at the intersections adjacent to the intersection, and the data processors set at each intersection are preset with a data update time interval T s , that is, the data update period, T s <15s, the starting time of data processing for data processors at all intersections is the same, which is recorded as data update 0 time. The data processor at each intersection enters the first data update cycle at data update 0 time, and then every T s , enter the next data update cycle, and the data processor at each intersection updates the data once each time it enters a data update cycle; a certain intersection is recorded as intersection n, and the number of entrance directions at intersection n is recorded as F n , the lth entrance direction at intersection n is called entrance direction l, l = 1, 2, ..., F n , the data update cycle currently entered by the data processor at intersection n is recorded as the tth data update cycle, t = 1, 2, 3, ...; the specific process of the data processor at intersection n performing data update in the tth data update cycle is:

[0047] S1, the data processor at the intersection n determines whether there is a vehicle in each entrance direction of the intersection n at this time according to the traffic data of the tth data update cycle transmitted by the video detector connected to it; if there is no traffic data of the vehicle in the traffic data transmitted to the data processor at the intersection n by the video detector of a certain entrance direction at the intersection n in the tth data update cycle, then there is no vehicle in the entrance direction at the intersection n in the tth data update cycle, that is, the number of vehicles is 0; otherwise, there is a vehicle in the entrance direction at the intersection n in the tth data update cycle;

[0048] S2. The number of vehicles at the entrance direction l of intersection n in the tth data update cycle is recorded as N l,n (t), the queue length of the entrance direction l of intersection n in the tth data update cycle is recorded as x l,n (t), the road occupancy rate of the entrance direction l of the intersection n in the tth data update period is recorded as θ l,n (t), according to N l,n (t) Get x l,n (t) and θ l,n (t), specifically:

[0049] If there is no vehicle in the direction of the lth entrance at intersection n in the tth data update cycle, that is, N l,n (t) = 0, then let x l,n (t) = 0, θ l,n (t) = 0;

[0050] If there is a vehicle at the lth entrance direction at intersection n in the tth data update cycle, the data processor obtains the number of vehicles N according to the identification codes of all vehicles. l,n (t), each vehicle identification code represents a vehicle, and then, according to formulas (1) and (2), we can calculate x l,n (t) and θ l,n (t):

[0051]

[0052] Among them, i l,n (t) = 1, 2, ..., N l,n (t); Indicates the i-th intersection n in the direction l of the t-th data update cycle l,n (t) The parking status of the vehicle. If the vehicle is currently parked, otherwise, Indicates the i-th intersection n in the direction l of the t-th data update cycle l,n (t) The length of the vehicle, λ l,n Indicates the number of lanes in the entrance direction l of intersection n;

[0053] S3, the data processor set at intersection n first uses the queue length x of the entrance direction l of intersection n in the tth data update cycle l,n (t) Update the queue length of the intersection n entrance direction l, using the road occupancy rate θ of the intersection n entrance direction l in the tth data update cycle l,n (t) updating the road occupancy rate of the entrance direction l of the intersection n at the location thereof, and then transmitting the queue length and road occupancy rate of the entrance direction l of the intersection n at the location thereof to the data processor of the adjacent intersection connected thereto, and receiving the queue length and occupancy rate data of each entrance direction of the adjacent intersection transmitted by the data processor of the adjacent intersection, and using the received queue length and occupancy rate data of each entrance direction of the adjacent intersection to correspondingly update the queue length and occupancy rate data of each entrance direction of the adjacent intersection;

[0054] Among them, at the data update time 0, the data processor set at intersection n will initialize the number of vehicles, queue length and road occupancy rate of all entrance directions of intersection n, so that the number of vehicles, queue length and road occupancy rate are all 0.

