Intersection dynamic signal control method, electronic equipment, storage medium and program product

Abstract

The invention provides an intersection dynamic signal control method, electronic equipment, a storage medium and a program product, and relates to the technical field of intelligent traffic. The method comprises the following steps: acquiring a phase set and a lane index set of a to-be-controlled intersection in a road network, and current queuing information, an acceleration information set and time headway information of each lane corresponding to the to-be-controlled intersection; determining a to-be-controlled target phase in the to-be-controlled intersection according to the current queuing information, the phase set and the lane index set; for each lane corresponding to the target phase, adjusting the current queuing information of the lane according to the speed increasing information set of the lane to obtain target queuing information; and based on the target phase, and the time headway information and the target queuing information of all lanes corresponding to the target phase, controlling the signal of the to-be-controlled intersection. According to the invention, the traffic signal control is more accurate, so that the road congestion is reduced, and the intersection passing efficiency is improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent transportation technology, and in particular to a method for controlling dynamic signals at intersections, electronic equipment, storage media, and program products. Background Art

[0002] With the acceleration of urbanization, the continuous expansion of urban areas has led to increasingly serious traffic congestion. As the core part of the urban transportation system, traffic signal control has the main task of intelligently regulating intersection lights, optimizing road resource allocation, improving traffic efficiency, and ensuring smooth travel for citizens.

[0003] Existing traffic signal control technology has gone through several stages of development. The early fixed-time cycle control presets the signal light change cycle based on historical traffic flow data. It is simple to operate but lacks flexibility and is difficult to cope with real-time traffic changes; the key road green wave control strategy forms a "green wave belt" by coordinating the green light time of the main road signal lights to reduce the frequency of vehicle stops and starts; the induction control technology uses detection equipment to monitor the traffic flow at the intersection in real time and dynamically adjusts the signal light phase accordingly; the maximum pressure rule calculates the difference between vehicles entering and leaving the lanes at the intersection, and gives priority to allocating green lights to the direction with high pressure.

[0004] However, existing traffic signal control methods still have the problem of poor control effect when facing complex and changeable traffic environments. Summary of the invention

[0005] The embodiments of the present application provide a dynamic intersection signal control method, electronic device, storage medium, and program product to solve the problem that the existing traffic signal control method still has poor control effect when facing a complex and changeable traffic environment.

[0006] In a first aspect, an embodiment of the present application provides a method for controlling dynamic traffic signals at an intersection, comprising:

[0007] Obtaining a phase set, a lane index set, and current queue information, a speed increase information set, and headway information of each lane corresponding to the intersection to be controlled in the road network, wherein the speed increase information set includes speed increase information of each lane of the intersection to be controlled in multiple time periods before the current time stored in the database, wherein each speed increase information is determined according to the lane queue length and collection time collected at the intersection to be controlled in the corresponding time period, and is used to indicate the change rate of the queue length of the corresponding lane at the intersection to be controlled during the red light process;

[0008] Determine a target phase to be controlled at the intersection to be controlled according to current queue information, a phase set and a lane index set;

[0009] For each lane corresponding to the target phase, the current queue information of the lane is adjusted according to the speed increase information set of the lane to obtain the target queue information;

[0010] Based on the target phase, and the headway time information and target queue information of all lanes corresponding to the target phase, the signal of the intersection to be controlled is controlled.

[0011] In a possible implementation, the current queue information of the lane is adjusted according to the speed increase information set of the lane to obtain the target queue information, including:

[0012] Determine a lane correlation parameter based on the historical speed information of all lanes in the road network, wherein the historical speed information of each lane is the speed information of the corresponding lane stored in the database before the current time, and the lane correlation parameter is used to indicate the traffic impact relationship between lanes;

[0013] According to the speed increase information set of the lane and the lane correlation parameters, the current queue information of the lane is adjusted to obtain the target queue information.

[0014] In a possible implementation, the lane correlation parameter is determined based on the historical speed information of all lanes in the road network, including:

[0015] Obtaining multiple historical growth rate information of each lane in the road network within a preset first time period;

[0016] Constructing a sample set according to multiple historical speed increase information of all lanes, wherein each sample includes a second time period in the first time period and historical speed increase information of all lanes in the corresponding second time period;

[0017] According to the adjacency relationship in the road network, a road network adjacency matrix is ​​constructed, and the road network adjacency matrix is ​​used to indicate the connectivity between every two lanes in the road network;

[0018] Based on the sample set and the road network adjacency matrix, the preset correlation parameters are adjusted to obtain the lane correlation parameters.

[0019] In a possible implementation, according to the speed increase information set of the lane and the lane correlation parameter, the current queue information of the lane is adjusted to obtain the target queue information, including:

[0020] Obtaining a sum of the speed information according to a speed information subset in the speed information set of the lane, wherein the speed information subset includes a plurality of speed information obtained by sampling the speed information of the lane stored in the database within a preset third time period;

[0021] Determine second speed information according to the lane correlation parameter and first speed information in the speed information set of the lane, where the first speed information is speed information of the lane in a second time period before the current time stored in the database;

[0022] The target queuing information is determined based on the sum of the speed increase information, the second speed increase information, the current queuing information and the signal decision time, wherein the signal decision time is the time to trigger the execution of the steps of obtaining the phase set, lane index set, and current queuing information, speed increase information set and vehicle headway information of each lane corresponding to the intersection to be controlled in the road network.

[0023] In a possible implementation, before obtaining a phase set, a lane index set, and current queue information, a speed increase information set, and headway information of each lane corresponding to the intersection to be controlled in the road network, the method further includes:

[0024] Obtain a decision request for signal control, which includes the signal decision time set by the user and the intersection to be controlled;

[0025] Accordingly, when obtaining the phase set, lane index set, and current queue information, speed increase information set, and headway information of each lane corresponding to the intersection to be controlled in the road network, the intersection to be controlled is the intersection to be controlled in the decision request;

[0026] or,

[0027] Determine the green light countdown length corresponding to all intersections in the road network, and the green light end time corresponding to each intersection at the current time;

[0028] For each intersection, if the time from the current time to the end time of the green light corresponding to the intersection is greater than or equal to the green light countdown length corresponding to the intersection, the current time is used as the signal decision time and the intersection is used as the intersection to be controlled.

[0029] In a possible implementation, the method further includes:

[0030] Get the phase status of traffic lights at all intersections in the road network;

[0031] For each intersection, when the phase state of the traffic light at the intersection indicates that the signal light turns red, obtain the first queue length and the first acquisition time of each lane corresponding to the intersection;

[0032] When the phase state of the signal light indicates that the signal light turns green, the second queue length, the second acquisition time, and the headway time data of each lane corresponding to the intersection are obtained, and after the phase state of the signal light indicates that the signal light turns red again, the headway time average of each lane is determined according to the headway time data; wherein the headway time information includes the headway time data and the headway time average of each lane;

[0033] For each lane corresponding to the intersection, determine the speed increase information of the lane in the second time period according to the first queue length, the first acquisition time, the second queue length and the second acquisition time of the lane;

[0034] The speed information of all lanes corresponding to all intersections is stored in the database.

[0035] In a possible implementation, determining the speed increase information of the lane in the second time period according to the first queue length, the first acquisition time, the second queue length, and the second acquisition time of the lane includes:

[0036] Determine an intermediate time between a first acquisition time and a second acquisition time of the lane;

[0037] If the intermediate moment is within the second time period, determining the speed increase information of the lane within the second time period based on a first difference between the second queue length and the first queue length of the lane and a second difference between the second acquisition time and the first acquisition time;

[0038] If the intermediate time is after the second time period, determining third speed increase information of the lane based on a first difference between the second queue length and the first queue length of the lane and a second difference between the second acquisition time and the first acquisition time;

[0039] Based on the third speed increase information and the fourth speed increase information, the speed increase information of the lane within the second time period is determined, wherein the fourth speed increase information is the speed increase information of the lane within the second historical time period stored in the database, and the second historical time period is the second time period corresponding to when the speed increase information of the lane was stored before the first acquisition time.

[0040] In a possible implementation, determining a target phase to be controlled at the intersection to be controlled according to current queuing information, a phase set, and a lane index set includes:

[0041] Determine lane flow information according to the phase set and the lane index set, where the lane flow information is used to indicate a mapping relationship between an upstream lane and a downstream lane in the intersection to be controlled;

[0042] According to the current queue information and lane flow information, determine the queue volume difference between the upstream lane queue length and the downstream lane queue length corresponding to each phase;

[0043] Based on the queue volume differences corresponding to all phases, a target phase in the phase set is determined.

