Traffic congestion judgment method and device based on traffic density

Abstract

The application discloses a traffic jam judgment method and device based on traffic density. The method is as follows: based on the environment sensing information collected by a sensor, a plurality of target lanes matched with a plurality of targets are determined; effective targets are selected from the plurality of targets matched with adjacent lanes; based on the sensing data of the effective targets, the traffic density of the adjacent lanes is determined; and based on the traffic density, the traffic jam condition of the adjacent lanes is determined. The method uses the environment sensing data collected by the sensor to determine the traffic density of the adjacent lanes, and uses the traffic density to determine the traffic jam condition of the adjacent lanes, so as to effectively judge the traffic jam condition of the adjacent lanes on the driving path of the vehicle, and provide a reliable environment sensing reference for the driving behavior of the automatic driving of the vehicle.

Description

Technical Field

[0001] This application relates to the field of environmental perception, and in particular to a method and device for judging traffic congestion based on vehicle flow density. Background Technology

[0002] Environmental perception methods in autonomous driving technology can generally be divided into lane-keeping perception and lane-change perception based on driving intent. Lane-keeping perception focuses on the motion state of the nearest or most threatening target in the current lane's driving direction. Lane-change perception, in addition to lane-keeping perception, adds the perception of the motion state of targets in the target lane that are on the vehicle's lane-change trajectory.

[0003] Existing perception methods tend to focus on perceiving individual targets, neglecting the traffic flow environment comprised of these individual targets. When there are no targets in the lane or the targets are far away, the perception system judges that there is ample driving space, and the autonomous driving control system will control the vehicle to travel at a higher speed. However, if the adjacent lanes are densely packed with vehicles, the driver's judgment of the driving space may deviate from the perception scheme, and excessive speed may cause the driver to feel subjectively nervous.

[0004] Therefore, how to determine the traffic congestion status of each lane on the vehicle's driving path and improve the user experience of autonomous driving has become a current research hotspot. Summary of the Invention

[0005] This application provides a method and apparatus for determining traffic congestion based on vehicle flow density, with the aim of determining the traffic congestion status of each lane on the vehicle's travel path.

[0006] To achieve the above objectives, this application provides the following technical solution:

[0007] A traffic congestion assessment method based on vehicle flow density includes:

[0008] Based on the environmental perception information collected by the sensors, lanes matching multiple targets are determined; the targets are vehicles located within the perception range of the sensors.

[0009] Valid targets are selected from multiple targets matched in adjacent lanes; the adjacent lanes are those adjacent to the lane where the sensor is located; the number of valid targets is at least three.

[0010] Based on the perception data of the effective target, the traffic density of the adjacent lane is determined;

[0011] Based on the traffic density, the traffic congestion status of the adjacent lanes is determined.

[0012] Optionally, based on environmental perception information collected by sensors, determine the lanes matching multiple targets, including:

[0013] The system acquires environmental perception information collected by sensors; the environmental perception information includes perception data of multiple targets; the perception data includes coordinate information and lane line parameters.

[0014] Based on the coordinate information of multiple targets and lane line parameters, the lateral position information of multiple targets is determined; the lateral position information includes the lateral position reference value of the left lane line, the lateral position reference value of the right lane line, the lateral position reference value of the left-left lane line, and the lateral position reference value of the right-right lane line.

[0015] Based on the coordinate and lateral position information of the multiple targets, the lanes that match the multiple targets are determined.

[0016] Optionally, based on the coordinate information and lateral position information of the multiple targets, the lanes matched to the multiple targets are determined, including:

[0017] For each target, a corresponding target reference value is determined based on the target's coordinate information;

[0018] If the target reference value is greater than or equal to the left-left lane line lateral position reference value and less than or equal to the left lane line lateral position reference value, the lane matched by the target is determined to be the left lane; the left lane is the lane adjacent to the left side of the lane where the sensor is located.

[0019] If the target reference value is greater than or equal to the right lane line lateral position reference value and less than or equal to the right lane line lateral position reference value, the lane matched by the target is determined to be the right lane; the right lane is the lane adjacent to the right side of the lane where the sensor is located.

[0020] Optionally, valid targets can be filtered from multiple targets matched in adjacent lanes, including:

[0021] Based on the coordinate information and index values ​​of multiple targets matched in adjacent lanes, the spacing between the multiple targets is determined; wherein, the index value represents the order in which the target is perceived by the sensor along a specified direction of the adjacent lane; the spacing represents the distance between the target and the next target; the index value of the next target is placed after the target, and the difference between the index value of the next target and the index value of the target is 1;

[0022] From the plurality of targets, determine a first target whose spacing is greater than a preset maximum longitudinal distance threshold;

[0023] Based on the index value of the first target, determine the corresponding invalid target;

[0024] Identify the valid targets among the multiple targets matched by the adjacent lanes, excluding the invalid targets.

[0025] Optionally, based on the index value of the first target, the corresponding invalid target is determined, including:

[0026] If the index value of the first target is greater than the target value, other targets whose index values ​​are ranked after the first target are determined to be invalid targets; the target value is determined based on the total number of targets matched by the adjacent lanes.

[0027] If the index value of the first target is less than or equal to the target value, it is determined to be an invalid target based on the first target and other targets whose index values ​​are ranked before the first target.

[0028] Optionally, determining the valid targets other than the invalid targets among the multiple targets matched by the adjacent lanes includes:

[0029] The perception data of multiple candidate targets, excluding the invalid targets, are determined from among the multiple targets matched by the adjacent lanes; the perception data includes coordinate information and vehicle length.