[0055] In this embodiment, high-definition video detectors are arranged at each entrance direction of the intersection to cover all vehicles within a 200-meter range. The vehicle identification code, location, length, speed and status can be collected in real time to accurately obtain the vehicle queue length and road occupancy rate data. The data processor updates the data according to the preset update cycle T s The collected data is processed to ensure that the traffic status information is highly real-time and accurate. At the same time, the data processor of each intersection not only processes the traffic data of the intersection, but also realizes distributed collaborative processing through data interaction with adjacent intersections to ensure the global coordination of signal control. The traffic queue length and road occupancy rate of each entrance direction are calculated by formulas (1) and (2), providing reliable input data for the subsequent maximum pressure model, which fully reflects the dynamic adaptability and efficiency of the control process.

[0056] In this embodiment, a time interval of 15 seconds is preset at the data processors set at all intersections that need to perform signal control. The time interval is the length of the signal control cycle. The starting time of controlling all intersections is recorded as control time 0, where control time 0 is the same as data update time 0. The data processor of each intersection enters the first signal control cycle 15 seconds after control time 0, and then enters the next signal control cycle every 15 seconds. The data processor at each intersection performs signal control once each time it enters a signal control cycle; the number of exit directions corresponding to the entrance direction l of intersection n is recorded as O. l,n The mth exit direction corresponding to the entrance direction l of intersection n is called the exit direction m of the entrance direction l of intersection n.l,n , m l,n =1,2,…,O l,n ; The signal control cycle currently entered by the data processor at intersection n is recorded as the kth signal control cycle, k = 1, 2, 3, ..., the specific process of the signal control of the data processor at intersection n in the kth signal control cycle is:

[0057] Step 1: The data processor set at intersection n calculates the weight of each entrance direction of intersection n according to the queue length and road occupancy rate of each entrance direction of intersection n and its adjacent signalized intersections pre-stored in the data processor, and the process is as follows:

[0058] S1.1. The average road occupancy rate of intersection n in the kth signal control cycle is recorded as Using formula (3) to calculate

[0059]

[0060] Among them, θ l ' ,n (k) is the latest data of the road occupancy rate of the entrance direction l of the intersection n currently stored in the data processor set at the intersection n;

[0061] S1.2 If Then, the weight of the entrance direction l of the intersection n in the kth signal control cycle is calculated according to formula (4):

[0062]

[0063] Among them, x l ' ,n (k) represents the latest data of the queue length in the entrance direction l of intersection n currently stored in the data processor of intersection n, Indicates that the vehicle is driving from the entrance direction l of intersection n to the m'th direction corresponding to the entrance direction l. l,n The probability of an exit direction, m' l,n =1,2,…,O l,n , O l,n is the number of exit directions corresponding to the entrance direction l of intersection n; Indicates the m'th corresponding to the entrance direction l of the current intersection n l,n The length of the queue in the exit direction is l,n When there is no corresponding intersection requiring traffic signal control in the exit direction, it is equal to 0. l,nWhen there is a corresponding intersection that needs traffic signal control for the exit direction, it is equal to the m'th entry direction l of intersection n currently saved in the data processor of intersection n. l,n The latest data of the queue length in the entrance direction of the adjacent intersection corresponding to the exit direction; Indicates the m'th corresponding to the entrance direction l of the current intersection n l,n The road occupancy rate of the exit direction is the m'th l,n When there is no corresponding intersection requiring traffic signal control in the exit direction, it is equal to 0. l,n When there is a corresponding intersection that needs traffic signal control for the exit direction, it is equal to the m'th entry direction l of intersection n currently saved in the data processor of intersection n. l,n The latest data on the road occupancy rate of the entrance direction of the adjacent intersection corresponding to the exit direction;

[0064] like Then, the weight of the entrance direction l of the intersection n in the kth signal control cycle is calculated according to formula (5):

[0065]

[0066] Step 2: The pressure of the entrance direction l of the intersection n in the kth signal control period is recorded as p l,n (k), and p is calculated according to formula (6) l,n (k):

[0067] p l,n (k) = w l,n (k)×1800 (6)