[0044] In a possible implementation, based on the target phase, and the headway information and target queue information of all lanes corresponding to the target phase, the signal of the intersection to be controlled is controlled, including:

[0045] For each lane corresponding to the target phase, based on the headway information of the lane, determine the starting time corresponding to the lane and the number of vehicles passing within the starting time;

[0046] Determine the lane pass time based on target queue information, vehicle number, start time, and headway information;

[0047] Determine a target passing time of the target phase according to the passing times of all lanes corresponding to the target phase, and the maximum phase time and the minimum phase time of the target phase;

[0048] Based on the target phase and the target passing time of the target phase, the signal of the intersection to be controlled is controlled.

[0049] In a second aspect, an embodiment of the present application provides a dynamic signal control device for an intersection, comprising:

[0050] An acquisition module, used to acquire a phase set, a lane index set, and current queue information, a speed increase information set, and headway information of each lane corresponding to the intersection to be controlled in the road network, wherein the speed increase information set includes speed increase information of each lane of the intersection to be controlled in multiple time periods before the current time stored in the database, wherein each speed increase information is determined according to the queue length and collection time of the lane collected at the intersection to be controlled in the corresponding time period, and is used to indicate the change rate of the queue length of the corresponding lane at the intersection to be controlled during the red light process;

[0051] A determination module, used to determine a target phase to be controlled in the intersection to be controlled according to current queuing information, a phase set and a lane index set;

[0052] An adjustment module, for adjusting the current queue information of each lane corresponding to the target phase according to the speed increase information set of the lane, to obtain the target queue information;

[0053] The control module is used to control the signal of the intersection to be controlled based on the target phase and the headway time information and target queue information of all lanes corresponding to the target phase.

[0054] In a third aspect, an embodiment of the present application provides an electronic device, including: a memory, a processor;

[0055] Memory stores computer-executable instructions;

[0056] The processor executes the computer-executable instructions stored in the memory, so that the processor executes the first aspect and / or various possible implementations of the first aspect as described above.

[0057] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed, they are used to implement the first aspect above and / or various possible implementation methods of the first aspect.

[0058] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program, which, when executed, implements the above first aspect and / or various possible implementation methods of the first aspect.

[0059] The dynamic signal control method, electronic device, storage medium and program product for intersections provided in the embodiments of the present application obtain relevant information of the intersection, including a phase set, a lane index set, and the current queue information, speed increase information and headway time information of each lane, wherein the speed increase information is calculated based on historical data and is used to indicate the rate of change of the lane queue length during the red light period; and based on the current queue information, the phase set and the lane index set, determine the target phase that needs to be adjusted to ensure that the signal control can be focused on the traffic flow that most needs to be optimized at the current time; for each lane under the target phase, use the speed increase information to adjust the current queue information to obtain more accurate target queue information; based on the target phase and the headway time information and target queue information of the lane, optimize the control of the traffic light. The method can dynamically adapt to changes in traffic flow, more accurately predict and adjust the switching timing of traffic lights, reduce the waiting time of vehicles at intersections and traffic congestion, and improve the traffic efficiency of traffic flow. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0061] Figure 1 A schematic diagram of a flow chart of a dynamic signal control method for an intersection provided in this application;

[0062] Figure 2 A flowchart of a lane correlation parameter determination method provided in this application;

[0063] Figure 3 A schematic diagram of another flow chart of a dynamic signal control method for an intersection provided in this application;

[0064] Figure 4 A flow chart of a signal decision time determination method provided in this application;

[0065] Figure 5 A flow chart of the red light queue increment and headway time collection process provided for this application;

[0066] Figure 6 A schematic diagram of the structure of the dynamic signal control device for intersections provided in this application;

[0067] Figure 7 A schematic diagram of the structure of the electronic device provided in this application.

[0068] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0069] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0070] In the existing technology, traffic signal control technology usually includes: early fixed time cycle control, which presets the signal light change cycle based on historical traffic flow data. It is simple to operate but lacks flexibility and is difficult to cope with real-time traffic changes; the key road green wave control strategy forms a "green wave belt" by coordinating the green light time of the signal lights on the main roads, reducing the frequency of vehicle stops and starts; the induction control technology uses detection equipment to monitor the traffic flow at the intersection in real time, and dynamically adjusts the phase of the signal light accordingly; the maximum pressure rule calculates the difference between vehicles entering and exiting the lanes at the intersection, and gives priority to allocating green lights to the direction with high pressure.

[0071] However, fixed-time period signal control has poor adaptability and cannot be flexibly adjusted according to real-time traffic flow, which will cause waiting waste during non-peak hours and congestion during peak hours, and resource allocation is unreasonable; the green wave control strategy for key roads has a limited scope of application and is only applicable to specific main roads. It is not effective in the face of complex and changeable traffic networks, and is highly dependent on precise time synchronization, making it difficult to adapt quickly to emergencies; the induction control technology is highly dependent on equipment, has high installation and maintenance costs, and has large data processing requirements. It is limited to control of a single intersection, and manual threshold setting relies on experience, is prone to errors, and requires a lot of monitoring and trial and error; the maximum pressure method is prone to uneven waiting time, lacks a learning mechanism, and is difficult to adapt to new situations. It is also limited to a single intersection and has poor effects on coordinated control of the road network.

[0072] With the development of artificial intelligence technology, existing traffic signal control methods have also incorporated reinforcement learning technology, such as using online reinforcement learning to iteratively optimize control strategies through continuous trial and error and receiving reward feedback, relying on a large amount of real-time interactive data and stable and effective learning algorithms to achieve optimization of traffic signal control; or using offline reinforcement learning to collect, organize and train historical data, and perform control according to established strategies after model training. However, the trial and error cost of online reinforcement learning is high, which can easily cause delays and safety risks in the traffic system, is difficult to train, requires a large amount of interactive data, and is difficult to determine in real time to adjust the training cycle; offline reinforcement learning has data dependence and leakage risks, slow update iterations, and cannot respond to changes in traffic structure or flow in real time.

[0073] In order to solve the above problems, the dynamic signal control method for intersections provided in the present application collects the phase set and lane index set of the intersection to be controlled, as well as the current queue information of the lane, the speed increase information set obtained from the historical time period data, and the headway information, and constructs the speed increase information by recording and analyzing the lane queue lengths and collection times in different time periods, thereby laying the foundation for subsequent decision-making; then, the target phase to be controlled is determined by combining the current actual queue situation with the intersection phase and lane layout information, and key control is highlighted; then, for the lanes involved in the target phase, the speed increase information set is used to dynamically adjust the current queue information, comprehensively considering the traffic flow inertia and changing trends, rather than just looking at the immediate status; finally, the traffic lights at the intersection to be controlled are regulated based on the adjusted target queue information and headway information. Therefore, with the help of multi-dimensional information, especially the growth rate information set, this application can not only accurately capture the dynamic changes in traffic flow characteristics of each lane and quickly adapt to changes in traffic flow, especially during peak hours, but also effectively reduce the occurrence of traffic congestion and improve intersection traffic efficiency. It can also focus on the target phase and targeted regulation of the corresponding lanes to avoid the waste of resources caused by indiscriminate unified control, and reasonably allocate green light time based on the headway time, reduce waiting time, and maximize road use efficiency.

[0074] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0075] The execution subject of the intersection dynamic signal control method provided in the embodiment of the present application can be a server. Among them, the server can be a mobile phone, a computer, a tablet and other devices. The embodiment of the present application does not impose any special restrictions on the implementation method of the execution subject, as long as the execution subject can obtain the phase set, lane index set, and the current queue information, speed increase information set and headway information of each lane corresponding to the intersection to be controlled in the road network, the speed increase information set includes the speed increase information of each lane of the intersection to be controlled in multiple time periods before the current time stored in the database, wherein each speed increase information is determined according to the lane queue length and collection time collected at the intersection to be controlled in the corresponding time period, and is used to indicate the change rate of the queue length of the corresponding lane at the intersection to be controlled during the red light process; according to the current queue information, the phase set and the lane index set, the target phase to be controlled in the intersection to be controlled is determined; for each lane corresponding to the target phase, according to the speed increase information set of the lane, the current queue information of the lane is adjusted to obtain the target queue information; based on the target phase, and the headway information and target queue information of all lanes corresponding to the target phase, the signal of the intersection to be controlled can be controlled.

[0076] It should be noted that the database can be a general term for a local storage database that stores data required to execute the dynamic signal control method for intersections, and it can include a traffic status database, a red light queue growth rate database, a headway time database, and a lane correlation database. Among them, the traffic status database can be used to store multiple categories of traffic data such as queue length and number of queued vehicles collected in the road network; the red light queue growth rate database can be used to store the growth rate information of each lane at each intersection in the road network corresponding to multiple time periods; the headway time database can be used to store the headway time information collected and calculated in the road network; the lane correlation database can be used to store lane correlation parameters in the road network, and based on the dynamic changes over time, the lane correlation database can have lane correlation parameters for multiple time periods.