[0030] Based on the coordinate information of multiple candidate targets and the overall vehicle length, the distance between the front ends of multiple candidate targets is determined; the distance between the front ends of a candidate target and the rear ends of the next candidate target is represented; the index value of the next candidate target is placed after the candidate target, and the difference between the index value of the next candidate target and the index value of the candidate target is 1;

[0031] For each candidate target, if the distance between the front of the candidate target and the vehicle is greater than or equal to a preset minimum longitudinal distance threshold, the candidate target is determined as a valid target.

[0032] Optionally, based on the perception data of the effective target, determining the traffic flow density of the adjacent lanes includes:

[0033] Determine perception data for at least three of the effective targets; the perception data includes coordinate information, vehicle length, and vehicle speed.

[0034] Based on the coordinate information of each of the effective targets, the length of the vehicle and the speed of the vehicle, the headway of each of the effective targets is determined.

[0035] The traffic density of the adjacent lanes is determined based on the arithmetic mean of the headway distances of each vehicle.

[0036] Optionally, based on the traffic flow density, determining the traffic congestion status of the adjacent lanes includes:

[0037] Based on the traffic flow of the adjacent lanes, a corresponding congestion score is determined; the traffic flow is determined based on the total number of targets matched by the adjacent lanes within a specified time period; the congestion score is a quantitative indicator for measuring the degree of traffic congestion.

[0038] When the traffic density is greater than a preset density threshold, the congestion score is adjusted upward according to a specified slope, so that the congestion score after adjustment is increased by a first specified score compared with the congestion score before adjustment.

[0039] When the traffic density is less than or equal to the preset density threshold, the congestion score is reduced according to the specified slope, so that the congestion score after reduction is lower than the congestion score before reduction by a second specified score.

[0040] Specifically, when the congestion score is greater than the first congestion calibration value, the traffic congestion status of the adjacent lane is determined to be congested; when the congestion score is less than the second congestion calibration value, the traffic congestion status of the adjacent lane is determined to be uncongested; the first congestion calibration value is greater than the second congestion calibration value.

[0041] Optionally, the type of the second specified score includes a first score and a second score, where the first score is less than the second score. When the traffic density is less than or equal to the preset density threshold, the congestion score is adjusted downwards according to the specified slope, so that the adjusted congestion score is lower than the original congestion score by a second specified score, including:

[0042] When the traffic density is less than or equal to the preset density threshold and the traffic volume meets the preset threshold, the congestion score is reduced according to the specified slope so that the reduced congestion score is lower than the original congestion score by the second score.

[0043] When the traffic density is less than or equal to the preset density threshold and the traffic volume does not meet the preset threshold, the congestion score is adjusted downward according to the specified slope, so that the congestion score after adjustment is lower than the congestion score before adjustment by the first score.

[0044] A traffic congestion judgment device based on vehicle flow density includes:

[0045] The lane matching unit is used to determine the lanes to be matched for multiple targets based on environmental perception information collected by sensors; the targets are vehicles located within the perception range of the sensors.

[0046] A target filtering unit is used to filter out valid targets from multiple targets matched in adjacent lanes; the adjacent lanes are lanes adjacent to the lane where the sensor is located; the number of valid targets is at least three;

[0047] A density determination unit is used to determine the traffic flow density of the adjacent lanes based on the perception data of the effective target.

[0048] The congestion determination unit is used to determine the traffic congestion status of the adjacent lanes based on the traffic flow density.

[0049] The technical solution provided in this application determines the lanes matched with multiple targets based on environmental perception information collected by sensors. Valid targets are then selected from the multiple targets matched with adjacent lanes. Based on the perception data of the valid targets, the traffic flow density of adjacent lanes is determined. Based on the traffic flow density, the traffic congestion status of adjacent lanes is determined. This application utilizes environmental perception data collected by sensors to determine the traffic flow density of adjacent lanes and uses this traffic flow density to determine the traffic congestion status of adjacent lanes, thereby achieving an effective judgment of the traffic congestion status of adjacent lanes along the vehicle's travel path and providing a reliable environmental perception reference for the driving behavior of autonomous vehicles. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 A flowchart illustrating a traffic congestion determination method based on vehicle flow density provided in an embodiment of this application;

[0052] Figure 2 A flowchart illustrating a method for determining targets in the left and right lanes, provided in an embodiment of this application;

[0053] Figure 3 A flowchart illustrating a method for identifying invalid targets in a lane, as provided in an embodiment of this application;

[0054] Figure 4 A flowchart illustrating an effective target discrimination method provided in an embodiment of this application;

[0055] Figure 5 A flowchart illustrating a method for determining traffic density provided in an embodiment of this application;

[0056] Figure 6 A flowchart illustrating a traffic condition determination method provided in an embodiment of this application;

[0057] Figure 7 A schematic diagram of the architecture of a traffic congestion judgment device based on vehicle flow density provided in an embodiment of this application;

[0058] Figure 8 A schematic diagram of a target reference point provided in an embodiment of this application;

[0059] Figure 9 This application provides a schematic diagram of target perception within adjacent lanes.

[0060] Figure 10 A perceptual schematic diagram of targets a, b, and c provided for an embodiment of this application;

[0061] Figure 11 A schematic diagram illustrating the perception of targets b, c, and d provided in an embodiment of this application;

[0062] Figure 12 This application provides a schematic diagram of target perception from far to near, as illustrated in an embodiment of the present application.

[0063] Figure 13 A schematic diagram illustrating the perception of an invalid target provided in an embodiment of this application;

[0064] Figure 14 A schematic diagram illustrating the perception of a remaining target provided in an embodiment of this application;

[0065] Figure 15 A schematic diagram illustrating the distribution of continuous targets provided in an embodiment of this application;

[0066] Figure 16 This is a schematic diagram of congestion point adjustment provided in an embodiment of this application. Detailed Implementation

[0067] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0068] In this application, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. The terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0069] like Figure 1 The diagram shown is a flowchart illustrating a traffic congestion judgment method based on vehicle flow density provided in an embodiment of this application, including the following steps.