[0068] Step 3: Get p 1,n (k) to The maximum value among the values ​​is used as the entrance direction of intersection n corresponding to the maximum value, which is recorded as the next entrance direction to be released, and is denoted as entrance direction S n * (k), and generates a corresponding release instruction to transmit to the signal light controller at the intersection n connected to it. When the signal light controller at the intersection n receives the release instruction, if S n * (k) The direction is consistent with the entrance currently being released,

[0069] Then the signal light controller set at intersection n is S at the entrance direction of intersection n. n * (k) Set the green light and the red light for other entrance directions, and re-record the entrance direction S n* (k) Green light duration;

[0070] If S n * (k) If the direction of the entrance is not consistent with the direction of the current entrance, the signal controller at intersection n will first set the yellow light for the direction of the entrance currently being released at intersection n, and the signal lights for other entrance directions will remain unchanged. After 3 seconds, the signal lights for the entrance direction S will be set to yellow. n * (k) Set the green light and set the red light for other entrance directions, and re-record the entrance direction S n * (k) The green light duration.

[0071] In this embodiment, the focus is on the control logic of the traffic light. Based on the signal control cycle, the weight w of each entrance direction of the intersection is calculated. l,n (k) and pressure value p l,n (k), dynamically evaluate the traffic conditions of each entrance direction of the intersection, and finally determine the entrance direction with priority. First, by calculating the average road occupancy rate θ l,n (k), combined with the set threshold, judge the occupancy of each entrance direction of the intersection, and calculate the corresponding weight value according to the occupancy rate to optimize the traffic signal control strategy. Secondly, the maximum pressure model is used to calculate the pressure value p of each entrance direction of the intersection l,n (k), and select the direction with the largest pressure value as the priority direction for current release. Finally, release control is achieved through the green, yellow and red light switching mechanism of the signal light, and the duration of the green light for release is recorded to ensure the stability of traffic flow and the accuracy of control. Finally, efficient, scientific and accurate port traffic signal control is achieved.

[0072] In order to verify the performance of the adaptive maximum pressure signal control method for port traffic of the present invention, an experimental verification was carried out in the SUMO simulation platform, and real-time data interaction was achieved through Python and SUMO API interfaces. The hardware equipment of the simulation environment is Intel Core i5-9300H CPU@2.40GHz, RAM is 8GB, the operating system is 64-bit Windows10, Python version is 3.10.13, and SUMO version is 1.20.0. The simulated road network structure is as follows Figure 2 As shown, Figure 2 The 4×4 grid-type traffic network shown in the middle left figure contains 16 intersections. Each intersection in the road network ( Figure 2As shown in the right figure, there are four entrance directions, each of which has three lanes, and the length of each lane is 200m. The traffic lights at each intersection can be dynamically controlled. The turning probabilities of vehicles entering from each entrance direction of each intersection are set as: 50% for straight, 30% for left turn, and 20% for right turn. The traffic flow is generated in the way that the north-south entrance of the road network is only composed of container trucks, and the east-west entrance of the road network is only composed of cars. The probabilities of straight, left turn, and right turn of all vehicles are equal, which are 0.5, 0.3, and 0.2 respectively. Container trucks and cars enter the road from the north-south and east-west directions of the road network respectively, and then they are randomly routed according to the turning probability. The maximum speed of all roads is set to 20m / s, the length of the container truck is 16m, the length of the car is 5m, the maximum speed of both vehicles is 20m / s, and the maximum acceleration of the container truck is 1.3m / s 2 and the maximum deceleration is 2.5m / s 2 , the maximum acceleration of the car is 2.6m / s 2 and the maximum deceleration is 4.5m / s 2 . All vehicles use the default Krauss vehicle following model. According to the characteristics of port traffic, five stages of traffic flow scenarios are set, covering three demand levels: low demand, medium demand and high demand. Among them, the north-south flow and east-west flow are set as: 300veh / h and 450veh / h in the first stage, 400veh / h and 600veh / h in the second stage, 500veh / h and 750veh / h in the third stage, 400veh / h and 600veh / h in the fourth stage, and 300veh / h and 450veh / h in the fifth stage.