[0077] Figure 1 The flowchart of a dynamic signal control method for intersections provided by the present application is shown in FIG. The execution subject of the method may be a server or other server storing the dynamic signal control method for intersections, and this embodiment is not particularly limited here. Figure 1 As shown, the method may include:

[0078] S101. Acquire a phase set, a lane index set, and current queue information, a speed increase information set, and vehicle headway information of each lane corresponding to the intersection to be controlled in the road network, wherein the speed increase information set includes speed increase information of each lane of the intersection to be controlled in multiple time periods before the current time stored in a database, wherein each speed increase information is determined based on the queue length and collection time of the lane collected at the intersection to be controlled in the corresponding time period, and is used to indicate the rate of change of the queue length of the corresponding lane at the intersection to be controlled during a red light process.

[0079] The phase set refers to the combination of different states of the traffic light in a traffic light cycle. For example, at a common intersection, the green light for east-west straight driving and the green light for north-south left turning are both on at the same time, which can be regarded as a phase. Different combinations of signal lights constitute a phase set, which is used to regulate the traffic order of vehicles in different directions at the intersection.

[0080] The lane index set can be a set of identifiers for all lanes in the controlled intersection, which is used to distinguish different lanes for data collection and signal control. For example, the eastbound straight lane at the intersection is numbered 1, the eastbound left turn lane is numbered 2, and so on. These indexes can accurately point to specific lanes.

[0081] The current queue information may refer to the length of the queues waiting to pass through the intersection on each lane at the current time, which may be obtained through detection means such as cameras or geomagnetic sensors. The queue information may also be the number of vehicles in the lane, and correspondingly, the speed increase information is determined according to the number of vehicles in the lane collected at the intersection to be controlled in the corresponding time period and the collection time.

[0082] The speed information set may be a data set containing changes in lane queue lengths over multiple time periods in the past. For example, the speed information corresponding to the minute closest to the current time in the past 30 minutes and the speed information corresponding to every other minute in the past 30 minutes constitute the speed information set for analyzing lane congestion trends. Optionally, the speed information is the speed information per minute obtained by dividing the queue length difference collected at adjacent times by the time interval.

[0083] Headway information refers to the time interval between the heads of two adjacent vehicles on the same lane. When a lane is passing, if the head of the rear vehicle reaches a certain observation point 3 seconds after the front vehicle passes the observation point, the 3 seconds is the headway at this time, which can be used to measure the density and speed of traffic flow.

[0084] S102: Determine a target phase to be controlled in the intersection to be controlled according to current queuing information, a phase set and a lane index set.

[0085] In this step, the current queue information obtained by monitoring the current time is used to determine the congestion level of each lane. For example, if the queues of vehicles in some lanes are long and the congestion is serious, then the phase set and lane index set are combined to check which phases involve these severely congested lanes, and the phases that are key to alleviating the current congestion are prioritized as target phases. For example, if the east-west straight lane is severely congested, and the east-west straight green light phase can directly regulate the flow direction, then the east-west straight green light phase is determined as the target phase.

[0086] S103. For each lane corresponding to the target phase, according to the speed increase information set of the lane, adjust the current queue information of the lane to obtain target queue information.

[0087] Furthermore, according to the growth rate information in the growth rate information set, the current queue information is predicted and adjusted to obtain more accurate target queue information. For example, the queue changes in the next five minutes are predicted using the data of the past hour, and the traffic lights are adjusted to reduce the queue length.

[0088] S104: Based on the target phase, and the headway information and target queue information of all lanes corresponding to the target phase, the signal of the intersection to be controlled is controlled.

[0089] Among them, the switching time and order of the signal lights can be dynamically adjusted according to the adjusted target queue information and headway information. For example, if the headway of a lane is small, it indicates that the traffic is dense, and the green light time of the lane can be appropriately extended. Furthermore, according to the target queue information and headway information, the preset optimization algorithm or calculation rule is used to calculate the green light duration of each lane, and a signal light control instruction is generated, which is executed by the traffic signal machine to achieve intersection traffic optimization.

[0090] The dynamic signal control method for intersections provided in the embodiment of the present application collects multi-dimensional information of intersections, uses historical growth rates to predict traffic flow changes, accurately locates key control phases, and then dynamically adjusts the queuing conditions of each lane to implement refined signal light control. This can improve the accuracy and adaptability of traffic signal control at intersections, flexibly allocate green light time according to real-time traffic conditions, reduce vehicle waiting time, improve intersection traffic efficiency, and alleviate urban traffic congestion.

[0091] On the basis of the above-mentioned embodiment, the method described in S103 for adjusting the current queuing information of the lane according to the speed increase information set of the lane to obtain the target queuing information may include: determining the lane correlation parameters according to the historical speed increase information of all lanes in the road network, wherein the historical speed increase information of each lane is the speed increase information of the corresponding lane stored in the database before the current time, and the lane correlation parameters are used to indicate the traffic impact relationship between the lanes; adjusting the current queuing information of the lane according to the speed increase information set of the lane and the lane correlation parameters to obtain the target queuing information.

[0092] The historical growth rate information may refer to the change rate of the queue length of each lane within a certain period of time in the past. These data are stored in the database and used to analyze the traffic flow trend of the lane.

[0093] In this step, the historical growth rate information of each lane in the road network over a long period of time before the current time can be extracted from the database; then, data analysis algorithms, such as correlation coefficient algorithms, are used to calculate all lanes pairwise to obtain a lane correlation matrix. Each element in the matrix is ​​a lane correlation parameter, which comprehensively reflects the traffic impact relationship between lanes.

[0094] For example, suppose that the current queue information of lane A corresponding to a certain target phase is a queue length of 30 meters, and its growth rate information set shows that the growth rate has increased and decreased in the past few periods. At the same time, the lane correlation parameter shows that lane A is strongly positively correlated with the adjacent lane B, and lane B is currently in a trend of increasing congestion (judged from its growth rate information). Then, considering all factors, it is estimated that lane A will be affected by lane B in the future, and the number of queued vehicles may increase. The current queue information is adjusted upward to obtain the target queue information, such as 35 meters. On the contrary, if the lane with a strong correlation with lane A is in a state of congestion relief, and lane A's own growth rate slows down, the current queue information is appropriately adjusted downward.

[0095] It should be noted that the lane correlation parameter may include one or more parameters, which are not particularly limited here. For example, the lane correlation parameter includes the correlation matrix W parameter and the bias b parameter. Among them, the b parameter can reflect some basic traffic characteristics or initial states of each lane itself without considering the mutual influence between lanes (reflected by the W parameter), such as the design capacity of the lane itself, the inherent traffic flow tendency caused by its position in the traffic network (such as lanes close to commercial areas may have higher basic flow), etc., which are reflected in the b parameter.

[0096] By introducing lane correlation parameters, we can more comprehensively consider the mutual influence of traffic between different lanes, and no longer view traffic changes in a single lane in isolation. When adjusting the target queue information, we can make a comprehensive prediction based on the conditions of surrounding lanes, making the prediction of lane queue conditions more accurate. In the subsequent signal control link, we can allocate green light time based on target queue information that is more in line with actual traffic dynamics, further improving the accuracy of traffic signal control at intersections, more efficiently reducing vehicle waiting time, and enhancing the traffic efficiency of the entire intersection and even the road network.

[0097] On the basis of the above embodiments, according to the historical growth rate information of all lanes in the road network, the method for determining the lane correlation parameters may include: obtaining multiple historical growth rate information of each lane in the road network within a preset first time period; constructing a sample set based on the multiple historical growth rate information of all lanes, wherein each sample includes a second time period in the first time period, and the historical growth rate information of all lanes in the corresponding second time period; constructing a road network adjacency matrix based on the adjacency relationship in the road network, and the road network adjacency matrix is ​​used to indicate the connectivity between every two lanes in the road network; based on the sample set and the road network adjacency matrix, adjusting the preset correlation parameters to obtain the lane correlation parameters.

[0098] The first time period may be a longer time range for collecting historical growth rate information, such as one day, one week, or one month. The second time period may be a shorter time segment for constructing a sample, such as one minute. Optionally, the time period corresponding to the growth rate information stored in the database is used as the second time period.

[0099] The adjacency relationship in the road network can be determined by the channelization state of the lanes. Lane channelization refers to the reasonable division and guidance of lanes on the road by setting up traffic signs, markings, diversion islands and other facilities. In the channelization state, each lane has a clear function and flow direction. For example, at an intersection, some lanes are specifically used for left turns and some are used for straight driving, and there is a specific connection relationship between lanes in different directions and lanes in other directions.

