Road congestion event identification method, electronic device, storage medium and computer program product

Abstract

The invention discloses a road congestion event identification method, electronic equipment, a storage medium and a computer program product. The method comprises the following steps: in a recognition period, obtaining vehicle sensing data and signal lamp data of a current lane of a current road section within a preset time before the recognition period; determining a corresponding congestion judgment logic according to the signal lamp data; and judging the road vehicle sensing data by adopting the congestion judgment logic, and determining and outputting the congestion information of the congestion event. According to the method, the running state of the vehicle is considered and serves as the judgment basis of congestion, the signal lamp data is combined, rule analysis and judgment are more comprehensively carried out, the data size is more comprehensive and accurate, and the accuracy of the intersection congestion degree is further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle-related technologies, and in particular to a road congestion event recognition method, electronic equipment, storage medium and computer program product. Background Art

[0002] With the establishment of national autonomous driving demonstration zones, the installation of roadside equipment, the access to perception data, and the analysis of traffic data based on big data, the autonomous driving demonstration zones will be given greater value.

[0003] However, how to use complex multi-source data to judge and evaluate traffic congestion is an important issue. In different standards, many traffic indicators are mentioned, as well as how to use traffic indicators to comprehensively judge road congestion. For example, GA / T115-2020 stipulates that the maximum average vehicle delay or maximum queue time index is used to define traffic congestion levels in different intervals. However, in actual engineering implementation, there is no strict standard definition in the industry on how to use real traffic data to calculate the maximum average vehicle delay or maximum queue time index and how to determine the interval threshold.

[0004] Patent document 1 (CN114023073A) proposes a method for real-time monitoring and congestion prediction of highway traffic conditions based on vehicle behavior analysis. By obtaining real-time vehicle data and road network structure data, the delay time ratio is calculated. The corresponding relationship between the delay time ratio and the urban traffic operation index and the traffic operation status level is combined to locate the different traffic operation conditions of different sections, which are divided into five levels: smooth, basically smooth, light congestion, moderate congestion and severe congestion.

[0005] Patent document 2 (CN111754780A) obtains the degree of congestion of a road section by the distance between vehicles in the road section within a unit time period. First, the inherent data such as the length of the monitored road section, the unit time length and the average vehicle length are determined. Then, a counter is set to record the vehicles entering and exiting the road section, and then the distance between vehicles is calculated according to the formula. Based on the relationship between the safe distance and the classification of the traffic operation level of the road section, the operation status is finally determined, which is divided into four levels: severe congestion, mild congestion, basically unblocked and unblocked.

[0006] However, the existing technology does not refer to traffic light data when judging congestion information, and the judgment is not accurate. Summary of the invention

[0007] Based on this, it is necessary to provide a road congestion event identification method, electronic device, storage medium and computer program product to address the technical problem that the existing technology does not refer to traffic light data and the judgment of congestion information is inaccurate.

[0008] The present invention provides a method for identifying a road congestion event, comprising:

[0009] During the recognition cycle, the vehicle perception data and signal light data of the current lane of the current road section within a preset time before the recognition cycle are obtained;

[0010] Determine corresponding congestion judgment logic according to the traffic light data;

[0011] The congestion judgment logic is used to judge the road vehicle perception data, and congestion information of the congestion event is determined and output.

[0012] Further, determining the corresponding congestion judgment logic according to the traffic light data includes:

[0013] When there is a traffic light control on the current road, determining the light color of the traffic light within the recognition cycle;

[0014] If all the traffic lights of the current lane are green during the identification period, determining that the corresponding congestion judgment logic is the green light congestion judgment logic;

[0015] If the traffic lights of the current lane are all red and / or yellow during the identification period, determining that the corresponding congestion judgment logic is a red light congestion judgment logic;

[0016] If the traffic lights of the current lane include green lights and red lights within the identification cycle, then the proportion of light color duration within the identification cycle is determined; if the proportion of green light duration within the identification cycle is greater than the sum of the proportions of red and yellow light durations, then the corresponding congestion judgment logic is determined to be the green light congestion judgment logic; if the proportion of green light duration within the identification cycle is less than the sum of the proportions of red and yellow light durations, then the corresponding congestion judgment logic is determined to be the red light congestion judgment logic.

[0017] Furthermore, the using the congestion judgment logic to judge the road vehicle perception data, determine and output congestion information of the congestion event, includes:

[0018] When the congestion judgment logic is the green light congestion judgment logic, calculating the duration of the green light at the beginning of the recognition cycle;

[0019] When the duration is less than the preset time threshold, the average speed of the first vehicle before the stop line in the current lane within the identification period is calculated according to the vehicle perception data; if the average speed of the first vehicle before the stop line within the identification period is less than or equal to the green light low speed threshold, the first vehicle before the stop line is judged to be at a low speed; otherwise, it is judged that the lane is not congested, the congestion information of the congestion event is output as no congestion, and the process ends;

[0020] When the duration is greater than or equal to a preset time threshold or when it is determined that the first vehicle before the stop line is at a low speed, all vehicles in the current lane within the identification period whose speed is less than or equal to the green light low speed threshold are selected as low-speed vehicles, and spatial clustering is performed based on the positions of all the low-speed vehicles in the current lane;

[0021] When the number of clusters of the clustering result of the spatial clustering is greater than 0, the clustering results are traversed, and for each cluster, the position of the first vehicle in the cluster is selected as the congestion starting point, the position of the last vehicle in the cluster is selected as the congestion end point, the length along the road from the congestion starting point to the congestion end point is calculated as the congestion length of the cluster, and the average of the average speeds of each vehicle in the cluster is calculated as the congestion speed of the cluster;

[0022] Determining a congestion level of each cluster according to the congestion length of each cluster and / or the congestion speed of the cluster;

[0023] The congestion level, congestion starting point, congestion end point, congestion length and / or congestion speed of each cluster are used as congestion information of the congestion event and output.

[0024] Furthermore, the using the congestion judgment logic to judge the road vehicle perception data, determine and output congestion information of the congestion event, includes:

[0025] When the congestion judgment logic is a red light congestion judgment logic, red light congestion information calculation is performed, and congestion information of the congestion event is output;

[0026] The red light congestion information calculation includes: selecting all vehicles in the current lane within the identification period whose speed is less than or equal to the red light low speed threshold as low-speed vehicles, and performing spatial clustering based on the positions of all the low-speed vehicles in the current lane;

[0027] When the number of clusters of the clustering result of the spatial clustering is greater than 0, the clustering results are traversed, and for each cluster, the position of the first vehicle in the cluster is selected as the congestion starting point, the position of the last vehicle in the cluster is selected as the congestion end point, the length along the road from the congestion starting point to the congestion end point is calculated as the congestion length of the cluster, and the average of the average speeds of each vehicle in the cluster is calculated as the congestion speed of the cluster;

[0028] Determine the congestion level of each cluster according to the congestion length of each cluster;

[0029] The congestion level, congestion starting point, congestion end point, congestion length and / or congestion speed of each cluster are used as congestion information.