[0073] The experiment uses Python and SUMO's Traci interface to realize data interaction, and calls SUMO's simulation data in real time in the Python environment to dynamically control the traffic lights. The signal control logic of the traditional Webster method is set to collect data every 5 minutes and update the signal timing as a comparison group. The adaptive maximum pressure signal control method suitable for port traffic of the present invention collects traffic data once a second, calculates and adjusts the signal timing in real time based on the maximum pressure model. During the experiment, SUMO generates traffic flow in real time and simulates vehicle traffic. Python obtains traffic data of each entrance direction of each intersection through the Traci interface, including parameters such as queue length, road occupancy rate and current number of vehicles. At the same time, the traffic pressure value of each entrance direction is calculated according to the adaptive maximum pressure signal control method suitable for port traffic of the present invention, and the traffic light control strategy is updated in real time.

[0074] In order to verify the reliability and scientificity of the experimental results, each method was independently run for 10 simulations in each traffic stage to eliminate the influence of random factors, and finally the average value was taken for comparative analysis. The experimental data mainly includes three evaluation indicators: the number of vehicles in the road network, the total delay and the average vehicle delay. The number of vehicles in the road network reflects the number of vehicles in the current network when the demand remains unchanged. The total delay represents the total waiting time of all vehicles on the road network, and the average vehicle delay measures the average waiting time of a single vehicle in the network. The three together characterize the overall performance of the signal control method. The specific experimental comparison results are shown in Table 1:

[0075] Table 1 Experimental comparison results of port traffic signal control methods

[0076]

[0077] The experimental results in Table 1 show that in all traffic stages, the adaptive maximum pressure signal control method for port traffic of the present invention is superior to the traditional Webster method, especially in the medium demand and high demand stages. In the low demand stage (stage 1), the adaptive maximum pressure signal control method for port traffic of the present invention can reduce the number of vehicles on the road network from 336.29 vehicles of the Webster method to 286.50 vehicles, the total delay from 35308.28 seconds to 23332.21 seconds, and the average vehicle delay from 104.83 seconds / vehicle to 81.34 seconds / vehicle, showing a stable traffic diversion capability. With the increase of traffic flow, in the medium demand stage (stage 2), the method of the present invention continues to maintain its advantages, with a reduction of 12.8% in the number of vehicles on the road network, a reduction of 30.3% in the total delay, and a reduction of 20.2% in the average vehicle delay.

[0078] In the high demand stage (Stage 3), the performance of the Webster method shows a significant decline, the number of vehicles on the road network increases to 812.65, the total delay and the average vehicle delay reach 149459.35 seconds and 176.08 seconds per vehicle, respectively, showing a large congestion accumulation problem. However, the adaptive maximum pressure signal control method for port traffic of the present invention has only 653.20 vehicles on the road network at this stage, a total delay of 84569.00 seconds, and an average vehicle delay of 126.91 seconds per vehicle, which are reduced by about 19.6%, 43.4% and 27.9% respectively compared with the Webster method, indicating its strong adaptability and significant optimization effect in high demand scenarios.

[0079] In the extremely high demand stage (Stage 4), the advantages of the adaptive maximum pressure signal control method for port traffic of the present invention are more prominent. The number of vehicles on the road network of the Webster method climbed to 878.84, and the total delay and average vehicle delay were as high as 274359.05 seconds and 300.25 seconds per vehicle, respectively. The adaptive maximum pressure signal control method for port traffic of the present invention can control the number of vehicles on the road network to 506.02, and the total delay and average vehicle delay are reduced to 63559.56 seconds and 121.87 seconds per vehicle, respectively. The optimization ranges of the three indicators are 42.4%, 76.8% and 59.4%, respectively. In the fifth stage when the traffic demand gradually decreases, the adaptive maximum pressure signal control method for port traffic of the present invention also performs well, and the number of vehicles on the road network, the total delay and the average vehicle delay are significantly lower than those of the Webster method, showing good recovery ability.