[0100] The method for adjusting the preset correlation parameters may be through statistical analysis or machine learning algorithms. For example, elastic net regression, time series analysis, linear regression or neural network model is used to input a sample set and an adjacency matrix, and output the adjusted correlation parameters. There is no particular restriction on the adjustment method.

[0101] In one example, nonlinear regression methods (such as polynomial regression, radial basis function regression, etc.) are used to estimate the growth rate of red light queues. These models can better capture the complex relationship between variables and may provide more accurate predictions. For example, by collecting traffic flow data, including vehicle queue length, time interval, etc.; using these data to train nonlinear regression models; using the trained model to predict the queue growth rate in real time, and adjusting the current queue information accordingly to optimize the control of traffic lights. This improves the accuracy of model predictions and can adapt to more complex traffic patterns.

[0102] In another example, time series analysis techniques (such as ARIMA models, seasonal decomposition, etc.) are used to predict the growth rate of red light queues. It can use the time dependency of historical data to improve the prediction accuracy. For example, collect traffic flow data over a period of time; apply a time series analysis model for training; and use the model output for real-time traffic control. This takes advantage of trends and seasonal patterns in historical data and performs better for time-related data.

[0103] In another example, neural networks, especially recurrent neural networks or long short-term memory networks, are used to predict the growth rate of red light queues. These networks are particularly suitable for processing sequence data and can capture long-term dependencies in time series. For example, collect detailed traffic flow and queue data; train neural network models to predict future queue growth rates; and apply the prediction results in real time to adjust traffic signals. This method can handle a large number of input variables and complex patterns, and the model performance can continue to improve as the amount of data increases.

[0104] Optionally, a parameter calculation cycle can be set for the method of determining the lane correlation parameters, so that the correlation parameters are dynamically updated according to the real-time updated speed increase information, to ensure that when the lane correlation parameters are subsequently used to adjust the lane queuing information and perform intersection signal control, they can better fit the actual traffic conditions.

[0105] Taking the elastic net regression method as an example, Figure 2 A flow chart of a lane correlation parameter determination method provided in this application. Figure 2 As shown, the method for obtaining lane correlation parameters includes: 1) setting the correlation matrix calculation cycle , set the road network adjacency matrix ; 2) Every time (otherwise wait), by sending a request to obtain the past from the red light queue speed database The growth rate of the red light queue of all lanes in the time period is H; 3) The growth rate of each lane i of N lanes And other lanes All historical data , forming a data set, N is the total number of lanes in the road network; 4) The lane correlation matrix and bias are calculated using the elastic net regression method and stored in the lane correlation database.

[0106] Below Figure 2 The method shown is exemplarily described:

[0107] 1) Based on the actual traffic data storage situation, Take the past 3 hours as the data, and then extract the data of the same time period in a week and perform data cleaning and preprocessing to form a data sample. For example, if the period of the past 3 hours is from 9 am to 12 noon on the same day, then extract the data from 9 am to 12 noon every day in the week before the current one. Through this extraction method, since 3 hours are extracted every day, the total extraction time for 7 days a week is 21 hours. Because some data does not meet the requirements or is null, the data is cleaned and preprocessed, and finally 11 hours of data is retained. The extracted data is divided into multiple time steps according to certain rules, with a total of 11×60=660 time step samples.

[0108] 2) Determine the road network adjacency correlation: First, create a road network adjacency matrix based on the connectivity representation of all lanes in the channelized state , which is a type The 0 / 1 binary space represents the connectivity between N lanes. In its adjacency matrix, the row direction and represent the in-degree, and the column direction and represent the out-degree. If , indicating that the jth lane will lead to the ith lane;

[0109] According to the configured road network adjacency matrix , construct a multi-order road network adjacency matrix ,nature express The n-order expression can be configured as follows: : indicates first-order adjacency), ( : represents second-order adjacency), ( : indicates third-order adjacency) and other multiple orders. In this embodiment, second-order neighbors are selected .

[0110] 3) Adjustment of preset correlation parameters: Creating a learnable parameter and bias . Each element of W is represented by Perform mask operation, where The elements in W corresponding to the positions of 0 in are set as non-learnable parameters and are permanently kept at 0. In addition, in order to avoid the influence of autoregression, a mask matrix with 0 on the main diagonal and 1 on the rest is used to perform a second mask operation on W to ensure that the main diagonal elements of W are also 0, so as to ensure that the model only learns the mutual influence between lanes, rather than its own historical data; finally, the parameters W and b are obtained, which will be used for the subsequent elastic net linear regression analysis.

[0111] Elastic Net Regression: Known Population Dataset samples, and divide them into training set and test set in the ratio of 8:2; elastic net regression is used to regress the training set at the parameter and b to obtain a linear regression model; verify it on the test set to ensure that a set of relatively high accuracy is obtained and b, which are lane correlation parameters; the lane correlation parameters are stored in the lane correlation database. Optionally, multiple calculated lane correlation parameters are retained in the lane correlation database, and the corresponding storage time is recorded to prevent the current lane correlation parameters from being incorrect due to external factors such as data loss when the latest lane correlation parameter is calculated. and b cannot be used. At the same time, when this happens, the lane correlation parameters with storage time close to the latest calculation time in the database can be directly used for subsequent calculation and processing.

[0112] By reasonably dividing the time periods to obtain rich historical growth rate samples, combined with the adjacency matrix that reflects the road network connection status, the algorithm is used to optimize the initial parameters, making the determination of lane correlation parameters more accurate and reliable; fully considering the dynamic changes of traffic flow in different time periods and the actual topological structure of the road network, so that the adjusted lane queuing information can be more in line with the actual traffic conditions.

[0113] On the basis of the above embodiment, according to the lane's speed increase information set and lane correlation parameters, the current queue information of the lane is adjusted to obtain the target queue information. The method may include: obtaining the sum of the speed increase information according to the speed increase information subset in the lane's speed increase information set, the speed increase information subset including multiple speed increase information obtained by sampling the speed increase information of the lane stored in the database within a preset third time period; determining the second speed increase information according to the lane correlation parameters and the first speed increase information in the lane's speed increase information set, the first speed increase information being the speed increase information of the lane stored in the database in a second time period before the current time; determining the target queue information according to the sum of the speed increase information, the second speed increase information, the current queue information and the signal decision time, wherein the signal decision time is the time to trigger the execution of the steps of obtaining the phase set of the intersection to be controlled in the road network, the lane index set, and the current queue information, the speed increase information set and the headway information of each lane corresponding to the intersection to be controlled.

[0114] In this embodiment, the acceleration information subset is a part of the data in the acceleration information set. For example, data within a preset time period is sampled from the database, such as 30 minutes before the current time (i.e., the third time period). Optionally, the third time period can be set according to actual analysis conditions. Normally, in order to obtain recent and timely lane acceleration dynamics, the third time period is greater than the second time period and less than the first time period.

[0115] The sum of the speed increase information may be the cumulative result of all the speed increase information in the speed increase information subset, and is used to quantify the overall queue change trend of the lane in the third time period.

[0116] Furthermore, the sum of the speed-up information and the second speed-up information may satisfy:

[0117] ;

[0118] in, is the sum of the growth rate information; For the third time period; is the growth rate information of the i-th time period in the third time period, ; is the second speed increase information; The influence factor indicates the degree to which the current intersection considers the influence of other intersections, usually between 0.2 and 0.5; is the multiplication operation; is the first speed information, ; and is the lane correlation parameter; is the identity matrix.

[0119] The target queue information can satisfy:

[0120] ;

[0121] in, Queue information for the target; Current queue information; Decision time for signal; is the lane index set corresponding to the target phase.

[0122] By combining the sum of speed information, lane correlation parameters and signal decision time, the lane queuing information can be predicted and adjusted more accurately.

[0123] On the basis of the above-mentioned embodiment, before S101, the dynamic signal control method for intersections may further include: obtaining a decision request for signal control, the decision request including a signal decision time set by a user and an intersection to be controlled; accordingly, when obtaining a phase set, a lane index set, and current queue information, a speed increase information set and vehicle headway information of each lane corresponding to the intersection to be controlled in the road network, the intersection to be controlled is the intersection to be controlled in the decision request; or, determining the green light countdown length corresponding to all intersections in the road network, and the green light end time corresponding to each intersection at the current time; for each intersection, if the time length from the current time to the green light end time corresponding to the intersection is greater than or equal to the green light countdown length corresponding to the intersection, the current time is used as the signal decision time, and the intersection is used as the intersection to be controlled.

[0124] In this embodiment, the signal decision time and the method for determining the intersection to be controlled may include two types, namely, passive decision type or active decision type. The decision request for signal control may be a request triggered by a user or a system, used to start the signal control process, and the request includes the signal decision time set by the user and the specific intersection to be controlled. The signal decision time may refer to a time point set by the user or automatically determined by the system, at which the signal control process is started.