[0030] Further, the using the congestion judgment logic to judge the road vehicle perception data, determine and output congestion information of the congestion event, also includes:

[0031] Determine the time range of the previous green light cycle of the current red light from the signal light data, and calculate the starting average speed of the first vehicle before the stop line on the current lane in the previous green light cycle within the starting time period of the previous green light cycle and the ending average speed within the ending time period of the previous green light cycle according to the vehicle perception data;

[0032] When the starting average speed is less than or equal to the red light low speed threshold, red light congestion information calculation is performed; when the starting average speed is greater than the red light low speed threshold, the ending average speed is determined; if the ending average speed is less than or equal to the red light low speed threshold, a flag is set, and then red light congestion information calculation is performed; otherwise, red light congestion information calculation is directly performed;

[0033] Determining the congestion level of each cluster according to the congestion length of each cluster includes: determining the congestion level of each cluster according to the congestion length of each cluster, and downgrading the congestion level if the flag is set.

[0034] Furthermore, the determining of the corresponding congestion judgment logic according to the traffic light data further includes:

[0035] When there is no traffic light control on the current road, the corresponding congestion judgment logic is determined to be a no-light congestion judgment logic.

[0036] Furthermore, the using the congestion judgment logic to judge the road vehicle perception data, determine and output congestion information of the congestion event, includes:

[0037] When the congestion judgment logic is a no-light congestion judgment logic, all vehicles in the current lane within the identification period whose speed is less than or equal to the no-signal control low-speed threshold are selected as low-speed vehicles, and spatial clustering is performed based on the positions of all the low-speed vehicles in the current lane;

[0038] When the number of clusters of the clustering result of the spatial clustering is greater than 0, the clustering results are traversed, and for each cluster, the position of the first vehicle in the cluster is selected as the congestion starting point, the position of the last vehicle in the cluster is selected as the congestion end point, the length along the road from the congestion starting point to the congestion end point is calculated as the congestion length of the cluster, and the average of the average speeds of each vehicle in the cluster is calculated as the congestion speed of the cluster;

[0039] Determining a congestion level of each cluster according to the congestion length of each cluster and / or the congestion speed of the cluster;

[0040] The congestion level, congestion starting point, congestion end point, congestion length and / or congestion speed of each cluster are used as congestion information of the congestion event and output.

[0041] The present invention provides an electronic device, comprising:

[0042] at least one processor; and,

[0043] a memory communicatively connected to at least one of the processors; wherein,

[0044] The memory stores instructions that can be executed by at least one of the processors. The instructions are executed by at least one of the processors to enable the at least one of the processors to execute the road congestion event identification method as described above.

[0045] The present invention provides a storage medium, wherein the storage medium stores computer instructions. When a computer executes the computer instructions, the storage medium is used to execute all steps of the road congestion event identification method as described above.

[0046] The present invention provides a computer program product, comprising a computer program / instruction, which implements the road congestion event identification method as described above when executed by a processor.

[0047] The present invention obtains the vehicle perception data and signal light data of the current lane of the current road section within the preset time before the identification cycle during the identification cycle, and determines the corresponding congestion judgment logic according to the signal light data, and then outputs the congestion information of the congestion event based on the determined congestion judgment logic. Therefore, the present invention not only considers the running state of the vehicle itself as the basis for judging congestion, but also combines the signal light data to conduct more comprehensive regularity analysis and judgment, and the data volume is more comprehensive and accurate, thereby improving the accuracy of road section congestion events. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a workflow diagram of a method for identifying road congestion events according to an embodiment of the present invention;

[0049] Figure 2 This is a workflow diagram of a road congestion event identification method according to another embodiment of the present invention;

[0050] Figure 3 This is a schematic diagram of the vehicle-road integration architecture of the best embodiment of the present invention;

[0051] Figure 4 This is a schematic diagram of data frequency reduction in the best embodiment of the present invention;

[0052] Figure 5 A schematic diagram of a congestion event identification process according to a preferred embodiment of the present invention;

[0053] Figure 6 A schematic diagram of the green light congestion judgment logic of the best embodiment of the present invention;

[0054] Figure 7 A schematic diagram of the red light congestion judgment logic of the best embodiment of the present invention;

[0055] Figure 8 A schematic diagram of the logic of judging congestion without lights according to the best embodiment of the present invention;

[0056] Fig. 9 A schematic diagram of clustering logic of the best embodiment of the present invention;

[0057] Fig.10 A schematic diagram of road section merging in the best embodiment of the present invention;

[0058] Fig.11 The figure is a schematic diagram of the hardware structure of an electronic device of the present invention. DETAILED DESCRIPTION

[0059] The specific embodiments of the present invention are further described below in conjunction with the accompanying drawings. The same components are represented by the same reference numerals. It should be noted that the words "front", "rear", "left", "right", "upper" and "lower" used in the following description refer to directions in the accompanying drawings, and the words "inner" and "outer" refer to directions toward or away from the geometric center of a specific component, respectively.

[0060] like Figure 1 The flowchart of a method for identifying a road congestion event according to an embodiment of the present invention includes:

[0061] Step S101, within the recognition cycle, obtaining vehicle perception data and signal light data of the current lane of the current road section within a preset time before the recognition cycle;

[0062] Step S102, determining corresponding congestion judgment logic according to the traffic light data;

[0063] Step S103: using the congestion judgment logic to judge the road vehicle perception data, determine and output congestion information of the congestion event.

[0064] Specifically, the present invention can be applied to electronic devices with processing capabilities, such as cloud servers.

[0065] Preferably, the road congestion event identification method of this embodiment realizes the judgment of traffic congestion events based on the vehicle-road-cloud integrated architecture.

[0066] The present invention is based on the vehicle-road-cloud integrated architecture, relying on the full amount of on-the-go vehicle data uploaded to the cloud control platform and the traffic light data of each intersection connected to the cloud control platform. After down-converting, cleaning and sorting the vehicle trajectory points, combined with high-precision map data, an algorithm for judging traffic congestion is proposed, and it supports being sent to on-the-go vehicles.

[0067] like Figure 3 The figure shows a vehicle-road integrated architecture, including: a cloud control platform 1, and a roadside device 2. Among them, a vehicle 3 traveling on the road communicates with the roadside device 2 through a communication protocol (for example, a V2X protocol). The roadside device 2 includes: a roadside unit (RSU) 21 and a roadside perception data measurement device 22. The roadside perception data measurement device 22 includes but is not limited to: a high-definition camera, a radar, etc. The roadside perception data measurement device 22 senses the vehicle 3 on the road, and uploads the roadside perception data to the cloud control platform 1 through, for example, an optical fiber 11. The cloud control platform 1 determines the congestion information based on the roadside perception data, and returns the congestion information to the vehicle 3 directly or through the RSU 21.

[0068] Specifically, the cloud control platform 1 is composed of a cloud control basic platform and cloud control applications, forming a "1+N" topology structure including "1" cloud control basic platform and "N" cloud control application platforms supported by the cloud control basic platform.

[0069] The following is a description of the terms used in the present invention, including:

[0070] An intersection is the intersection of two or more roads;

[0071] Stop line refers to the solid white line across the front of the intersection (mainly the traffic light intersection), indicating the position of vehicles waiting for the release of the traffic signal. There may be multiple stop lines at an intersection;

[0072] uuid refers to the unique identifier of the vehicle trajectory data in the cloud control platform;

[0073] Geometry refers to the spatial attributes in map data, which records the longitude and latitude of points, lines or surfaces representing the shapes of spatial objects.