[0080] Combined with the comparison chart for further analysis, Figure 3 It is shown that the adaptive maximum pressure signal control method for port traffic of the present invention maintains a low average vehicle delay curve during the entire simulation process, especially in the stage where traffic demand fluctuates greatly, the curve changes more smoothly; Figure 4 and Figure 5 The differences between the two methods in total delay and cumulative total delay are shown respectively. The adaptive maximum pressure signal control method for port traffic of the present invention significantly suppresses the rapid accumulation of delay. Figure 6 The adaptive maximum pressure signal control method for port traffic of the present invention is demonstrated to effectively reduce the number of vehicles on the road network during high demand phases. Figures 3 to 6 In FIG. 1 , DEAMP Mean represents the adaptive maximum pressure signal control method applicable to port traffic of the present invention, and Webster Mean represents the traditional Webster method as a control group.

[0081] The above experimental results fully verify the superiority of the adaptive maximum pressure signal control method suitable for port traffic of the present invention.

[0082] In summary, in a complex port traffic environment, the adaptive maximum pressure signal control method suitable for port traffic of the present invention can effectively reduce the number of vehicles on the road network, reduce vehicle delays, and significantly improve traffic operation efficiency through real-time data acquisition and dynamic optimization, and has broad practical application value.

Claims

1. An adaptive maximum pressure signal control method suitable for port traffic, characterized in that By setting a video detector at each entrance direction of each intersection where traffic signal control is required in the port collection and distribution network, and setting a data processor with a pre-stored maximum pressure model at each intersection, the data processor at each intersection is respectively connected to the video detectors set at each entrance direction of the intersection, the data processors of adjacent intersections and the signal light controller of the intersection; the video detectors at each entrance direction of each intersection collect traffic data of the entrance direction in real time according to a preset time interval, and transmit the collected traffic data to the data processor at the intersection, the data processor at the intersection calculates the queue length and road occupancy rate of each entrance direction of the intersection according to the received traffic data, and calculates The result is transmitted to the data processor at the intersection adjacent to the intersection; when controlling a certain intersection, the data processor at the intersection calculates the signal light control logic of the intersection based on the maximum pressure model according to the queue length and road occupancy rate of each entrance direction of the intersection, as well as the queue length and road occupancy rate of the corresponding entrance direction of each intersection adjacent to the intersection, obtains the pressure value of each entrance direction of the intersection, and then determines the next entrance direction to be released of the intersection, and outputs the corresponding release instruction to the signal light controller of the intersection. After receiving the release instruction, the signal light controller switches the signal light display state of the intersection to be consistent with the release instruction, thereby realizing periodic dynamic control of the signal light at the intersection.