[0125] For example, Figure 3 A flow chart of another method for controlling dynamic signal at an intersection provided by the present application is shown as follows: Figure 3 As shown, when an external decision request (passive decision type) is detected, data preparation is performed through the database, and the target phase and the headway information and target queue information (decision result) of all lanes corresponding to the target phase are determined according to the decision calculation formula (corresponding to the method described in the above embodiment); then the decision result is pushed to the signal control machine (traffic signal control machine) / signal control platform (traffic signal control platform) to achieve control of the signal at the intersection to be controlled. When the signal control process is an active decision type, each intersection is traversed, and the green light countdown length corresponding to the intersection and the green light end time corresponding to each intersection at the current time are used to determine whether the intersection needs a decision (i.e., execute the signal control process) at the current time point. If necessary, the execution steps are similar to those of the passive decision type, and will not be repeated here.

[0126] The green light countdown length may refer to the countdown length of the green light state set for the intersection at the current time in the road network, and the green light end time refers to the specific time point at which the current green light state is expected to end.

[0127] Figure 4 A flow chart of a signal decision time determination method provided in this application, such as Figure 4As shown, when the signal control process is of active decision-making type, it is detected that the current time point is exactly at the beginning of the green light countdown (that is, the time length from the current time to the end time of the green light corresponding to the intersection is equal to the length of the green light countdown corresponding to the intersection), then it is considered that a decision needs to be made, and the current time is used as the signal decision time.

[0128] By introducing signal control decision requests and a mechanism for automatically determining intersections to be controlled, it is possible to flexibly respond to user needs and real-time traffic conditions, ensure that signal control decisions are made at the most appropriate time, improve the initiative and accuracy of traffic signal control, and more effectively adapt to dynamic traffic environments.

[0129] On the basis of the above embodiments, the dynamic signal control method for intersections may further include: obtaining the phase status of signal lights at all intersections in the road network; for each intersection, when the phase status of the signal lights at the intersection indicates the start of a red light signal, obtaining the first queue length and the first acquisition time of each lane corresponding to the intersection; when the phase status of the signal lights indicates the start of a green light signal, obtaining the second queue length and the second acquisition time, and the headway data of each lane corresponding to the intersection, and determining the average headway data of each lane after the phase status of the signal lights indicates the start of a red light signal again; wherein the headway information includes the headway data and the average headway data of each lane; for each lane corresponding to the intersection, determining the speed increase information of the lane in the second time period according to the first queue length, the first acquisition time, the second queue length and the second acquisition time of the lane; and storing the speed increase information of all lanes corresponding to all intersections in a database.

[0130] Among them, the signal light phase state can refer to the current state of the traffic light, including different phases such as red light, green light and yellow light, which is used to indicate the control state of traffic flow. The queue length can refer to the number of vehicles or queue length waiting to pass when the red light starts or the green light starts on each lane; the acquisition time is the specific time point at which the corresponding queue length is recorded. For example, a queue increment accumulator is set in each lane of each intersection. This accumulator can be a first-in-first-out queue structure. Whenever the red light starts at the intersection, the initial queue and time point are recorded. When the green light starts at the corresponding intersection, the queue and time point are recorded again.

[0131] The headway data may refer to the time interval between vehicles, which is collected only during the green light period; the headway average is the average of these intervals, which is calculated at the end of each green light cycle and is used to measure the density and speed of traffic flow. The second time period may refer to the time period corresponding to the calculation of the growth rate information based on the first and second queue lengths and their acquisition time, which is used to store in the database, for example, the growth rate information is stored once every minute.

[0132] In one example, Figure 5 The flowchart of the red light queue increment and headway collection shown in the figure obtains the current control information of the signal control machine (i.e., the phase state of the signal light); determines whether the current control information is the red light information of a certain intersection; if it is a red light, stops collecting the green light headway data, determines the average headway value and stores it in the headway database, and records the current first queue length and first acquisition time If it is a green light, start collecting the headway data of the green light vehicles and record the current length of the second queue. and the second acquisition time ; If the first queue length and the first acquisition time exist, then Store it in the current lane queue change cache; clean it according to the existing data and store it in the red light queue growth rate database.

[0133] Optionally, the structure of the red light queue speed increase database is: lane identification - lane_id, the start time of the second time period - timepoint_start, the end time of the second time period - timepoint_end. The setting of the trigger in the database is optional, and the queue data structure can be used to store the queue speed increase data. For example, 259200 (60 minutes × 24 hours × 180 days) is the actual set value, that is, the red light queue speed increase information related data for every minute in 180 days is stored. If the data of a lane exceeds 259200, the data record with the earliest timestamp of the lane is automatically deleted to retain the latest data and save system resources. Taking the storage of speed increase information once every minute as an example, in the red light queue speed increase database, a storage record includes: lane identification, the start time and end time corresponding to one minute, and the speed increase information of the corresponding intersection.

[0134] By acquiring and analyzing traffic light phase status, queue length and headway data in real time, the speed increase information of each lane can be accurately calculated and stored for subsequent traffic signal optimization, improving the understanding of traffic flow trends.

[0135] On the basis of the above embodiment, the method for determining the speed increase information of the lane in the second time period according to the first queue length, the first acquisition time, the second queue length and the second acquisition time of the lane may include: determining the middle moment between the first acquisition time and the second acquisition time of the lane; if the middle moment is within the second time period, determining the speed increase information of the lane in the second time period based on the first difference between the second queue length and the first queue length of the lane, and the second difference between the second acquisition time and the first acquisition time; if the middle moment is after the second time period, determining the third speed increase information of the lane based on the first difference between the second queue length and the first queue length of the lane, and the second difference between the second acquisition time and the first acquisition time; determining the speed increase information of the lane in the second time period based on the third speed increase information and the fourth speed increase information, wherein the fourth speed increase information is the speed increase information of the lane in the second historical time period stored in the database, and the second historical time period is the second time period corresponding to the storage of the speed increase information of the lane before the first acquisition time.

[0136] Furthermore, for a certain intersection, when adding the speed information of a lane in the second time period (target storage record) to the database, it is necessary to obtain the start time and end time of the latest storage record of the lane stored in the database corresponding to one minute (i.e., the second historical time period), and use the end time of the storage record corresponding to one minute as the target start time corresponding to one minute of the target storage record, and then add 1 minute backward based on the target start time to obtain the target end time corresponding to one minute of the target storage record;

[0137] If the middle moment Between the target start time and the target end time (in the second time period), the speed information of the lane in the second time period satisfies , the target storage record includes: the corresponding lane mark, the target start time and the target end time, ;

[0138] If the middle moment After the second time period, As the third speed information of the lane, As a second storage record ,in, As the second storage record corresponds to the start time of one minute , As the second storage record corresponds to the end time of one minute , As speed information ;

[0139] The latest storage record of the lane stored in the database is used as the first storage record ,in, is the fourth speed increase information;

[0140] Assume missing time point For and If the data between , the speed information of the lane in the second time period satisfies , the target storage record includes: the corresponding lane mark, the target start time and the target end time, .

[0141] By introducing intermediate moment judgments and combining historical growth rate information, the traffic flow changes in the lane within a specific time period can be more accurately reflected. At the same time, the differences within and outside the time period and the influence of historical data are taken into account, which further improves the accuracy and reliability of the growth rate information, thereby optimizing the decision-making process of traffic signal control.

[0142] On the basis of the above embodiments, the method described in S102 for determining the target phase to be controlled in the intersection to be controlled according to the current queuing information, the phase set and the lane index set may include: determining the lane flow direction information according to the phase set and the lane index set, the lane flow direction information being used to indicate the mapping relationship between the upstream lane and the downstream lane in the intersection to be controlled; determining the queue volume difference between the upstream lane queue length and the downstream lane queue length corresponding to each phase according to the current queuing information and the lane flow direction information; and determining the target phase in the phase set based on the queue volume difference corresponding to all phases.

[0143] The lane flow information may refer to a data structure used to represent the mapping relationship between the upstream lane and the downstream lane in the intersection to be controlled. It indicates the traffic flow direction of each lane, thereby helping to identify which lanes are related to each other. The queue volume difference may refer to the difference between the queue length of the upstream lane and the queue length of the downstream lane corresponding to each phase.

[0144] Prior to this embodiment, it is necessary to construct an initialization queue quantity difference vector , whose value is filled with 0 for initialization, indicating The queue volume difference of each phase is calculated, and then for each phase of the i-th intersection at this moment The queue length of the upstream lane and the queue length in the downstream lane .