[0070] S101: Based on the environmental perception information collected by sensors, determine the lanes that match multiple targets.

[0071] The target is a vehicle located within the sensor's sensing range.

[0072] In some examples, sensors can be mounted on vehicles and move in real time along with the vehicle, collecting environmental perception information in real time, including perception data of the vehicle itself and other vehicles on the road.

[0073] It is understandable that the vehicle to which the sensor belongs is also within the sensor's sensing range, and in possible implementations, all vehicles sensed by the sensor can be identified as targets.

[0074] Optionally, the process of determining the lanes matching multiple targets based on environmental perception information collected by sensors can be found in [reference needed]. Figure 2 The steps shown are accompanied by corresponding explanations.

[0075] S102: Select valid targets from multiple targets matched in adjacent lanes.

[0076] The adjacent lanes are those adjacent to the lane where the sensor is located, and the number of valid targets must be at least three. Generally speaking, the lane where the sensor is located can be understood as the lane currently occupied by the vehicle to which the sensor belongs.

[0077] In some examples, adjacent lanes include left and right lanes. In possible implementations, vehicles in adjacent lanes travel in the same direction, and the direction of travel of vehicles in adjacent lanes may be the same as or different from the direction of travel of the vehicle to which the sensor belongs.

[0078] It should be noted that in order to ensure the effectiveness of traffic congestion estimation, it is necessary to ensure the necessity and effectiveness of the targets participating in the subsequent calculation of adjacent lanes. For this purpose, it is necessary to select effective targets from the multiple targets matched by adjacent lanes.

[0079] In some examples, the necessity of an objective is reflected in whether all objectives in adjacent lanes are included in the calculation. Optionally, the process for determining necessity can be found in [link to relevant documentation]. Figure 3 The steps shown are accompanied by corresponding explanations.

[0080] In some examples, the validity of a target is reflected in whether the target is truly effective. Optionally, the process for determining validity can be found in [reference needed]. Figure 4 The steps shown are accompanied by corresponding explanations.

[0081] S103: Determine the traffic density of adjacent lanes based on the perception data of effective targets.

[0082] Traffic congestion is characterized by high vehicle density, which can be defined as the number of vehicles per unit length of road at a given moment.

[0083] In some examples, traffic density The calculation formula can be found in formula (7).

[0084] (7)

[0085] In formula (7), Represents the number of vehicles. Represents road length. Represents traffic rate, Represents the average speed of traffic flow. Represents time.

[0086] In a possible implementation, the average speed of the traffic flow This can be monitored in real time by sensors; therefore, only the perception data of the effective targets needs to be used to calculate the corresponding traffic rate. This allows us to determine the traffic density of adjacent lanes.

[0087] Optionally, after identifying the effective targets, the corresponding traffic rate is calculated using the perception data of the effective targets to further determine the traffic flow density of adjacent lanes. This process can be found in [reference needed]. Figure 5 The steps shown are accompanied by corresponding explanations.

[0088] S104: Determine the traffic congestion status of adjacent lanes based on traffic density.

[0089] Traffic congestion is a continuous scenario, so an integral method can be used to estimate the state of the traffic congestion scenario to ensure the effectiveness and continuity of traffic congestion assessment. Optionally, the implementation process of this state estimation can be found in [link to documentation]. Figure 6 The steps shown are accompanied by corresponding explanations.

[0090] Combination Figures 2-6 The method shown in this embodiment introduces a method for calculating macroscopic traffic flow density, combined with congestion integral estimation, to effectively determine whether the current driving environment (i.e., adjacent lanes) is congested. In addition to providing single-target perception for autonomous driving, it can effectively perceive the traffic flow in adjacent lanes, and then, based on the environmental perception results, it can control the entire vehicle, such as limiting the speed of autonomous driving based on the traffic congestion of adjacent lanes.

[0091] It should be noted that the method shown in this embodiment can determine not only the traffic congestion status of adjacent lanes, but also the traffic congestion status of the lane in which the vehicle is currently located.

[0092] It is understood that the process of traffic congestion judgment in this embodiment involves lane matching for targets sensed by sensors and filtering targets in adjacent lanes to select effective targets that form congestion environment characteristics. The macroscopic traffic flow density calculation method is applied to the local driving environment, thereby estimating the traffic congestion status of adjacent lanes through traffic flow density. The degree of traffic congestion is estimated by combining congestion integral. By designing different types of second specified scores, the problem of sensor recognition accuracy error is overcome, and a stable and reliable traffic congestion judgment result is obtained.

[0093] The processes shown in S101-S104 above utilize environmental perception data collected by sensors to determine the traffic flow density of adjacent lanes, and use the traffic flow density to determine the traffic congestion status of adjacent lanes, so as to achieve an effective judgment on the traffic congestion status of adjacent lanes on the vehicle's driving path and provide a reliable environmental perception reference for the driving behavior of autonomous driving.

[0094] like Figure 2 The diagram shown is a flowchart illustrating a method for determining targets in the left and right lanes according to an embodiment of this application, including the following steps.

[0095] S201: Obtain environmental perception information collected by the sensor.

[0096] Among them, environmental perception information includes perception data of multiple targets, and the perception data includes coordinate information and lane line parameters.

[0097] In some examples, the coordinate information includes the x and y coordinates of the target's reference point in the sensor coordinate system. , Represents the x-axis. Represents the vertical axis.