[0074] The roadside equipment 2 collects vehicle information on the road and reports it to the cloud control platform 1 at a certain frequency. The data types on the cloud control platform 1 include the driving trajectory of the motor vehicle (including vehicle information such as vehicle uuid, longitude, latitude, instantaneous speed, etc.).

[0075] In summary, the data of the cloud control platform has the following characteristics:

[0076] 1. Real-time. Roadside equipment reports perception data to the cloud control platform at a fixed frequency, and streaming data can truly reflect the current operating status of the vehicle;

[0077] 2. Comprehensiveness. Roadside equipment can sense the information of all vehicles on the road. Due to the redundancy of sensors, the reported perception data will cover all targets on the road;

[0078] At the same time, due to the influence of objective factors such as network problems and roadside equipment, there will be some problems with the perception data reported to the cloud control platform 1, such as the lack of data in a certain period of time.

[0079] To ensure the accuracy of traffic index calculation, the perception data needs to be preprocessed before entering the algorithm, including:

[0080] Data downsampling: Get all the perception data of a uuid from the streaming data every second, sort the perception data by timestamp, and take the last data of the sorted perception data as the downsampling data within that second. In this way, for example, 10hz perception data can be downsampled to 1hz. Due to data delay or accumulation, the data received in one second may not be exactly 10. In this case, the last data after sorting is also taken, such as Figure 4 As shown: for data t0, t1, t2, ..., t(i-1), ti, take the last data ti after sorting.

[0081] Data cleaning: Get the perception data after frequency reduction. After accumulating a computing cycle, for the perception data under a uuid, if one or more of the following conditions are met:

[0082] a) The total amount of data is less than or equal to 1;

[0083] b) There are one or more trajectory points whose instantaneous speed is greater than a preset speed threshold, and the speed threshold is preferably 50 m / s;

[0084] The perception data of the uuid is deleted and does not participate in the calculation as valid data.

[0085] Data smoothing: Due to errors in the perception data, the longitude and latitude of the vehicle trajectory point may deviate from the actual position of the vehicle, and the perception data after frequency reduction within a calculation cycle needs to be smoothed. The present invention uses the Kalman filter algorithm to smooth the vehicle trajectory points under a single uuid. Kalman filtering is a linear quadratic estimation algorithm used for state estimation of linear dynamic systems. It has strong real-time performance and high fusion accuracy, and is widely used in the field of autonomous driving. The present invention will not go into details about the use of Kalman filtering.

[0086] The traffic light data of each intersection on the road is connected to the cloud control platform 1 through a third-party platform. The traffic light data refers to the data items of the car-cloud group standard protocol, such as the light state, cycle, remaining time, etc. of the traffic light.

[0087] The cloud control platform 1 executes the road congestion event identification method described in the present invention, determines the congestion information of the road surface, and sends the congestion information to the vehicle 3.

[0088] Specifically, first, the cloud control platform 1 executes step S101, and obtains the vehicle perception data and signal light data of the current lane of the current road section within a preset time before the recognition cycle during the recognition cycle.

[0089] The road congestion event recognition method of this embodiment is to recognize congestion events for a certain lane of a certain road section. The current road section is the road section for which congestion event information needs to be determined. The current lane refers to the lane for which congestion event information needs to be determined. Preferably, the road section is a road section in one direction of travel between two adjacent intersections.

[0090] Preferably, the recognition cycle of congestion event recognition is defined as 5s, wherein the basic data that needs to be input includes vehicle perception data and traffic light data. The time span of the data is within the preset time forward of the cycle. Preferably, the preset time is 3min. For example, if the current recognition cycle is 00:03:00-00:03:05 (the T cycle in the subsequent flow chart and logic is the recognition cycle), the vehicle perception data and traffic light data within 00:00:05-00:03:05 need to be input.

[0091] Step S102, determining corresponding congestion judgment logic according to the traffic light data;

[0092] Specifically, the traffic light data is the traffic light data of the current lane of the current road section. The traffic light data includes the light state, cycle, remaining time, etc. of the traffic light. According to the light color within the recognition cycle, the congestion judgment logic corresponding to the light color is selected to adapt to the driving conditions of vehicles under different light colors and accurately judge the congestion situation.

[0093] Step S103: using the congestion judgment logic to judge the road vehicle perception data, determine and output congestion information of the congestion event.

[0094] After the congestion information is determined, the congestion information is sent to the vehicle.

[0095] The present invention obtains the vehicle perception data and signal light data of the current lane of the current road section within the preset time before the identification cycle during the identification cycle, and determines the corresponding congestion judgment logic according to the signal light data, and then outputs the congestion information of the congestion event based on the determined congestion judgment logic. Therefore, the present invention not only considers the running state of the vehicle itself as the basis for judging congestion, but also combines the signal light data to conduct more comprehensive regularity analysis and judgment, and the data volume is more comprehensive and accurate, thereby improving the accuracy of intersection congestion.

[0096] like Figure 2 The flowchart of a method for identifying a road congestion event in another embodiment of the present invention includes:

[0097] Step S201, within a recognition cycle, obtain vehicle perception data and traffic light data of a current lane of a current road section within a preset time before the recognition cycle.

[0098] Step S202, when there is a traffic light control on the current road, determining the light color of the traffic light within the recognition cycle;

[0099] If all the traffic lights of the current lane are green during the identification period, determining that the corresponding congestion judgment logic is the green light congestion judgment logic;

[0100] If all the traffic lights of the current lane are red lights during the recognition period, determining that the corresponding congestion judgment logic is a red light congestion judgment logic, and the red light is a red light or a yellow light;

[0101] If the traffic lights of the current lane include green lights and red lights within the identification cycle, the proportion of light color duration within the identification cycle is determined; if the proportion of green light duration within the identification cycle is greater than the proportion of red light duration, the corresponding congestion judgment logic is determined to be the green light congestion judgment logic; if the proportion of green light duration within the identification cycle is less than the proportion of red light duration, the corresponding congestion judgment logic is determined to be the red light congestion judgment logic.

[0102] Step S203, when there is no traffic light control on the current road, determining that the corresponding congestion judgment logic is a no-light congestion judgment logic.

[0103] Step S204: using the congestion judgment logic to judge the road vehicle perception data, determine and output congestion information of the congestion event.

[0104] Specifically, step S201 is first executed to obtain, within the recognition cycle, vehicle perception data and traffic light data of the current lane of the current road section within a preset time before the recognition cycle.

[0105] Then, step S202 is executed to determine the light color of the signal light within the recognition cycle when there is a signal light control on the current road;

[0106] If all the traffic lights of the current lane are green during the identification period, determining that the corresponding congestion judgment logic is the green light congestion judgment logic;

[0107] If all the traffic lights of the current lane are red lights during the recognition period, determining that the corresponding congestion judgment logic is a red light congestion judgment logic, and the red light is a red light or a yellow light;

[0108] If the traffic lights of the current lane include green lights and red lights within the identification cycle, the proportion of light color duration within the identification cycle is determined; if the proportion of green light duration within the identification cycle is greater than the proportion of red light duration, the corresponding congestion judgment logic is determined to be the green light congestion judgment logic; if the proportion of green light duration within the identification cycle is less than the proportion of red light duration, the corresponding congestion judgment logic is determined to be the red light congestion judgment logic.