[0145] In this embodiment, according to the phase set (for With phase The index set corresponding to the lane) and the lane index set (i is the intersection to be controlled), determine the lane flow information (i.e. upstream lane To the downstream lane mapping relationship); according to the current queue information ( ) and lane flow information to calculate the upstream queue length for each phase and downstream queue length , and the difference in queue volume between the upstream and downstream lanes, and fill in In Chinese, that is:

[0146] ;

[0147] For filling values Sort the values ​​(from large to small) to get a sorted phase index set ; Take the first phase in the ranking as the target phase of the final decision output ( ).

[0148] By introducing lane flow information and queue volume difference calculation, it is possible to more accurately identify phases of unbalanced traffic flow, thereby optimizing signal control; by giving priority to adjusting phases with the largest queue volume difference, the system can effectively alleviate traffic congestion and improve traffic flow efficiency, thereby enhancing the intelligence and responsiveness of traffic signal control, and further improving the overall efficiency and adaptability of the traffic management system.

[0149] On the basis of the above embodiment, the method described in S104 for controlling the signal at the intersection to be controlled based on the target phase and the headway information and target queue information of all lanes corresponding to the target phase may include:

[0150] For each lane corresponding to the target phase, based on the headway information of the lane, determine the starting time corresponding to the lane and the number of vehicles passing within the starting time;

[0151] Determine the lane pass time based on target queue information, vehicle number, start time, and headway information;

[0152] Determine a target passing time of the target phase according to the passing times of all lanes corresponding to the target phase, and the maximum phase time and the minimum phase time of the target phase;

[0153] Based on the target phase and the target passing time of the target phase, the signal of the intersection to be controlled is controlled.

[0154] In this step, the starting time may refer to the time required for the vehicle to start moving and pass through the intersection to be controlled after the signal light turns green. The number of vehicles may be the number of vehicles that can pass through the intersection to be controlled within the starting time, which is usually affected by the headway and traffic flow density of the corresponding lane. The passing time may refer to the length of time that the lane allows vehicles to pass during the target phase. The maximum phase time and the minimum phase time are the longest and shortest passing times allowed in the target phase, respectively, to ensure the safety and efficiency of signal control. The target passing time is the actual passing time determined after considering the passing time, maximum phase time and minimum phase time of all lanes.

[0155] In some examples, a negative exponential distribution model can be used to calculate the start time, assuming that at a certain moment, 90% of the headway has reached the minimum headway. (the average headway time in the headway time information), the vehicle volume from this moment onwards will follow the minimum headway time. In the negative exponential distribution, the headway time The probability distribution function of is:

[0156] ;

[0157] There is a starting time The headway distance has a probability of reaching the minimum headway distance of 90%:

[0158] ;

[0159] because , so the starting time finally solved is also in vector form , represents the starting time of each lane. The number of vehicles passing during the starting time for:

[0160] ;

[0161] in, is the starting headway information, that is, the headway data in the headway information. .

[0162] Queue Passing Time Function , which means that by inputting a five-tuple Calculate the travel time:

[0163] ;

[0164] Finally, for the target phase , its travel time is the maximum value of the travel time corresponding to all lanes, that is:

[0165] ;

[0166] in, is the maximum value function.

[0167] In order to standardize the phase time and push it, it is also necessary to calculate the maximum phase time corresponding to the target phase. and minimum phase time , normalized phase time:

[0168] ;

[0169] in, , Binary operators for minimum and maximum values.

[0170] Finally, the decision results Push to the control end, such as the signal control machine / signal control platform, to control the signal changes at the intersection to be controlled.

[0171] Optionally, if the phase decision of a whole cycle is required to be provided at one time due to equipment limitation or control end limitation, the lead time for the advance decision :

[0172] 1) First, based on the method of the above embodiment, a sorted phase index set is obtained: ;

[0173] 2) Then, referring to the method of the above embodiment, we obtain , , , , ,as follows:

[0174] ;

[0175] in, A time list representing the phase decisions of the entire cycle;

[0176] 3) Finally, Push to the control end to control the signal changes at the intersection to be controlled.

[0177] By analyzing the headway information and queue information, the travel time of each lane can be calculated more accurately, thereby optimizing the signal control of the target phase; by setting a reasonable target travel time, the system can effectively reduce vehicle waiting time and further improve traffic flow efficiency.

[0178] In one example, the dynamic signal control method for intersections provided by the embodiment of the present application may include 4 processes and four databases, wherein process 1-data monitoring process is used to obtain traffic data in the road network (collecting many types of data including queue length and number of queued vehicles); process 2-is used to collect red light queue increments and (green light) headway time, such as Figure 5 As shown; Process 3-Lane correlation parameter calculation process, such as Figure 2 As shown; Process 4-decision-making process, that is, the process that triggers the execution signal control process, including active decision-making type and passive decision-making type, such as Figure 3 shown.

[0179] Furthermore, the traffic data in the road network is monitored and collected through process 1 and stored in the traffic status database. The red light queue increment and (green light) headway are collected through process 2, that is, the signal light phase state of the traffic signal controller at the corresponding time is obtained according to the acquisition time period set in advance or the received data acquisition request; if the signal light phase state changes to the red light state, the headway data collection is stopped, the headway average value is determined and stored in the headway database; and the current first queue length and first acquisition time are recorded; if the signal light phase state changes to the green light state, the headway data collection starts, and the current second queue length and second acquisition time are recorded; if the first queue length and first time recorded in this collection task already exist, (first acquisition time, second acquisition time, first queue length, second queue length) are stored in the current lane queue change cache, if not, it is recorded when the signal light phase state changes to the red light state; the existing data in the cache is cleaned (including removing interference data and calculating the growth rate information) and stored in the red light queue growth rate database.

[0180] Set the correlation matrix calculation period through process 3 , set the road network adjacency matrix; every time (otherwise wait), by sending a request to obtain the past from the red light queue speed database stored in process 2 The growth rate of red light queues in all lanes within a time period; the growth rate of each lane in the road network and all historical data of other lanes constitute a data set; the lane correlation matrix and offset are calculated using the elastic net regression method and stored in the lane correlation database.

[0181] The signal control process is executed through process 4. If it is an active decision-making type, traverse each intersection in the road network; determine whether a decision is needed at the current time point (the length of time from the preset signal decision time to the end of the green light is greater than or equal to the length of the green light countdown itself, or the current moment is about to count down); if not, update the current time point and continue to traverse each intersection; if yes, perform data preparation (i.e. extract the multi-dimensional data of the corresponding intersection from the 4 databases stored in processes 1, 2, and 3); follow the decision calculation formula to determine the target phase of the final decision output and the target travel time corresponding to the target phase. For example, initialize the queue volume difference vector, calculate the upstream queue length and downstream queue length and queue volume difference of each phase according to the current queue information and lane flow information, fill and sort the vector to select the target phase of the final decision output (the one with the largest numerical sorting); determine the third time according to the speed increase information subset The sum of the speed information in the interval; the second speed information is determined according to the correlation parameter and the first speed information; the target queue information is determined according to the sum of the speed information, the second speed information, the current queue information and the signal decision time; the starting time is calculated using the negative exponential distribution model according to the minimum headway, and the number of vehicles passing within the starting time is determined according to the starting time, the minimum headway and the starting headway; the passing time of each lane is calculated according to the target queue information, the number of vehicles, the starting time, the minimum headway and the starting headway, and the maximum passing time of all lanes in the target phase is selected as the passing time of the phase; the passing time of the phase is adjusted according to the maximum phase time and the minimum phase time corresponding to the target phase to obtain the target passing time; the target phase and the corresponding target passing time are pushed to the signal control machine / signal control platform to control the signal change of the intersection. If it is a passive decision type, the data preparation is performed by receiving the decision request of the signal control; the result is obtained by following the decision calculation formula; the result is pushed to the signal control machine / signal control platform (same as the active decision type).