[0098] In a possible implementation, the reference point of the target can be the tail center point M, as detailed in [reference needed]. Figure 8 As shown.

[0099] In some examples, lane line parameters include the width of the left lane line. Heading angle curvature rate of change of curvature and the width of the right lane line Heading angle curvature rate of change of curvature .

[0100] S202: Determine the lateral position information of multiple targets based on the coordinate information of multiple targets and lane line parameters.

[0101] The lateral position information includes the lateral position reference values ​​for the left lane line, the right lane line, the left-left lane line, and the right-right lane line.

[0102] In some examples, the lateral position reference value of the left lane line The calculation formula can be found in formula (1).

[0103] (1)

[0104] In some examples, the lateral position reference value of the right lane line The calculation formula can be found in formula (2).

[0105] (2)

[0106] In some examples, the lateral position reference value of the left-left lane line The calculation formula can be found in formula (3).

[0107] (3)

[0108] In some examples, the lateral position reference value of the right lane line The calculation formula can be found in formula (4).

[0109] (4)

[0110] S203: Based on the coordinate and lateral position information of multiple targets, determine the lanes that match the multiple targets.

[0111] Among them, obtaining the reference value of the lateral position of the left lane line of the target. Reference value for the lateral position of the right lane line Reference values ​​for the lateral position of the left and right lane lines and reference values ​​for the lateral position of the right lane lines Then, the target's ordinate can be used in conjunction with this. Determine the lane that matches the target.

[0112] Optionally, the process of determining the lanes matched by multiple targets based on their coordinate and lateral position information may include: for each target, determining a corresponding target reference value based on its coordinate information; if the target reference value is greater than or equal to the lateral position reference value of the left-left lane line and less than or equal to the lateral position reference value of the left lane line, determining that the lane matched by the target is the left lane; the left lane is the lane adjacent to the left side of the lane where the sensor is located; if the target reference value is greater than or equal to the lateral position reference value of the right lane line and less than or equal to the lateral position reference value of the right-right lane line, determining that the lane matched by the target is the right lane; the right lane is the lane adjacent to the right side of the lane where the sensor is located.

[0113] It should be noted that this can be based on the target's ordinate. The target reference value is determined as the target. In a possible implementation, see formula (5). If the center point M of the tail of the target is located between the left lane line and the left-left lane line, then the target is classified as the left lane (i.e., the lane matched by the target is determined to be the left lane). See formula (6). If the center point M of the tail of the target is located between the right lane line and the right-right lane line, then the target is classified as the right lane (i.e., the lane matched by the target is determined to be the right lane).

[0114] (5)

[0115] (6)

[0116] The process described in S201-S203 above can use the coordinate information of multiple targets collected by the sensor and lane line parameters to determine the lane matched by each target, thereby determining the targets contained in the lanes on the left and right sides of the lane where the vehicle to which the sensor belongs is currently located.

[0117] like Figure 3 The diagram shown is a flowchart of a lane invalid target identification method provided in an embodiment of this application, which includes the following steps.

[0118] S301: Determine the spacing between multiple targets based on the coordinate information and index values ​​of multiple targets matched in adjacent lanes.

[0119] The index value represents the order in which a target is perceived by the sensor along a specified direction in an adjacent lane (e.g., the specified driving direction of the adjacent lane), and the spacing represents the distance between the target and the next target. The index value of the next target is listed after the target's index value, and the difference between the index value of the next target and the target's index value is 1.

[0120] In some examples, assuming the index of the target is n, then the index of the next target is n+1.

[0121] In some examples, the spacing between targets can be determined based on their x-coordinates and index values. Specifically, the spacing between target n can be expressed as... , The x-coordinate represents the target n. The x-coordinate represents the next target n+1.

[0122] S302: Determine the first target from multiple targets whose spacing is greater than the preset maximum longitudinal distance threshold.

[0123] The preset maximum longitudinal distance threshold can be denoted as: Generally speaking, if the distance of target n Greater than If so, then target n is determined as the first target.

[0124] S303: Determine the corresponding invalid target based on the index value of the first target.

[0125] In this process, after determining one or more first targets, the index value of each first target can be determined first. Then, by using the index value of each first target and combining it with the total number of targets matched by adjacent lanes (i.e., the total number of targets contained in adjacent lanes), the corresponding invalid targets can be determined.

[0126] Optionally, the process of determining the corresponding invalid target based on the index value of the first target can be as follows: if the index value of the first target is greater than the target value, other targets whose index values ​​are ranked after the first target are determined as the corresponding invalid targets; the target value is determined based on the total number of targets matched by adjacent lanes; if the index value of the first target is less than or equal to the target value, other targets whose index values ​​are ranked before the first target are determined as the corresponding invalid targets.

[0127] In some examples, the total number of targets matched by adjacent lanes is denoted as N, and the target value can be denoted as... .

[0128] In a possible implementation, the process of determining the implementation of the corresponding invalid target based on the index value of the first target can be summarized as follows: If so, then the first to the nth targets in the adjacent lanes will be determined as invalid targets; If so, the (n+1)th target to the last target in the adjacent lane will be determined as invalid targets.

[0129] See in some examples Figure 9 As shown, three consecutive near-end targets (target a, target b, and target c) constitute a basic congestion scenario. If the sensor identifies a target d at a distant end, such as... Figure 10 As shown, including target d in the calculation is equivalent to increasing the following distance of the three near-end congested targets (including target a, target b, and target c), making it impossible to effectively identify near-end congestion. Similarly, see... Figure 11 As shown, if the near target a is included in the calculation, the three congested targets at the far end (i.e., target a, target c and target d) will be ignored, resulting in the inability to effectively identify far-end congestion.