[0109] Specifically, the intuitive manifestation of congestion is that the traffic is running at a low speed. First, the low-speed traffic is defined. In this embodiment, during the green light period, vehicles with a speed lower than the green light low-speed threshold are defined as low speed, at the intersection without signal control, vehicles with a speed lower than the no-signal control low-speed threshold are defined as low speed, and during the red or yellow light period, vehicles with a speed lower than the red light low-speed threshold are defined as low speed.

[0110] Preferably, the green light low speed threshold is greater than the red light low speed threshold, and the no signal control low speed threshold is greater than the red light low speed threshold.

[0111] Preferably, the green light low speed threshold is 4 m / s, the no signal control low speed threshold is 4 m / s, and the red light low speed threshold is 2 m / s.

[0112] When there is a traffic light control on the current road, the congestion judgment logic is selected according to the color of the traffic light in the recognition cycle, wherein the judgment logic of the red light and the yellow light is the same, and the judgment logic of the green light is different from that of the red light and the yellow light. Specifically:

[0113] If the traffic lights of the current lane are all green during the recognition cycle, the corresponding congestion judgment logic is determined to be the green light congestion judgment logic;

[0114] If the traffic lights of the current lane are all red and / or yellow during the recognition cycle, the corresponding congestion judgment logic is determined to be the red light congestion judgment logic;

[0115] If the traffic lights of the current lane include green and red lights within the recognition cycle, the proportion of the light color duration within the recognition cycle is determined. If the proportion of the green light duration within the recognition cycle is greater than the sum of the proportions of the red and yellow light durations, the corresponding congestion judgment logic is determined to be the green light congestion judgment logic. If the proportion of the green light duration within the recognition cycle is less than the sum of the proportions of the red and yellow light durations, the corresponding congestion judgment logic is determined to be the red light congestion judgment logic. If the proportion of the green light duration within the recognition cycle is equal to the sum of the proportions of the red and yellow light durations, the corresponding congestion judgment logic is determined to be the green light congestion judgment logic or the red light congestion judgment logic.

[0116] When there is no traffic light control on the current road, step S203 is executed to determine that the corresponding congestion judgment logic is the no-light congestion judgment logic.

[0117] Specifically, Figure 5The figure shows a schematic diagram of a congestion event identification process according to a preferred embodiment of the present invention, including:

[0118] Step S501, input 3 minutes of vehicle perception data and signal light data to determine whether the current lane has signal light control, if yes, proceed to step S502; if no, proceed to no light logic;

[0119] Step S502, if the current lane is controlled by a traffic light, determine the color of the traffic light in the T cycle. If all lights are green within 5 seconds, execute step S503 to enter the green light congestion judgment logic; if all lights are red or yellow within 5 seconds, execute step S504 to enter the red light logic; if there are green, red, and yellow lights, execute step S505 to enter the logic of the light color with the highest duration.

[0120] Then, step S204 is executed to judge the road vehicle perception data using the congestion judgment logic to determine and output congestion information of the congestion event.

[0121] Specifically, the congestion information includes but is not limited to the congestion level, the congestion starting point, the congestion end point, the congestion length and / or the congestion speed.

[0122] This embodiment not only considers the congestion judgment of signalized intersections, but also considers the scenario of intersections without signalized lights. The method of this embodiment adopts the accurate data source of the cloud control platform. It not only considers the operating status of the vehicle itself as the basis for judging congestion, but also combines the signal light data to conduct a more comprehensive regularity analysis and judgment. At the same time, this embodiment takes into account the serviceability of the vehicle, and not only outputs the congestion judgment results and congestion levels, but also accurately outputs the starting and end points of the congestion and the average speed of the congested vehicles. In addition, the method of this patent is based on the full data of the cloud control platform, including map data, vehicle trajectory points after perception fusion, and signal light data, so in the calculation process, the amount of data is more comprehensive and accurate, thereby improving the accuracy of intersection congestion.

[0123] In summary, this embodiment proposes a method for accurately judging urban road congestion, which can output relevant information of lane-level congested sections based on the data of the cloud control platform under the vehicle-road-cloud integrated architecture. This information can be sent to vehicles on the way through different communication links, such as road cloud, vehicle cloud, etc., which not only serves the driver, but also can be used as input for the autonomous driving decision-making module.

[0124] In one embodiment, the using the congestion judgment logic to judge the road vehicle perception data, determine and output congestion information of the congestion event, includes:

[0125] When the congestion judgment logic is the green light congestion judgment logic, calculating the duration of the green light at the beginning of the recognition cycle;

[0126] When the duration is less than the preset time threshold, the average speed of the first vehicle before the stop line in the current lane within the identification period is calculated according to the vehicle perception data; if the average speed of the first vehicle before the stop line within the identification period is less than or equal to the green light low speed threshold, the first vehicle before the stop line is judged to be at a low speed; otherwise, it is judged that the lane is not congested, the congestion information of the congestion event is output as no congestion, and the process ends;

[0127] When the duration is greater than or equal to a preset time threshold or when it is determined that the first vehicle before the stop line is at a low speed, all vehicles in the current lane within the identification period whose speed is less than or equal to the green light low speed threshold are selected as low-speed vehicles, and spatial clustering is performed based on the positions of all the low-speed vehicles in the current lane;

[0128] When the number of clusters of the clustering result of the spatial clustering is greater than 0, the clustering results are traversed, and for each cluster, the position of the first vehicle in the cluster is selected as the congestion starting point, the position of the last vehicle in the cluster is selected as the congestion end point, the length along the road from the congestion starting point to the congestion end point is calculated as the congestion length of the cluster, and the average of the average speeds of each vehicle in the cluster is calculated as the congestion speed of the cluster;

[0129] Determining a congestion level of each cluster according to the congestion length of each cluster and / or the congestion speed of the cluster;

[0130] The congestion level, congestion starting point, congestion end point, congestion length and / or congestion speed of each cluster are used as congestion information of the congestion event and output.

[0131] Specifically, in the green light congestion judgment logic, the duration of the green light at the beginning of the recognition cycle is first judged. The start time of the green light can be found from the signal light data of the current lane of the current road section within a preset time before the recognition cycle, so as to calculate the duration of the green light at the beginning of the recognition cycle.

[0132] If the duration is less than the preset time threshold, that is, the duration does not exceed the preset time threshold, then the average speed of the first vehicle before the stop line in the current lane during the recognition period is calculated based on the vehicle perception data, and the first vehicle before the stop line is the vehicle closest to the stop line. The preset time threshold is preferably 10 seconds.

[0133] If the average speed of the first vehicle before the stop line in the identification period is less than or equal to the green light low speed threshold, the first vehicle before the stop line is judged to be at a low speed, otherwise if the duration is greater than or equal to the preset time threshold, it is judged that the lane is not congested, the congestion information of the congestion event is output as no congestion, and the process ends. Among them, the green light low speed threshold is preferably 4m / s.

[0134] When the duration is greater than or equal to the preset time threshold or the first vehicle before the stop line is judged to be slow, the speed of all vehicles in the recognition cycle is calculated, non-slow-speed vehicles are eliminated, and the last 1s position of all vehicles in the current lane and current section in the recognition cycle is taken to perform spatial clustering on the slow-speed vehicles. If there is a valid clustering result, that is, the number of clusters is greater than 0, proceed to the next step. Otherwise, it is judged to be non-congested and the process ends.