[0182] Therefore, the dynamic signal control method for intersections provided by the embodiment of the present application records and processes the vehicle queue data within the traffic light cycle by setting a lane queue increment accumulator and a standardized database; uses historical data to build a model, and fits and corrects the data through elastic net regression to achieve real-time dynamic optimization control of traffic flows at multiple intersections and improve traffic efficiency. Further effects may include: 1) Continuous and real-time traffic data processing and optimization. By collecting traffic flow data in real time and dynamically updating queue growth rate information using a first-in-first-out queue structure, the problem of difficulty in continuously obtaining traffic data is effectively solved, which not only improves the real-time and accuracy of the data, but also enables the traffic control system to quickly respond to actual traffic changes, thereby managing traffic flow more effectively. 2) Accurate traffic decision support. By introducing the red light queue growth rate correlation matrix and the road network connectivity adjacency matrix, vehicle collaborative control between multiple intersections becomes possible. By considering the impact of vehicle flow at multiple intersections, the system can more accurately predict and adjust the signal light changes at each intersection, optimize traffic flow, and reduce congestion. This method is particularly suitable for areas with large traffic volume and many intersections in cities, and can significantly improve the overall efficiency of the transportation system. 3) Smooth transition from transition design to full adaptive control. By retaining the concept of cycle and gradually introducing data-driven decision-making mechanism, a robust transition is provided for the future fully automated traffic control system. This design not only ensures the stable operation of the current system, but also provides data support and technical verification for the future upgrade of the system, reducing the risk and cost of technology upgrade. 4) Improve the scalability and flexibility of the traffic system. Through modular design, it allows flexible adjustment and expansion under different traffic scenarios and requirements. For example, through variant design, the control strategy can be adjusted according to the actual hardware and platform limitations, so that the technical solution can be widely applied to traffic networks of different sizes and types. Based on the method corresponding to the above embodiment, it also solves the problems of difficulty in obtaining traffic data in the prior art (through real-time data collection and processing mechanism, traffic status information is continuously updated to ensure the timeliness and accuracy of decision-making basis), inaccurate advance decision-making (by introducing correlation analysis between lanes and collaborative control of multiple intersections, the accuracy and reliability of decision-making are improved), and the inability to meet the needs of transition to full adaptive control (by gradually introducing data-driven elements and retaining the concept of periodicity, a smooth transition to a more advanced control strategy is made, laying the foundation for the implementation of the full adaptive control method).

[0183] Figure 6 The schematic diagram of the structure of the dynamic signal control device for the intersection provided in this application is as follows: Figure 6 As shown, the intersection dynamic signal control device 60 provided in this embodiment includes:

[0184] The acquisition module 601 is used to acquire a phase set, a lane index set, and current queue information, a speed increase information set, and headway information of each lane corresponding to the intersection to be controlled in the road network, wherein the speed increase information set includes speed increase information of each lane of the intersection to be controlled in multiple time periods before the current time stored in the database, wherein each speed increase information is determined according to the queue length and collection time of the lane collected at the intersection to be controlled in the corresponding time period, and is used to indicate the change rate of the queue length of the corresponding lane at the intersection to be controlled during the red light process;

[0185] A determination module 602, configured to determine a target phase to be controlled in the intersection to be controlled according to current queuing information, a phase set and a lane index set;

[0186] An adjustment module 603 is used to adjust the current queue information of each lane corresponding to the target phase according to the speed increase information set of the lane to obtain the target queue information;

[0187] The control module 604 is used to control the signal of the intersection to be controlled based on the target phase and the headway time information and target queue information of all lanes corresponding to the target phase.

[0188] In a possible implementation, the adjustment module 603 may also be used to:

[0189] The lane correlation parameters are determined based on the historical growth rate information of all lanes in the road network, where the historical growth rate information of each lane is the growth rate information of the corresponding lane stored in the database before the current time, and the lane correlation parameters are used to indicate the traffic impact relationship between lanes; based on the lane growth rate information set and the lane correlation parameters, the current queue information of the lane is adjusted to obtain the target queue information.

[0190] In a possible implementation, the adjustment module 603 may also be used to:

[0191] Acquire multiple historical growth rate information of each lane in the road network within a preset first time period; construct a sample set based on the multiple historical growth rate information of all lanes, wherein each sample includes a second time period in the first time period and the historical growth rate information of all lanes in the corresponding second time period; construct a road network adjacency matrix based on the adjacency relationship in the road network, and the road network adjacency matrix is ​​used to indicate the connectivity between every two lanes in the road network; based on the sample set and the road network adjacency matrix, adjust the preset correlation parameters to obtain the lane correlation parameters.

[0192] In a possible implementation, the adjustment module 603 may also be used to:

[0193] According to a subset of speed increase information in a speed increase information set of a lane, a sum of speed increase information is obtained, wherein the speed increase information subset includes a plurality of speed increase information obtained by sampling the speed increase information of the lane stored in a database within a preset third time period; according to a lane correlation parameter and a first speed increase information in a speed increase information set of the lane, a second speed increase information is determined, wherein the first speed increase information is the speed increase information of the lane stored in the database in a second time period before the current time; according to the sum of speed increase information, the second speed increase information, the current queuing information and the signal decision time, the target queuing information is determined, wherein the signal decision time is the time for triggering the execution of the steps of obtaining the phase set, the lane index set, and the current queuing information, the speed increase information set and the headway information of each lane corresponding to the intersection to be controlled in the road network.

[0194] In a possible implementation, the acquisition module 601 may also be used to:

[0195] Obtain a decision request for signal control, the decision request includes a signal decision time set by the user and an intersection to be controlled; accordingly, when obtaining a phase set, a lane index set, and current queue information, a speed increase information set, and vehicle headway information of each lane corresponding to the intersection to be controlled in the road network, the intersection to be controlled is the intersection to be controlled in the decision request; or, determine the green light countdown length corresponding to all intersections in the road network, and the green light end time corresponding to each intersection at the current time; for each intersection, if the time length from the current time to the green light end time corresponding to the intersection is greater than or equal to the green light countdown length corresponding to the intersection, then the current time is used as the signal decision time, and the intersection is used as the intersection to be controlled.

[0196] In a possible implementation, the acquisition module 601 may also be used to:

[0197] The signal light phase status of all intersections in the road network is obtained; for each intersection, when the signal light phase status of the intersection indicates that the signal light starts to turn red, the first queue length and the first acquisition time of each lane corresponding to the intersection are obtained; when the signal light phase status indicates that the signal light starts to turn green, the second queue length and the second acquisition time and the headway data of each lane corresponding to the intersection are obtained, and after the signal light phase status indicates that the signal light starts to turn red again, the headway average of each lane is determined according to the headway data; wherein the headway information includes the headway data and the headway average of each lane; for each lane corresponding to the intersection, the speed increase information of the lane in the second time period is determined according to the first queue length, the first acquisition time, the second queue length and the second acquisition time of the lane; the speed increase information of all lanes corresponding to all intersections is stored in a database.

[0198] In a possible implementation, the acquisition module 601 may also be used to:

[0199] Determine the middle moment between the first acquisition time and the second acquisition time of the lane; if the middle moment is within the second time period, determine the speed increase information of the lane within the second time period based on the first difference between the second queue length and the first queue length of the lane and the second difference between the second acquisition time and the first acquisition time; if the middle moment is after the second time period, determine the third speed increase information of the lane based on the first difference between the second queue length and the first queue length of the lane and the second difference between the second acquisition time and the first acquisition time; determine the speed increase information of the lane within the second time period based on the third speed increase information and the fourth speed increase information, wherein the fourth speed increase information is the speed increase information of the lane within the second historical time period stored in the database, and the second historical time period is the second time period corresponding to the time when the speed increase information of the lane was stored before the first acquisition time.

[0200] In a possible implementation, the determination module 602 may also be used to:

[0201] According to the phase set and the lane index set, the lane flow direction information is determined, and the lane flow direction information is used to indicate the mapping relationship between the upstream lane and the downstream lane in the intersection to be controlled; according to the current queuing information and the lane flow direction information, the queue volume difference between the upstream lane queue length and the downstream lane queue length corresponding to each phase is determined; based on the queue volume difference corresponding to all phases, the target phase in the phase set is determined.

[0202] In a possible implementation, the control module 604 may also be used to:

[0203] For each lane corresponding to the target phase, the starting time corresponding to the lane and the number of vehicles passing within the starting time are determined based on the headway time information of the lane; the passage time of the lane is determined according to the target queuing information, the number of vehicles, the starting time and the headway time information; the target passage time of the target phase is determined according to the passage time of all lanes corresponding to the target phase and the maximum phase time and minimum phase time of the target phase; based on the target phase and the target passage time of the target phase, the signal of the controlled intersection is controlled.

[0204] The dynamic signal control device for intersections provided in this embodiment can execute the method provided in the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be described in detail here.

[0205] Figure 7 This is a schematic diagram of the structure of the electronic device provided in this application. Figure 7As shown, the electronic device 70 provided in this embodiment includes: at least one processor 701 and a memory 702. Optionally, the device 70 further includes a communication component 703. The processor 701, the memory 702 and the communication component 703 are connected via a bus 704.

[0206] In a specific implementation process, at least one processor 701 executes the computer-executable instructions stored in the memory 702, so that at least one processor 701 executes the above method.

[0207] The specific implementation process of the processor 701 can be found in the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be repeated here.

[0208] In the above embodiments, it should be understood that the processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the invention can be directly implemented as a hardware processor, or can be implemented by a combination of hardware and software modules in the processor.

[0209] The memory may include a high-speed memory (Random Access Memory, RAM), and may also include a non-volatile memory (NVM), such as at least one disk storage.

[0210] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, the bus in the drawings of this application is not limited to only one bus or one type of bus.