[0130] and The use of this method can be seen as a dichotomy approach: if the first target is detected in the first half of the target group matched by the adjacent lane, the targets in the first half can be considered invalid targets; if the first target is detected in the second half of the target group matched by the adjacent lane, the targets in the second half can be considered invalid targets.

[0131] In some examples, if there are multiple first targets, the priority of the multiple first targets can be determined in descending order of spacing. Then, based on the index value of each first target, the invalid targets corresponding to each first target can be determined in descending order of priority.

[0132] In possible implementations, see Figure 12 As shown, the process starts traversing from the furthest target f in the adjacent lane. First, the maximum distance between target d and target e is detected (i.e., target e is determined as the first target). Since target e is in the first half, target e and target f are considered invalid targets for easy filtering. Then, the maximum distance between target b and target a is detected (i.e., target b is determined as the first target). Since target n is in the second half, target a is considered an invalid target for easy filtering.

[0133] In a possible implementation, if the number of remaining targets in adjacent lanes, excluding invalid targets, is three, see [reference needed]. Figure 13 As shown, the remaining targets can be considered valid targets (i.e., target a, target b, and target c). If the number of remaining targets in adjacent lanes, excluding invalid targets, is less than three, see [reference needed]. Figure 14As shown, the remaining targets can be considered invalid targets, and the traffic congestion status of the adjacent lane will no longer be judged. The traffic congestion status of the adjacent lane can be directly determined as not congested.

[0134] S304: Identify valid targets among multiple targets matched by adjacent lanes, excluding invalid targets.

[0135] After identifying invalid targets in adjacent lanes, these invalid targets can be directly filtered out, and the remaining targets are considered valid targets. However, some remaining targets may be irregularly shaped vehicles, such as vans, causing a single target to be identified as two partially overlapping targets, leading to misjudgments in traffic congestion assessment. Therefore, further assessment of the remaining targets is necessary. Optionally, this further assessment process can be found in [link to documentation]. Figure 4 The steps shown are accompanied by corresponding explanations.

[0136] The processes shown in S301-S304 above utilize the coordinate information and index value of targets in adjacent lanes, combined with the total number of targets matched in adjacent lanes, to determine invalid targets in adjacent lanes, thereby increasing the necessity of valid targets and improving the reliability of traffic congestion situation determination results.

[0137] like Figure 4 The diagram shown is a flowchart of an effective target discrimination method provided in an embodiment of this application, including the following steps.

[0138] S401: Determine the perception data of multiple candidate targets, excluding invalid targets, among multiple targets matched by adjacent lanes.

[0139] The perception data includes coordinate information and the length of the entire vehicle.

[0140] S402: Based on the coordinate information of multiple candidate targets and the length of the vehicle, determine the distance between the front ends of multiple candidate targets.

[0141] Here, the distance between the front of the vehicle and the rear of the next candidate target represents the distance between the front of the candidate target and the rear of the next candidate target. The index value of the next candidate target is placed after the candidate target, and the difference between the index value of the next candidate target and the index value of the candidate target is 1.

[0142] In some examples, the distance to the front of the candidate target can be denoted as... , The x-coordinate represents the candidate target n. The x-coordinate represents the next candidate target n+1. The total vehicle length represents candidate target n; see details below. Figure 15 As shown.

[0143] S403: For each candidate target, if the distance between the front of the candidate target and the vehicle is greater than or equal to the preset minimum longitudinal distance threshold, the candidate target is determined as a valid target.

[0144] The preset minimum longitudinal distance threshold can be denoted as: Understandably, if the distance to the front of the candidate target is less than... If the candidate target is filtered out, it will no longer be considered a valid target.

[0145] In some examples, the process of filtering candidate targets using a preset minimum vertical distance threshold can be summarized as follows: If the probability of the first overlapping target being true is relatively high, then the second overlapping target will be filtered out.

[0146] The processes shown in S401-S403 above use the coordinate information of the candidate target and the length of the vehicle to determine the distance to the front of the candidate target, and use the distance to identify whether the candidate target is a valid target, so as to improve the effectiveness of the valid target and thus improve the reliability of the traffic congestion situation determination result.

[0147] like Figure 5 The diagram shown is a flowchart illustrating a method for determining traffic density according to an embodiment of this application, including the following steps.

[0148] S501: Determine the perception data for at least three valid targets.

[0149] The perception data includes coordinate information, vehicle length, and vehicle speed.

[0150] S502: Based on the coordinate information of each valid target, the overall vehicle length, and the vehicle speed, determine the headway of each valid target.

[0151] The process of determining the headway of the effective target by using the coordinate information of the effective target, the length of the vehicle and the speed of the vehicle can be found in formula (8).

[0152] (8)

[0153] In formula (8), The headway of the train represents the effective target. Index representing valid targets, The x-coordinate represents the effective target n. The x-coordinate represents the next valid target n+1. The total length of the vehicle represents the effective target n. The vehicle speed represents the effective target n.

[0154] S503: Determine the traffic density of adjacent lanes based on the arithmetic mean of the headway of each vehicle.

[0155] In this process, after determining the headway of each valid target, the traffic rate is calculated using the headway. Then based on traffic rate and average speed of traffic flow Determine the traffic density of adjacent lanes.

[0156] In some examples, the traffic rate is calculated using the headway of each vehicle. The implementation process can be seen in formula (9).

[0157] (9)

[0158] In formula (9), This represents the arithmetic mean of the headway distances of each locomotive.