[0135] When the number of clusters in the clustering result of spatial clustering is greater than 0, traverse the clustering results. For each cluster, select the position of the first vehicle in each cluster as the starting point of the congestion and the position of the last vehicle as the end point of the congestion. Calculate the length along the road from the starting point to the end point as the congestion length, and take the average speed of all vehicles in each cluster as the congestion speed of the cluster.

[0136] Then, the first congestion level is determined according to the congestion length of each cluster: if the congestion length is less than 100m, the first congestion level is mild congestion; if the congestion length is greater than or equal to 100m and less than 200m, the first congestion level is moderate congestion; if the congestion length is greater than or equal to 200m and less than 300m, the first congestion level is severe congestion;

[0137] The second congestion level is determined according to the congestion speed of each cluster: if the congestion speed is less than 5km / h, the second congestion level is severe congestion; if the congestion speed is greater than or equal to 5km / h and less than 10km / h, the second congestion level is moderate congestion; if the congestion speed is greater than or equal to 10km / h and less than 15km / h, the second congestion level is mild congestion;

[0138] The congestion event recognition result is the less severe result of the first congestion level and the second congestion level calculated by the above two logics as the final congestion level. The severity of mild congestion is less severe than that of moderate congestion, and the severity of moderate congestion is less severe than that of severe congestion.

[0139] Finally, it is determined that a congestion event occurs, and the congestion level, congestion starting point, congestion end point, congestion length and / or congestion speed of each cluster are taken as congestion information of the congestion event and output.

[0140] This embodiment implements the green light congestion judgment logic. First, it judges whether there is congestion based on the duration of the green light, and can quickly filter out the congestion situation. When it is judged that there is congestion, the vehicles are clustered by clustering, and the congestion starting point, congestion end point, congestion length and / or congestion speed of the congestion are calculated and output to provide more abundant congestion information. At the same time, the congestion level is determined based on the congestion length and / or congestion speed, so that the judgment of the congestion level is more accurate.

[0141] like Figure 6As shown, the green light congestion judgment logic of the best embodiment of the present invention includes:

[0142] Step S601, calculate how many seconds the green light has lasted when the T period starts. If the duration does not exceed 10 seconds, proceed to step S602, otherwise proceed to step S603;

[0143] Step S602, calculate the average speed of the first vehicle in the current lane (the one closest to the stop line) within the T period, and determine whether the first vehicle is at a low speed. If so, proceed to step S603, otherwise it is considered that there is no congestion at present, and the process ends;

[0144] Step S603, calculate the speed of all vehicles in the T period, remove non-low-speed vehicles, take the last 1s position of all vehicles on the lane in the T period, and perform spatial clustering on the low-speed vehicles. If there is a valid clustering result, that is, the number of clusters is greater than 0, go to step S604, otherwise it is determined that there is no congestion;

[0145] Step S604, traverse the clustering results, select the position of the first car as the congestion starting point, the position of the last car as the congestion end point, calculate the length along the road from the starting point to the end point as the congestion length, calculate the average speed of each car as the congestion speed, output the clustering results within the T period, including the congestion starting point, congestion end point, congestion length, and congestion speed, and enter step S605;

[0146] Step S605, determining the congestion level according to the congestion speed and congestion length within the T period.

[0147] The congestion level determination logic during the green light period is as follows:

[0148] First congestion level: if the congestion length is less than 100m, it is mild congestion; if the congestion length is greater than or equal to 100m and less than 200m, it is moderate congestion; if the congestion length is greater than or equal to 200m and less than 300m, it is severe congestion;

[0149] Second congestion level: if the average speed is less than 5km / h, it is severe congestion; if the average speed is greater than or equal to 5km / h and less than 10km / h, it is moderate congestion; if the average speed is greater than or equal to 10km / h and less than 15km / h, it is light congestion;

[0150] For the congestion event identification result, the less severe result of the first congestion level and the second congestion level 2 calculated by the above two logics is taken as the final congestion level.

[0151] In one embodiment, the using the congestion judgment logic to judge the road vehicle perception data, determine and output congestion information of the congestion event, includes:

[0152] When the congestion judgment logic is a red light congestion judgment logic, red light congestion information calculation is performed, and congestion information of the congestion event is output;

[0153] The red light congestion information calculation includes: selecting all vehicles in the current lane within the identification period whose speed is less than or equal to the red light low speed threshold as low-speed vehicles, and performing spatial clustering based on the positions of all the low-speed vehicles in the current lane;

[0154] When the number of clusters of the clustering result of the spatial clustering is greater than 0, the clustering results are traversed, and for each cluster, the position of the first vehicle in the cluster is selected as the congestion starting point, the position of the last vehicle in the cluster is selected as the congestion end point, the length along the road from the congestion starting point to the congestion end point is calculated as the congestion length of the cluster, and the average of the average speeds of each vehicle in the cluster is calculated as the congestion speed of the cluster;

[0155] Determine the congestion level of each cluster according to the congestion length of each cluster;

[0156] The congestion level, congestion starting point, congestion end point, congestion length and / or congestion speed of each cluster are used as congestion information.

[0157] Specifically, the congestion judgment logic for the red light and the yellow light is the same, which is the red light congestion judgment logic.

[0158] In the red light congestion judgment logic, the red light congestion information calculation is performed and the congestion information of the congestion event is output;

[0159] The calculation of red light congestion information includes: calculating the speed of all vehicles in the recognition period, eliminating non-low-speed vehicles, taking the last 1s position of all vehicles in the current lane and the current section in the recognition period, and spatially clustering the low-speed vehicles. If there is a valid clustering result, that is, the number of clusters is greater than 0, proceed to the subsequent steps. Otherwise, it is judged as not congested and the process ends.

[0160] When the number of clusters in the clustering result of spatial clustering is greater than 0, traverse the clustering results. For each cluster, select the position of the first vehicle in each cluster as the starting point of the congestion and the position of the last vehicle as the end point of the congestion. Calculate the length along the road from the starting point to the end point as the congestion length, and take the average speed of all vehicles in each cluster as the congestion speed of the cluster.

[0161] Then, the congestion level of each cluster is determined according to the congestion length of each cluster.

[0162] Specifically, the congestion level determination logic during the red and / or yellow light period is: if the congestion length is less than 100m, the congestion level is mild congestion; if the congestion length is greater than or equal to 100m and less than 200m, the congestion level is moderate congestion; if the congestion length is greater than or equal to 200m and less than 300m, the congestion level is severe congestion.

[0163] Finally, the congestion level, congestion starting point, congestion end point, congestion length and / or congestion speed of each cluster are used as congestion information.

[0164] This embodiment implements the red light and / or yellow light congestion judgment logic, clusters the vehicles by clustering, calculates and outputs the congestion starting point, congestion end point, congestion length and / or congestion speed of the congestion, and provides more abundant congestion information. At the same time, the congestion judgment of the red light and / or yellow light is taken into account, so the congestion level is determined only based on the congestion length, making the judgment of the congestion level more accurate.

[0165] In one embodiment, the using the congestion judgment logic to judge the road vehicle perception data, determine and output congestion information of the congestion event, also includes:

[0166] Determine the time range of the previous green light cycle of the current red light from the signal light data, and calculate the starting average speed of the first vehicle before the stop line on the current lane in the previous green light cycle within the starting time period of the previous green light cycle and the ending average speed within the ending time period of the previous green light cycle according to the vehicle perception data;

[0167] When the starting average speed is less than or equal to the red light low speed threshold, red light congestion information calculation is performed; when the starting average speed is greater than the red light low speed threshold, the ending average speed is determined; if the ending average speed is less than or equal to the red light low speed threshold, a flag is set, and then red light congestion information calculation is performed; otherwise, red light congestion information calculation is directly performed;

[0168] Determining the congestion level of each cluster according to the congestion length of each cluster includes: determining the congestion level of each cluster according to the congestion length of each cluster, and downgrading the congestion level if the flag is set.