2. The adaptive maximum pressure signal control method for port traffic according to claim 1, characterized in that The specific method of setting a video detector at each entrance direction of each intersection that needs to be controlled by traffic signals in the port collection and distribution network is as follows: a video detector is installed above or beside the road in each entrance direction of each intersection that needs to be controlled in the port collection and distribution network, and the shooting area of ​​each video detector can cover all vehicles within a range of 200 meters extending upstream from the stop line in all lanes of the corresponding entrance direction. The video detectors set at each entrance direction of each intersection are connected to the data processor set at the intersection. Each video detector is used to collect traffic flow images in its coverage area in real time through a high-definition camera, and each time a traffic flow image is collected, the traffic flow image is first identified by using image processing technology to obtain the vehicle model, speed and position of all vehicles in the traffic flow image, and then the state of the vehicle is judged according to the speed. If the vehicle speed is less than 1m / s, the vehicle is in a parked state, otherwise it is in a driving state, and the vehicle length is obtained according to the vehicle model of each vehicle and a unique identification code is generated for the vehicle. Finally, the identification code, position, length and state of each vehicle are used to form the traffic data of the vehicle, and the obtained traffic data of all vehicles are transmitted to the data processor connected to it; The video detectors at all entrance directions of each intersection have the same acquisition frequency and the same acquisition time, that is, the video detectors at all entrance directions of each intersection synchronously transmit the acquired traffic data of all vehicles to the data processor connected thereto; When a data processor at a certain intersection receives traffic data transmitted by all video detectors connected to it, the data processor at the intersection processes the received traffic data to obtain the queue length and road occupancy rate of all entrance directions of the intersection, and transmits it to the data processors set at the intersections adjacent to the intersection. The data processors set at each intersection are preset with a data update time interval T s , that is, the data update period, T s <15s, the starting time of data processing for data processors at all intersections is the same, which is recorded as data update 0 time. The data processor at each intersection enters the first data update cycle at data update 0 time, and then every T s , enters the next data update cycle, and the data processor at each intersection performs a data update every time it enters a data update cycle; Let a certain intersection be intersection n, and the number of entrance directions at intersection n be F n , the lth entrance direction at intersection n is called entrance direction l, l = 1, 2, ..., F n , the data update cycle currently entered by the data processor at intersection n is recorded as the tth data update cycle, t = 1, 2, 3, ...; the specific process of the data processor at intersection n performing data update in the tth data update cycle is: S1, the data processor at the intersection n determines whether there is a vehicle in each entrance direction of the intersection n at this time according to the traffic data of the tth data update cycle transmitted by the video detector connected to it; if there is no traffic data of the vehicle in the traffic data transmitted to the data processor at the intersection n by the video detector of a certain entrance direction at the intersection n in the tth data update cycle, then there is no vehicle in the entrance direction at the intersection n in the tth data update cycle, that is, the number of vehicles is 0; otherwise, there is a vehicle in the entrance direction at the intersection n in the tth data update cycle; S2. The number of vehicles at the entrance direction l of intersection n in the tth data update cycle is recorded as N l,n (t), the queue length of the entrance direction l of intersection n in the tth data update cycle is recorded as x l,n (t), the road occupancy rate of the entrance direction l of the intersection n in the tth data update period is recorded as θ l,n (t), according to N l,n (t) Get x l,n (t) and θ l,n (t), specifically: If there is no vehicle in the direction of the lth entrance at intersection n in the tth data update cycle, that is, N l,n (t) = 0, then let x l,n (t) = 0, θ l,n (t) = 0; If there is a vehicle at the lth entrance direction at intersection n in the tth data update cycle, the data processor obtains the number of vehicles N according to the identification codes of all vehicles. l,n (t), each vehicle identification code represents a vehicle, and then, according to formulas (1) and (2), we can calculate x l,n (t) and θ l,n (t): Among them, i l,n (t) = 1, 2, ..., N l,n (t); Indicates the i-th intersection n in the direction l of the t-th data update cycle l,n (t) The parking status of the vehicle. If the vehicle is currently parked, otherwise, Indicates the i-th intersection n in the direction l of the t-th data update cycle l,n (t) The length of the vehicle, λ l,n Indicates the number of lanes in the entrance direction l of intersection n; S3, the data processor set at intersection n first uses the queue length x of the entrance direction l of intersection n in the tth data update cycle l,n (t) Update the queue length of the intersection n entrance direction l, using the road occupancy rate θ of the intersection n entrance direction l in the tth data update cycle l,n (t) updating the road occupancy rate of the entrance direction l of the intersection n at the location thereof, and then transmitting the queue length and road occupancy rate of the entrance direction l of the intersection n at the location thereof to the data processor of the adjacent intersection connected thereto, and receiving the queue length and occupancy rate data of each entrance direction of the adjacent intersection transmitted by the data processor of the adjacent intersection, and using the received queue length and occupancy rate data of each entrance direction of the adjacent intersection to correspondingly update the queue length and occupancy rate data of each entrance direction of the adjacent intersection; Among them, at the data update time 0, the data processor set at intersection n will initialize the number of vehicles, queue length and road occupancy rate of all entrance directions of intersection n, so that the number of vehicles, queue length and road occupancy rate are all 0.