[0159] It is important to note that, based on the understanding of traffic congestion, the traffic flow density of adjacent lanes is only calculated when there are at least three valid targets in adjacent lanes. If there are only two targets in adjacent lanes, there is only one time interval between them, and therefore no concept of an arithmetic mean. For this reason, this embodiment applies the macroscopic traffic flow state estimation method to the microscopic autonomous vehicle operating environment, and evaluates the traffic flow state within the sensor's perception range by calculating the traffic flow density.

[0160] The processes shown in S501-S503 above use the coordinate information of each valid target, the length of the vehicle, and the speed of the vehicle to determine the headway of each valid target, and use the headway of each vehicle to determine the traffic density of adjacent lanes, thereby realizing real-time tracking and calculation of traffic density.

[0161] like Figure 6 The diagram shown is a flowchart of a traffic condition determination method provided in an embodiment of this application, including the following steps.

[0162] S601: Determine the corresponding congestion score based on the traffic flow of adjacent lanes.

[0163] Among them, traffic flow is determined based on the total number of targets matched by adjacent lanes within a specified time, and congestion score is a quantitative indicator for measuring the degree of traffic congestion.

[0164] S602: When the traffic density is greater than the preset density threshold, the congestion score is increased according to the specified slope so that the increased congestion score is higher than the original congestion score by a first specified number of points.

[0165] Specifically, when the congestion score is greater than the first congestion benchmark, the traffic congestion status of the adjacent lane is determined to be congested; when the congestion score is less than the second congestion benchmark, the traffic congestion status of the adjacent lane is determined to be uncongested. The first congestion benchmark is greater than the second congestion benchmark.

[0166] In some examples, the first congestion calibration value is denoted as The second congestion benchmark is denoted as , as well as Form the corresponding hysteresis interval When the congestion score is greater than The traffic congestion status of adjacent lanes is congested, when the congestion score is less than 1. The traffic congestion status of the adjacent lane is not congested, or the traffic congestion status of the adjacent lane has been relieved from congestion to smooth.

[0167] In a possible implementation, the first specified score can be denoted as This can be understood as the slope of the congestion integral increase per unit step (i.e., a preset calculation step, such as 0.2 seconds). When the traffic density is greater than a preset density threshold, the congestion integral increases along a specified slope. For details, please refer to Figure 16 As shown ( Figure 16 shown (Represents the congestion score before adjustment).

[0168] S603: When the traffic density is less than or equal to the preset density threshold, the congestion score is reduced according to the specified slope so that the congestion score after the reduction is lower than the congestion score before the reduction by a second specified number of points.

[0169] There are generally two reasons for a decrease in traffic density: one is the reduction in traffic density due to the easing of congestion, and the other is the decrease in traffic density due to sensor recognition errors, resulting in a temporary loss of target identification within the traffic flow (i.e., there is an error in the traffic flow). Based on the different reasons for the decrease in traffic density, two types of decrease slopes can be set. Therefore, the type of the second specified score includes a first score and a second score, with the first score being less than the second score.

[0170] Optionally, when the traffic density is less than or equal to a preset density threshold and the traffic volume meets the preset threshold, the congestion score is reduced by a specified slope so that the reduced congestion score is lower than the original congestion score by a second score; when the traffic density is less than or equal to a preset density threshold and the traffic volume does not meet the preset threshold, the congestion score is reduced by a specified slope so that the reduced congestion score is lower than the original congestion score by a first score.

[0171] In some examples, the first fraction can be denoted as The second fraction can be denoted as Both the first and second fractions can be considered as the slope of the congestion integral decrease per unit step. For application scenarios of the first and second fractions, please refer to... Figure 16 As shown.

[0172] The processes shown in S601-S603 above use traffic flow density to adjust the congestion score, and use the adjusted congestion score to determine the traffic congestion status of adjacent lanes, thereby obtaining a stable and reliable traffic congestion judgment result.

[0173] like Figure 7 The diagram shown is a schematic representation of the architecture of a traffic congestion judgment device based on vehicle flow density provided in an embodiment of this application, including the following units.

[0174] The lane matching unit 100 is used to determine the lanes to be matched for multiple targets based on environmental perception information collected by sensors; the targets are vehicles located within the sensor's perception range.

[0175] Optionally, the lane matching unit 100 is specifically used to: obtain environmental perception information collected by sensors; the environmental perception information includes perception data of multiple targets; the perception data includes coordinate information and lane line parameters; based on the coordinate information and lane line parameters of multiple targets, determine the lateral position information of multiple targets; the lateral position information includes the lateral position reference value of the left lane line, the lateral position reference value of the right lane line, the lateral position reference value of the left-left lane line, and the lateral position reference value of the right-right lane line; based on the coordinate information and lateral position information of multiple targets, determine the lane matched to the multiple targets.

[0176] Optionally, the lane matching unit 100 is specifically used to: for each target, determine the corresponding target reference value based on the target's coordinate information; if the target reference value is greater than or equal to the left-left lane line lateral position reference value and less than or equal to the left lane line lateral position reference value, determine that the lane matched by the target is the left lane; the left lane is the lane adjacent to the left side of the lane where the sensor is located; if the target reference value is greater than or equal to the right lane line lateral position reference value and less than or equal to the right-right lane line lateral position reference value, determine that the lane matched by the target is the right lane; the right lane is the lane adjacent to the right side of the lane where the sensor is located.

[0177] The target filtering unit 200 is used to filter out valid targets from multiple targets matched in adjacent lanes; the adjacent lanes are the lanes adjacent to the lane where the sensor is located; the number of valid targets is at least three.