[0169] Specifically, based on the traffic light data, find the time range of the previous green light cycle of the current red light, then find the data corresponding to the first vehicle in front of the stop line in the previous green light cycle in the perception data, and calculate the starting average speed of the first vehicle in front of the stop line in the starting time period of the previous green light cycle and the ending average speed in the ending time period. The starting time period is preferably 5s, and the ending time period is preferably 5s.

[0170] Then, determine whether the speed is low during the start time period. If so, directly perform the red light congestion information calculation. Otherwise, determine whether the first vehicle before the stop line is at a low speed during the end time period of the previous green light cycle. If so, generate a mark of 1, otherwise mark 0.

[0171] When judging the congestion level, if it is determined whether there is a mark and the mark is equal to 1, if so, the congestion level is reduced by one level. If it is slightly congested, it is not congested and the congestion result is output.

[0172] When the traffic light just turns green, there will be a certain delay between the driver reading the green light and the vehicle actuator starting. Therefore, this embodiment adds a speed judgment for the first vehicle before the stop line within the previous green light cycle of the current red light to avoid misjudgment.

[0173] like Figure 7 The red light and yellow light congestion judgment logic of the best embodiment of the present invention is shown, including:

[0174] Step S701, based on the traffic light data, find the time range of the previous green light cycle of the current red light, then find the data corresponding to the first vehicle in front of the stop line in the previous green light cycle in the perception data, calculate the average speed of the first vehicle in front of the stop line in the first 5 seconds and the last 5 seconds of the previous green light cycle, and determine whether the first 5 seconds is a low speed. If so, go to step S703, otherwise go to step S702;

[0175] Step S702, determine whether the first vehicle before the stop line is slowing down 5 seconds after the last green light cycle ends. If yes, generate a flag of 1, otherwise 0, and proceed to step S703;

[0176] Step S703, calculate the speed of all vehicles in the T period, remove non-low-speed vehicles, take the last 1s position of all vehicles on the lane in the T period, and perform spatial clustering on the low-speed vehicles. If there is a valid clustering result, that is, the number of clusters is greater than 0, go to step S704, otherwise it is judged that there is no congestion;

[0177] Step S704, traverse the clustering results, select the position of the first car as the congestion starting point, the position of the last car as the congestion end point, calculate the length along the road from the starting point to the end point as the congestion length, calculate the average speed of each car as the congestion speed, output the clustering results within the T period, including the congestion starting point, congestion end point, congestion length, and congestion speed, and enter step S705;

[0178] Step S705, determining the congestion level according to the congestion length within the T period;

[0179] The congestion level determination logic during the red light period is: if the congestion length is less than 100m, it is mild congestion; if the congestion length is greater than or equal to 100m and less than 200m, it is moderate congestion; if the congestion length is greater than or equal to 200m and less than 300m, it is severe congestion;

[0180] Step S706, determine whether there is a mark and the mark is equal to 1. If so, the congestion level is reduced by one level, otherwise the congestion result is directly output. If there is light congestion, it is not congested and the congestion result is output.

[0181] In one embodiment, the using the congestion judgment logic to judge the road vehicle perception data, determine and output congestion information of the congestion event, includes:

[0182] When the congestion judgment logic is a no-light congestion judgment logic, all vehicles in the current lane within the identification period whose speed is less than or equal to the no-signal control low-speed threshold are selected as low-speed vehicles, and spatial clustering is performed based on the positions of all the low-speed vehicles in the current lane;

[0183] When the number of clusters of the clustering result of the spatial clustering is greater than 0, the clustering results are traversed, and for each cluster, the position of the first vehicle in the cluster is selected as the congestion starting point, the position of the last vehicle in the cluster is selected as the congestion end point, the length along the road from the congestion starting point to the congestion end point is calculated as the congestion length of the cluster, and the average of the average speeds of each vehicle in the cluster is calculated as the congestion speed of the cluster;

[0184] Determining a congestion level of each cluster according to the congestion length of each cluster and / or the congestion speed of the cluster;

[0185] The congestion level, congestion starting point, congestion end point, congestion length and / or congestion speed of each cluster are used as congestion information of the congestion event and output.

[0186] Calculate the speed of all vehicles in the recognition cycle, exclude non-low-speed vehicles, take the last 1s position of all vehicles in the current lane and current section in the recognition cycle, and perform spatial clustering on low-speed vehicles. If there is a valid clustering result, that is, the number of clusters is greater than 0, proceed to the next step. Otherwise, it is judged as not congested and the process ends.

[0187] When the number of clusters in the clustering result of spatial clustering is greater than 0, traverse the clustering results. For each cluster, select the position of the first vehicle in each cluster as the starting point of the congestion and the position of the last vehicle as the end point of the congestion. Calculate the length along the road from the starting point to the end point as the congestion length, and take the average speed of all vehicles in each cluster as the congestion speed of the cluster.

[0188] Then, the first congestion level is determined according to the congestion length of each cluster: if the congestion length is less than 100m, the first congestion level is mild congestion; if the congestion length is greater than or equal to 100m and less than 200m, the first congestion level is moderate congestion; if the congestion length is greater than or equal to 200m and less than 300m, the first congestion level is severe congestion;

[0189] The second congestion level is determined according to the congestion speed of each cluster: if the congestion speed is less than 5km / h, the second congestion level is severe congestion; if the congestion speed is greater than or equal to 5km / h and less than 10km / h, the second congestion level is moderate congestion; if the congestion speed is greater than or equal to 10km / h and less than 15km / h, the second congestion level is mild congestion;

[0190] The congestion event recognition result is the less severe result of the first congestion level and the second congestion level calculated by the above two logics as the final congestion level. The severity of mild congestion is less severe than that of moderate congestion, and the severity of moderate congestion is less severe than that of severe congestion.

[0191] Finally, the congestion level, congestion starting point, congestion end point, congestion length and / or congestion speed of each cluster are output as congestion information of the congestion event.

[0192] This embodiment implements the no-light congestion judgment logic, clusters the vehicles by clustering, calculates and outputs the congestion starting point, congestion end point, congestion length and / or congestion speed of the congestion, and provides more abundant congestion information. At the same time, the congestion level is determined based on the congestion length and / or congestion speed, making the judgment of the congestion level more accurate.

[0193] like Figure 8 The following is a no-light congestion judgment logic of the best embodiment of the present invention, including:

[0194] Step S801, calculate the speed of all vehicles within the T period, remove non-low-speed vehicles, take the last 1s position of all vehicles on the lane within the T period, and perform spatial clustering on the low-speed vehicles. If there is a valid clustering result, that is, the number of clusters is greater than 0, proceed to step S802, otherwise it is determined that there is no congestion;

[0195] Step S802, traverse the clustering results, select the position of the first car as the congestion starting point, the position of the last car as the congestion end point, calculate the length along the road from the starting point to the end point as the congestion length, calculate the average speed of each car as the congestion speed, output the clustering results within the T period, including the congestion starting point, congestion end point, congestion length, and congestion speed, and enter step S803;

[0196] Step S803, determining the congestion level according to the congestion speed and congestion length within the T period.