3. The adaptive maximum pressure signal control method for port traffic according to claim 2, characterized in that The data processors set at all intersections that need to be signal controlled are preset with a time interval of 15 seconds, which is the length of the signal control cycle. The starting time of controlling all intersections is recorded as the control time 0, where the control time 0 is the same as the data update time 0. The data processor of each intersection enters the first signal control cycle 15 seconds after the control time 0, and then enters the next signal control cycle every 15 seconds. The data processor at each intersection performs signal control once every time it enters a signal control cycle; The number of exit directions corresponding to the entrance direction l of intersection n is recorded as O l,n The mth exit direction corresponding to the entrance direction l of intersection n is called the exit direction m of the entrance direction l of intersection n. l,n , m l,n =1,2,…,O l,n ; The signal control cycle currently entered by the data processor at intersection n is recorded as the kth signal control cycle, k = 1, 2, 3, ..., the specific process of the signal control of the data processor at intersection n in the kth signal control cycle is: Step 1: The data processor set at intersection n calculates the weight of each entrance direction of intersection n according to the queue length and road occupancy rate of each entrance direction of intersection n and its adjacent signalized intersections pre-stored in the data processor, and the process is as follows: S1.

1. The average road occupancy rate of intersection n in the kth signal control cycle is recorded as Using formula (3) to calculate Among them, θ l ' ,n (k) is the latest data of the road occupancy rate of the entrance direction l of the intersection n currently stored in the data processor set at the intersection n; S1.2 If Then, the weight of the entrance direction l of the intersection n in the kth signal control cycle is calculated according to formula (4): Among them, x l ' ,n (k) represents the latest data of the queue length in the entrance direction l of intersection n currently stored in the data processor of intersection n, Indicates that the vehicle is driving from the entrance direction l of intersection n to the m'th direction corresponding to the entrance direction l. l,n The probability of an exit direction, O l,n is the number of exit directions corresponding to the entrance direction l of intersection n; Indicates the m'th corresponding to the entrance direction l of the current intersection n l,n The length of the queue in the exit direction is l,n When there is no corresponding intersection requiring traffic signal control in the exit direction, it is equal to 0. l,n When there is a corresponding intersection that needs traffic signal control for the exit direction, it is equal to the m'th entry direction l of intersection n currently saved in the data processor of intersection n. l,n The latest data of the queue length in the entrance direction of the adjacent intersection corresponding to the exit direction; Indicates the m'th corresponding to the entrance direction l of the current intersection n l,n The road occupancy rate of the exit direction is the m'th l,n When there is no corresponding intersection requiring traffic signal control in the exit direction, it is equal to 0. l,n When there is a corresponding intersection that needs traffic signal control for the exit direction, it is equal to the m'th entry direction l of intersection n currently saved in the data processor of intersection n. l,n The latest data on the road occupancy rate of the entrance direction of the adjacent intersection corresponding to the exit direction; like Then, the weight of the entrance direction l of the intersection n in the kth signal control cycle is calculated according to formula (5): Step 2: The pressure of the entrance direction l of the intersection n in the kth signal control period is recorded as p l,n (k), and p is calculated according to formula (6) l,n (k): p l,n (k)=w l,n (k)×1800 (6) Step 3: Get p 1,n (k) to The maximum value among the values ​​is used as the entrance direction of intersection n corresponding to the maximum value, which is recorded as the entrance direction S n * (k), and generates a corresponding release instruction to transmit to the signal light controller at the intersection n connected to it. When the signal light controller at the intersection n receives the release instruction, if S n * (k) The direction is consistent with the entrance currently being released, Then the signal light controller set at intersection n is S at the entrance direction of intersection n. n * (k) Set the green light and the red light for other entrance directions, and re-record the entrance direction S n * (k) Green light duration; If S n * (k) If the direction of the entrance is not consistent with the direction of the current entrance, the signal controller at intersection n will first set the yellow light for the direction of the entrance currently being released at intersection n, and the signal lights for other entrance directions will remain unchanged. After 3 seconds, the signal lights for the entrance direction S will be set to yellow. n * (k) Set the green light and set the red light for other entrance directions, and re-record the entrance direction S n * (k) The green light duration.

Images