[0178] Optionally, the target filtering unit 200 is specifically used to: determine the spacing between multiple targets based on the coordinate information and index values ​​of multiple targets matched in adjacent lanes; wherein, the index value represents the order in which the target is perceived by the sensor along a specified direction of the adjacent lane; the spacing represents the distance between the target and the next target; the index value of the next target is placed after the target, and the difference between the index value of the next target and the index value of the target is 1; determine a first target whose spacing is greater than a preset maximum longitudinal distance threshold from the multiple targets; determine the corresponding invalid target based on the index value of the first target; and determine the valid targets other than the invalid targets among the multiple targets matched in adjacent lanes.

[0179] Optionally, the target filtering unit 200 is specifically used to: if the index value of the first target is greater than the target value, determine the other targets after the first target based on their index values ​​as the corresponding invalid targets; the target value is determined based on the total number of targets matched in adjacent lanes; if the index value of the first target is less than or equal to the target value, determine the first target and the other targets before the first target based on their index values ​​as the corresponding invalid targets.

[0180] Optionally, the target filtering unit 200 is specifically used to: determine the perception data of multiple candidate targets (excluding invalid targets) among multiple targets matched in adjacent lanes; the perception data includes coordinate information and vehicle length; based on the coordinate information and vehicle length of multiple candidate targets, determine the head-to-head distance of multiple candidate targets; the head-to-head distance represents the distance between the front of a candidate target and the rear of the next candidate target; the index value of the next candidate target is ranked after the candidate target, and the difference between the index value of the next candidate target and the index value of the candidate target is 1; for each candidate target, if the head-to-head distance of the candidate target is greater than or equal to a preset minimum longitudinal distance threshold, the candidate target is determined as a valid target.

[0181] The density determination unit 300 is used to determine the traffic flow density of adjacent lanes based on the perception data of effective targets.

[0182] Optionally, the density determination unit 300 is specifically used to: determine the perception data of at least three valid targets; the perception data includes coordinate information, vehicle length, and vehicle speed; determine the headway of each valid target based on the coordinate information, vehicle length, and vehicle speed; and determine the traffic density of adjacent lanes based on the arithmetic mean of the headway of each valid target.

[0183] The congestion determination unit 400 is used to determine the traffic congestion status of adjacent lanes based on traffic flow density.

[0184] Optionally, the congestion determination unit 400 is specifically used for: determining the corresponding congestion score based on the traffic flow of adjacent lanes; the traffic flow is determined based on the total number of targets matched by adjacent lanes within a specified time; the congestion score is a quantitative indicator for measuring the degree of traffic congestion; when the traffic density is greater than a preset density threshold, the congestion score is increased by a specified slope so that the increased congestion score is increased by a first specified score compared to the original congestion score; when the traffic density is less than or equal to the preset density threshold, the congestion score is decreased by a specified slope so that the decreased congestion score is decreased by a second specified score compared to the original congestion score; wherein, when the congestion score is greater than a first congestion calibration value, the traffic congestion status of the adjacent lanes is determined to be congested; when the congestion score is less than a second congestion calibration value, the traffic congestion status of the adjacent lanes is determined to be uncongested; the first congestion calibration value is greater than the second congestion calibration value.

[0185] Optionally, the type of the second specified score includes a first score and a second score, where the first score is less than the second score. The congestion determination unit 400 is specifically used to: when the traffic density is less than or equal to a preset density threshold and the traffic volume meets the preset threshold, adjust the congestion score downwards according to a specified slope so that the adjusted congestion score is lower than the original congestion score by a second score; when the traffic density is less than or equal to a preset density threshold and the traffic volume does not meet the preset threshold, adjust the congestion score downwards according to a specified slope so that the adjusted congestion score is lower than the original congestion score by a first score.

[0186] The units described above use environmental perception data collected by sensors to determine the traffic density of adjacent lanes, and use the traffic density to determine the traffic congestion status of adjacent lanes, so as to effectively judge the traffic congestion status of adjacent lanes on the vehicle's driving path and provide a reliable environmental perception reference for the driving behavior of autonomous driving.

[0187] This application also provides a computer-readable storage medium including a stored program, wherein the program executes the traffic congestion judgment method based on vehicle flow density provided in this application.

[0188] This application also provides a vehicle, including a processor, a memory, and a bus. The processor and the memory are connected via the bus. The memory is used to store a program, and the processor is used to run the program. When the program runs, it executes the traffic congestion judgment method based on vehicle flow density provided in this application.

[0189] Furthermore, the functions described above in the embodiments of this application can be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that can be used include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SOCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0190] While several specific implementation details are included in the foregoing discussion, these should not be construed as limiting the scope of this application. Certain features described in the context of individual embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented individually or in any suitable sub-combination in multiple embodiments.

[0191] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of disclosure in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this application.

Claims

1. A method for judging traffic congestion based on vehicle flow density, characterized in that, include: Based on the environmental perception information collected by sensors, the lanes matching multiple targets are determined; The target is a vehicle located within the sensing range of the sensor; Valid targets are selected from multiple targets matched in adjacent lanes; the adjacent lanes are those adjacent to the lane where the sensor is located. The number of valid targets is at least three; wherein, the process of selecting valid targets from multiple targets matched by adjacent lanes includes: determining the spacing between multiple targets based on the coordinate information and index values ​​of multiple targets matched by adjacent lanes; wherein, the index value represents the order in which the target is perceived by the sensor along a specified direction of the adjacent lane; the spacing represents the distance between the target and the next target; the index value of the next target is ranked after the target, and the difference between the index value of the next target and the index value of the target is 1; determining a first target whose spacing is greater than a preset maximum longitudinal distance threshold from multiple targets; if the index value of the first target is greater than a target value, determining other targets whose index values ​​are ranked after the first target as corresponding invalid targets; the target value is determined based on the total number of targets matched by adjacent lanes; if the index value of the first target is less than or equal to the target value, determining other targets whose index values ​​are ranked before the first target as corresponding invalid targets; determining valid targets other than the invalid targets from multiple targets matched by adjacent lanes; Based on the perception data of the effective target, the traffic density of the adjacent lane is determined; Based on the traffic density, the traffic congestion status of the adjacent lanes is determined.