[0197] The logic for judging the congestion level at intersections without signal control is as follows:

[0198] First congestion level: if the congestion length is less than 100m, it is mild congestion; if the congestion length is greater than or equal to 100m and less than 200m, it is moderate congestion; if the congestion length is greater than or equal to 200m and less than 300m, it is severe congestion;

[0199] Second congestion level: if the average speed is less than 5km / h, it is severe congestion; if the average speed is greater than or equal to 5km / h and less than 10km / h, it is moderate congestion; if the average speed is greater than or equal to 10km / h and less than 15km / h, it is light congestion;

[0200] For the congestion event identification result, the less severe result of the first congestion level and the second congestion level calculated by the above two logics is taken as the final congestion level.

[0201] In some embodiments, spatial clustering uses the DBSCAN clustering algorithm, with the spatial distance threshold set to 100m and the minimum number of cluster elements set to 2, including:

[0202] The original set creation step includes: creating an original set A including all low-speed vehicles, sorting the vehicles in the original set from large to small according to the distance from the stop line, initializing the cluster identifier to 0, and initializing the cluster set to an empty set;

[0203] The initialization step includes: initializing the temporary cluster set to an empty set, adding the first unselected car in the original set to the temporary cluster set, traversing from the first car in the original set, selecting a car as the current car each time, and executing the traversal step;

[0204] The traversal step includes: calculating the distance between the current vehicle and the next vehicle in the original set, determining whether the distance is less than a distance threshold, if so, adding the next vehicle to the temporary cluster set, deleting the current vehicle from the original set, selecting the next vehicle as the current vehicle, if there are still unselected vehicles in the original set, executing the traversal step again, if there are no unselected vehicles in the original set or the distance is greater than or equal to the distance threshold, executing the cluster adding step;

[0205] The cluster adding step includes: judging the number of elements in the temporary cluster set, if the number is greater than or equal to 2, adding the temporary cluster set as a clustering result to the cluster set, and increasing the cluster identifier by one, if there are still unselected vehicles in the original set, executing the initialization step again, until all vehicles in the original set are selected, and then executing the judging step;

[0206] In the judgment step, after traversing all the elements in the original set, it is judged whether the cluster identifier is greater than 0. If so, it means there is a valid clustering result, otherwise it means there is no valid clustering result.

[0207] like Fig. 9 The clustering logic diagram of the best embodiment of the present invention is shown, including:

[0208] Step S901, create a set A of all vehicles, sort the vehicles in A from large to small according to the distance from the stop line, initialize the cluster identifier ClusterID = 0, and the cluster set ClusterSet = {};

[0209] Step S902, initialize the temporary cluster set CurrentSet = [], add the first car to CurrentSet, and start traversing from the first car to be traversed in A;

[0210] Step S903, calculate the distance between the current car and the next car, and determine whether the distance is less than 100m. If so, add the next car to CurrentSet and delete the current car from A. Repeat step S903 until there are no untraversed vehicles in A, or the distance between the current car and the next car is greater than or equal to 100, then enter step S904;

[0211] Step S904, determine the number of elements in CurrentSet. If the number is greater than or equal to 2, add CurrentSet as a clustering result to ClusterSet, and add ClusterID+1. If the elements in A have not been traversed, execute step S902 again, otherwise execute step S905;

[0212] Step S905, after traversing all elements in A, determine whether ClusterID is greater than 0. If so, it means there is a valid clustering result, otherwise it means there is no valid clustering result.

[0213] In some embodiments, each cluster is regarded as a congested road section, the position of the first vehicle in the cluster is the starting position of the congested road section, and the position of the last vehicle in the cluster is the ending position of the congested road section.

[0214] In some embodiments, if the length of the middle congested road segment among three consecutive congested road segments is less than the merged road segment length threshold, the three congested road segments are merged into one congested road segment, and the congested road segment is a cluster.

[0215] like Fig.10As shown, in this algorithm, if it is determined that there are multiple discontinuous congested sections 4, 5, and 6, but the length of the discontinuous congested section 5 is less than or equal to the merged section length threshold D1 (configuration value: 50m in urban areas and 200m on highways), the congested sections 4, 5, and 6 are merged, and the output congestion is continuous;

[0216] If it is determined that there are multiple discontinuous congested sections, but the length of the discontinuous congested sections is greater than the merged section length threshold D1 (configuration value TBD: 50m in urban areas, 200m on highways), the algorithm outputs multiple congested sections.

[0217] In this embodiment, the lane-level congestion start and end points, as well as the average speed and congestion level of the congested road section can be output.

[0218] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0219] like Fig.11 The figure shows a hardware structure diagram of an electronic device of the present invention, including:

[0220] at least one processor 1101; and,

[0221] A memory 1102 is communicatively connected to at least one of the processors 1101; wherein:

[0222] The memory 1102 stores instructions that can be executed by at least one of the processors. The instructions are executed by at least one of the processors so that the at least one of the processors can execute the road congestion event identification method as described above.

[0223] Fig.11 A processor 1101 is taken as an example.

[0224] The electronic device may further include: an input device 1103 and a display device 1104 .

[0225] The processor 1101, the memory 1102, the input device 1103 and the display device 1104 may be connected via a bus or other means, and the figure takes the connection via a bus as an example.

[0226] The memory 1102 is a non-volatile computer-readable storage medium that can be used to store non-volatile software programs, non-volatile computer executable programs and modules, such as program instructions / modules corresponding to the road congestion event identification method in the embodiment of the present application, for example, Figure 1 , Figure 2The processor 1101 executes various functional applications and data processing by running the non-volatile software programs, instructions and modules stored in the memory 1102, that is, the road congestion event recognition method in the above embodiment is implemented.

[0227] The memory 1102 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application required by at least one function; the data storage area may store data created according to the use of the road congestion event identification method, etc. In addition, the memory 1102 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 1102 may optionally include a memory remotely arranged relative to the processor 1101, and these remote memories may be connected to the device for executing the road congestion event identification method via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0228] The input device 1103 can receive user clicks and generate signal inputs related to user settings and function controls of the road congestion event recognition method. The display device 1104 can include display devices such as a display screen.

[0229] The one or more modules are stored in the memory 1102 , and when executed by the one or more processors 1101 , the road congestion event recognition method in any of the above method embodiments is executed.

[0230] The present invention obtains the vehicle perception data and signal light data of the current lane of the current road section within the preset time before the identification cycle during the identification cycle, and determines the corresponding congestion judgment logic according to the signal light data, and then outputs the congestion information of the congestion event based on the determined congestion judgment logic. Therefore, the present invention not only considers the running state of the vehicle itself as the basis for judging congestion, but also combines the signal light data to conduct more comprehensive regularity analysis and judgment, and the data volume is more comprehensive and accurate, thereby improving the accuracy of intersection congestion.

[0231] An embodiment of the present invention provides a storage medium, wherein the storage medium stores computer instructions. When a computer executes the computer instructions, the storage medium is used to execute all steps of the road congestion event identification method as described above.