[0211] The present application also provides a computer program product, including a computer program, which implements the above method when executed by a processor.

[0212] The present application also provides a computer-readable storage medium, in which computer-executable instructions are stored. When a processor executes the computer-executable instructions, the above method is implemented.

[0213] The above-mentioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The readable storage medium can be any available medium that can be accessed by a general or special-purpose computer.

[0214] An exemplary readable storage medium is coupled to a processor so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can be located in an application specific integrated circuit (Application Specific Integrated Circuits, referred to as: ASIC). Of course, the processor and the readable storage medium can also exist in the device as discrete components.

[0215] The division of units is only a logical function division, and there may be other divisions in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0216] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0217] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0218] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0219] Those skilled in the art can understand that all or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk and other media that can store program codes.

[0220] Finally, it should be noted that those skilled in the art will readily conceive of other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. The present invention is intended to cover any variations, uses or adaptations of the present invention, which follow the general principles of the present invention and include common knowledge or customary technical means in the art not disclosed by the present invention, are not limited to the precise structure described above and shown in the drawings, and may be modified and changed in various ways without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. A method for controlling dynamic signals at intersections, characterized in that: include: Acquire a phase set and a lane index set of an intersection to be controlled in a road network, and current queue information, a speed increase information set and headway information of each lane corresponding to the intersection to be controlled, wherein the speed increase information set includes speed increase information of each lane of the intersection to be controlled in multiple time periods before the current time stored in a database, wherein each speed increase information is determined according to the queue length and collection time of the lane collected at the intersection to be controlled in the corresponding time period, and is used to indicate the change rate of the queue length of the corresponding lane at the intersection to be controlled during a red light process; Determining a target phase to be controlled in the intersection to be controlled according to the current queuing information, the phase set and the lane index set; For each lane corresponding to the target phase, according to the speed increase information set of the lane, current queue information of the lane is adjusted to obtain target queue information; Based on the target phase, and the headway time information and target queue information of all lanes corresponding to the target phase, the signal of the intersection to be controlled is controlled.

2. The method according to claim 1, characterized in that The step of adjusting the current queue information of the lane according to the speed increase information set of the lane to obtain the target queue information includes: Determining a lane correlation parameter according to historical speed information of all lanes in the road network, wherein the historical speed information of each lane is the speed information of the corresponding lane stored in the database before the current time, and the lane correlation parameter is used to indicate the traffic impact relationship between the lanes; According to the speed increase information set of the lane and the lane correlation parameter, the current queue information of the lane is adjusted to obtain the target queue information.

3. The method according to claim 2, characterized in that The determining of lane correlation parameters according to the historical speed information of all lanes in the road network includes: Acquire multiple historical speed increase information of each lane in the road network within a preset first time period; Constructing a sample set according to multiple historical speed increase information of all lanes, wherein each sample includes a second time period in the first time period and historical speed increase information of all lanes in the corresponding second time period; Constructing a road network adjacency matrix according to the adjacency relationship in the road network, wherein the road network adjacency matrix is ​​used to indicate the connectivity between every two lanes in the road network; Based on the sample set and the road network adjacency matrix, the preset correlation parameters are adjusted to obtain lane correlation parameters.

4. The method according to claim 2 or 3, characterized in that: The adjusting the current queue information of the lane according to the speed increase information set of the lane and the lane correlation parameter to obtain the target queue information includes: obtaining a sum of the speed information according to a speed information subset in the speed information set of the lane, wherein the speed information subset includes a plurality of speed information obtained by sampling the speed information of the lane stored in the database within a preset third time period; Determine second speed information according to the lane correlation parameter and first speed information in the speed information set of the lane, where the first speed information is speed information of the lane in a second time period before the current time stored in the database; The target queuing information is determined according to the sum of the speed increase information, the second speed increase information, the current queuing information and the signal decision time, wherein the signal decision time is the time for triggering the execution of the steps of obtaining the phase set of the intersection to be controlled in the road network, the lane index set, and the current queuing information, speed increase information set and vehicle headway information of each lane corresponding to the intersection to be controlled.

5. The method according to claim 1, characterized in that Before acquiring the phase set, lane index set, and current queue information, speed increase information set, and headway information of each lane corresponding to the intersection to be controlled in the road network, the method further includes: Obtain a decision request for signal control, wherein the decision request includes a signal decision time set by a user and an intersection to be controlled; Correspondingly, when acquiring the phase set, lane index set, and current queue information, speed increase information set, and headway information of each lane corresponding to the intersection to be controlled in the road network, the intersection to be controlled is the intersection to be controlled in the decision request; or, Determine the green light countdown length corresponding to all intersections in the road network, and the green light end time corresponding to each intersection at the current time; For each intersection, if the time length from the current time to the green light end time corresponding to the intersection is greater than or equal to the green light countdown length corresponding to the intersection, the current time is used as the signal decision time and the intersection is used as the intersection to be controlled.

6. The method according to claim 1, characterized in that The method further comprises: Obtaining the phase status of traffic lights at all intersections in the road network; For each intersection, when the phase state of the traffic light at the intersection indicates that the signal light turns red, obtain a first queue length and a first acquisition time for each lane corresponding to the intersection; When the phase state of the signal light indicates that the signal light turns green, the second queue length, the second acquisition time and the headway time data of each lane corresponding to the intersection are obtained, and after the phase state of the signal light indicates that the signal light turns red again, the headway time average of each lane is determined according to the headway time data; wherein the headway time information includes the headway time data and the headway time average of each lane; For each lane corresponding to the intersection, determining the speed increase information of the lane in a second time period according to the first queue length, the first acquisition time, the second queue length and the second acquisition time of the lane; The speed increase information of all lanes corresponding to all intersections is stored in the database.

7. The method according to claim 6, characterized in that The determining, according to the first queue length, the first acquisition time, the second queue length, and the second acquisition time of the lane, the speed increase information of the lane in the second time period includes: determining an intermediate time between a first acquisition time and a second acquisition time of the lane; If the intermediate moment is within the second time period, determining the speed increase information of the lane within the second time period based on a first difference between a second queue length and a first queue length of the lane and a second difference between a second acquisition time and a first acquisition time; If the intermediate moment is after the second time period, determining third speed increase information of the lane based on a first difference between a second queue length and a first queue length of the lane and a second difference between a second acquisition time and a first acquisition time; Based on the third speed increase information and the fourth speed increase information, the speed increase information of the lane within the second time period is determined, wherein the fourth speed increase information is the speed increase information of the lane within the second historical time period stored in the database, and the second historical time period is the second time period corresponding to when the speed increase information of the lane was stored before the first acquisition time.

8. The method according to claim 1, characterized in that The determining, according to the current queuing information, the phase set and the lane index set, a target phase to be controlled in the intersection to be controlled, comprises: Determining lane flow direction information according to the phase set and the lane index set, wherein the lane flow direction information is used to indicate a mapping relationship between an upstream lane and a downstream lane in the intersection to be controlled; Determine, according to the current queue information and the lane flow direction information, a queue volume difference between an upstream lane queue length and a downstream lane queue length corresponding to each phase; A target phase in the phase set is determined based on queue amount differences corresponding to all phases.

9. The method according to claim 1, characterized in that: The controlling of the signal of the to-be-controlled intersection based on the target phase and the headway time information and target queue information of all lanes corresponding to the target phase includes: For each lane corresponding to the target phase, determining a starting time corresponding to the lane and a number of vehicles passing within the starting time based on headway information of the lane; Determining the lane passing time according to the target queue information, the number of vehicles, the starting time and the headway information; Determining a target passing time of the target phase according to the passing times of all lanes corresponding to the target phase, and the maximum phase time and the minimum phase time of the target phase; Based on the target phase and the target passing time of the target phase, the signal of the intersection to be controlled is controlled.

10. A dynamic signal control device for an intersection, characterized in that: include: An acquisition module, used to acquire a phase set, a lane index set, and current queue information, a speed increase information set, and headway information of each lane corresponding to the intersection to be controlled in the road network, wherein the speed increase information set includes speed increase information of each lane of the intersection to be controlled in multiple time periods before the current time stored in a database, wherein each speed increase information is determined according to the lane queue length and collection time collected at the intersection to be controlled in the corresponding time period, and is used to indicate the change rate of the queue length of the corresponding lane at the intersection to be controlled during the red light process; A determination module, configured to determine a target phase to be controlled in the intersection to be controlled according to the current queuing information, the phase set and the lane index set; an adjustment module, configured to adjust, for each lane corresponding to the target phase, the current queue information of the lane according to the speed increase information set of the lane, to obtain the target queue information; A control module is used to control the signal of the intersection to be controlled based on the target phase and the headway time information and target queue information of all lanes corresponding to the target phase.

Images