[0232] In the context of the present disclosure, a storage medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or apparatus. The storage medium may be a machine-readable signal medium or a machine-readable storage medium. Alternatively, the storage medium may be a non-temporary computer-readable storage medium, for example, a non-temporary computer-readable storage medium may be a ROM, a random access memory (Random Access Memory, RAM), a compact disc read-only memory (Compact Disc ROM, CD-ROM), a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0233] An embodiment of the present invention provides a computer program product, including a computer program / instruction, which implements the road congestion event identification method as described above when executed by a processor.

[0234] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for identifying road congestion events, characterized in that: include: During the recognition cycle, the vehicle perception data and signal light data of the current lane of the current road section within a preset time before the recognition cycle are obtained; Determine corresponding congestion judgment logic according to the traffic light data; The congestion judgment logic is used to judge the road vehicle perception data, and congestion information of the congestion event is determined and output.

2. The method for identifying road congestion events according to claim 1, characterized in that: The determining, according to the traffic light data, corresponding congestion judgment logic includes: When there is a traffic light control on the current road, determining the light color of the traffic light within the recognition cycle; If all the traffic lights of the current lane are green during the identification period, determining that the corresponding congestion judgment logic is the green light congestion judgment logic; If the traffic lights of the current lane are all red and / or yellow in the recognition cycle, determining that the corresponding congestion judgment logic is a red light congestion judgment logic; If the traffic lights of the current lane include green lights and red lights within the identification cycle, then the proportion of light color duration within the identification cycle is determined; if the proportion of green light duration within the identification cycle is greater than the sum of the proportions of red and yellow light durations, then the corresponding congestion judgment logic is determined to be the green light congestion judgment logic; if the proportion of green light duration within the identification cycle is less than the sum of the proportions of red and yellow light durations, then the corresponding congestion judgment logic is determined to be the red light congestion judgment logic.

3. The method for identifying road congestion events according to claim 2, characterized in that: The using the congestion judgment logic to judge the road vehicle perception data, determine and output congestion information of the congestion event, includes: When the congestion judgment logic is the green light congestion judgment logic, calculating the duration of the green light at the beginning of the recognition cycle; When the duration is less than the preset time threshold, the average speed of the first vehicle before the stop line in the current lane within the identification period is calculated according to the vehicle perception data; if the average speed of the first vehicle before the stop line within the identification period is less than or equal to the green light low speed threshold, the first vehicle before the stop line is judged to be at a low speed; otherwise, it is judged that the lane is not congested, the congestion information of the congestion event is output as no congestion, and the process ends; When the duration is greater than or equal to a preset time threshold or when it is determined that the first vehicle before the stop line is at a low speed, all vehicles in the current lane within the identification period whose speed is less than or equal to the green light low speed threshold are selected as low-speed vehicles, and spatial clustering is performed based on the positions of all the low-speed vehicles in the current lane; When the number of clusters of the clustering result of the spatial clustering is greater than 0, the clustering results are traversed, and for each cluster, the position of the first vehicle in the cluster is selected as the congestion starting point, the position of the last vehicle in the cluster is selected as the congestion end point, the length along the road from the congestion starting point to the congestion end point is calculated as the congestion length of the cluster, and the average of the average speeds of each vehicle in the cluster is calculated as the congestion speed of the cluster; Determining a congestion level of each cluster according to the congestion length of each cluster and / or the congestion speed of the cluster; The congestion level, congestion starting point, congestion end point, congestion length and / or congestion speed of each cluster are used as congestion information of the congestion event and output.

4. The method for identifying road congestion events according to claim 2, characterized in that: The using the congestion judgment logic to judge the road vehicle perception data, determine and output congestion information of the congestion event, includes: When the congestion judgment logic is a red light congestion judgment logic, red light congestion information calculation is performed, and congestion information of the congestion event is output; The red light congestion information calculation includes: selecting all vehicles in the current lane within the identification period whose speed is less than or equal to the red light low speed threshold as low-speed vehicles, and performing spatial clustering based on the positions of all the low-speed vehicles in the current lane; When the number of clusters of the clustering result of the spatial clustering is greater than 0, the clustering results are traversed, and for each cluster, the position of the first vehicle in the cluster is selected as the congestion starting point, the position of the last vehicle in the cluster is selected as the congestion end point, the length along the road from the congestion starting point to the congestion end point is calculated as the congestion length of the cluster, and the average of the average speeds of each vehicle in the cluster is calculated as the congestion speed of the cluster; Determine the congestion level of each cluster according to the congestion length of each cluster; The congestion level, congestion starting point, congestion end point, congestion length and / or congestion speed of each cluster are used as congestion information.

5. The method for identifying road congestion events according to claim 4, characterized in that: The using the congestion judgment logic to judge the road vehicle perception data, determine and output congestion information of the congestion event, also includes: Determine the time range of the previous green light cycle of the current red light from the signal light data, and calculate the starting average speed of the first vehicle before the stop line on the current lane in the previous green light cycle within the starting time period of the previous green light cycle and the ending average speed within the ending time period of the previous green light cycle according to the vehicle perception data; When the starting average speed is less than or equal to the red light low speed threshold, red light congestion information calculation is performed; when the starting average speed is greater than the red light low speed threshold, the ending average speed is determined; if the ending average speed is less than or equal to the red light low speed threshold, a flag is set, and then red light congestion information calculation is performed; otherwise, red light congestion information calculation is directly performed; Determining the congestion level of each cluster according to the congestion length of each cluster includes: determining the congestion level of each cluster according to the congestion length of each cluster, and downgrading the congestion level if the flag is set.

6. The method for identifying road congestion events according to claim 1, characterized in that: The determining the corresponding congestion judgment logic according to the traffic light data further includes: When there is no traffic light control on the current road, the corresponding congestion judgment logic is determined to be a no-light congestion judgment logic.

7. The method for identifying road congestion events according to claim 6, characterized in that: The using the congestion judgment logic to judge the road vehicle perception data, determine and output congestion information of the congestion event, includes: When the congestion judgment logic is a no-light congestion judgment logic, all vehicles in the current lane within the identification period whose speed is less than or equal to the no-signal control low-speed threshold are selected as low-speed vehicles, and spatial clustering is performed based on the positions of all the low-speed vehicles in the current lane; When the number of clusters of the clustering result of the spatial clustering is greater than 0, the clustering results are traversed, and for each cluster, the position of the first vehicle in the cluster is selected as the congestion starting point, the position of the last vehicle in the cluster is selected as the congestion end point, the length along the road from the congestion starting point to the congestion end point is calculated as the congestion length of the cluster, and the average of the average speeds of each vehicle in the cluster is calculated as the congestion speed of the cluster; Determining a congestion level of each cluster according to the congestion length of each cluster and / or the congestion speed of the cluster; The congestion level, congestion starting point, congestion end point, congestion length and / or congestion speed of each cluster are used as congestion information of the congestion event and output.

8. An electronic device, characterized in that: include: at least one processor; as well as, a memory communicatively connected to at least one of the processors; wherein, The memory stores instructions that can be executed by at least one of the processors, and the instructions are executed by at least one of the processors to enable at least one of the processors to execute the road congestion event identification method as described in any one of claims 1 to 7.

9. A storage medium, characterized in that: The storage medium stores computer instructions, and when a computer executes the computer instructions, it is used to execute all steps of the road congestion event identification method as described in any one of claims 1 to 7.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, the road congestion event identification method as described in any one of claims 1 to 7 is implemented.

Images