2. The method according to claim 1, characterized in that, Based on environmental perception information collected by sensors, lanes matching multiple targets are determined, including: The system acquires environmental perception information collected by sensors; the environmental perception information includes perception data of multiple targets; the perception data includes coordinate information and lane line parameters. Based on the coordinate information of multiple targets and lane line parameters, the lateral position information of multiple targets is determined; the lateral position information includes the lateral position reference value of the left lane line, the lateral position reference value of the right lane line, the lateral position reference value of the left-left lane line, and the lateral position reference value of the right-right lane line. Based on the coordinate and lateral position information of the multiple targets, the lanes that match the multiple targets are determined.

3. The method according to claim 2, characterized in that, Based on the coordinate and lateral position information of multiple targets, determine the lanes that match the multiple targets, including: For each target, a corresponding target reference value is determined based on the target's coordinate information; If the target reference value is greater than or equal to the left-left lane line lateral position reference value and less than or equal to the left lane line lateral position reference value, the lane matched by the target is determined to be the left lane; the left lane is the lane adjacent to the left side of the lane where the sensor is located. If the target reference value is greater than or equal to the right lane line lateral position reference value and less than or equal to the right lane line lateral position reference value, the lane matched by the target is determined to be the right lane; the right lane is the lane adjacent to the right side of the lane where the sensor is located.

4. The method according to claim 1, characterized in that, Determining valid targets (excluding invalid targets) from among multiple targets matched by the adjacent lanes includes: The perception data of multiple candidate targets, excluding the invalid targets, are determined from among the multiple targets matched by the adjacent lanes; the perception data includes coordinate information and vehicle length. Based on the coordinate information of multiple candidate targets and the overall vehicle length, the distance between the front ends of multiple candidate targets is determined; the distance between the front ends of a candidate target and the rear ends of the next candidate target is represented; the index value of the next candidate target is placed after the candidate target, and the difference between the index value of the next candidate target and the index value of the candidate target is 1; For each candidate target, if the distance between the front of the candidate target and the vehicle is greater than or equal to a preset minimum longitudinal distance threshold, the candidate target is determined as a valid target.

5. The method according to claim 1, characterized in that, Based on the perception data of the effective target, the traffic flow density of the adjacent lanes is determined, including: Determine perception data for at least three of the effective targets; the perception data includes coordinate information, vehicle length, and vehicle speed. Based on the coordinate information of each of the effective targets, the length of the vehicle and the speed of the vehicle, the headway of each of the effective targets is determined. The traffic density of the adjacent lanes is determined based on the arithmetic mean of the headway distances of each vehicle.

6. The method according to claim 1, characterized in that, Based on the traffic density, the traffic congestion status of the adjacent lanes is determined, including: Based on the traffic flow of the adjacent lanes, a corresponding congestion score is determined; the traffic flow is determined based on the total number of targets matched by the adjacent lanes within a specified time period; the congestion score is a quantitative indicator for measuring the degree of traffic congestion. When the traffic density is greater than a preset density threshold, the congestion score is adjusted upward according to a specified slope, so that the congestion score after adjustment is increased by a first specified score compared with the congestion score before adjustment. When the traffic density is less than or equal to the preset density threshold, the congestion score is reduced according to the specified slope, so that the congestion score after reduction is lower than the congestion score before reduction by a second specified score. Specifically, when the congestion score is greater than the first congestion calibration value, the traffic congestion status of the adjacent lane is determined to be congested; when the congestion score is less than the second congestion calibration value, the traffic congestion status of the adjacent lane is determined to be uncongested; the first congestion calibration value is greater than the second congestion calibration value.

7. The method according to claim 6, characterized in that, The second specified score can be of the type of a first score and a second score, where the first score is less than the second score. When the traffic density is less than or equal to the preset density threshold, the congestion score is adjusted downwards according to the specified slope, so that the adjusted congestion score is lower than the original congestion score by a second specified score, including: When the traffic density is less than or equal to the preset density threshold and the traffic volume meets the preset threshold, the congestion score is reduced according to the specified slope so that the reduced congestion score is lower than the original congestion score by the second score. When the traffic density is less than or equal to the preset density threshold and the traffic volume does not meet the preset threshold, the congestion score is adjusted downward according to the specified slope, so that the congestion score after adjustment is lower than the congestion score before adjustment by the first score.

8. A traffic congestion judgment device based on vehicle flow density, characterized in that, include: The lane matching unit is used to determine the lanes that match multiple targets based on environmental perception information collected by sensors. The target is a vehicle located within the sensing range of the sensor; The target filtering unit is used to filter out valid targets from multiple targets matched in adjacent lanes; the adjacent lanes are the lanes adjacent to the lane where the sensor is located. The number of valid targets is at least three; wherein, the process of selecting valid targets from multiple targets matched by adjacent lanes includes: determining the spacing between multiple targets based on the coordinate information and index values ​​of multiple targets matched by adjacent lanes; wherein, the index value represents the order in which the target is perceived by the sensor along a specified direction of the adjacent lane; the spacing represents the distance between the target and the next target; the index value of the next target is ranked after the target, and the difference between the index value of the next target and the index value of the target is 1; determining a first target whose spacing is greater than a preset maximum longitudinal distance threshold from multiple targets; if the index value of the first target is greater than a target value, determining other targets whose index values ​​are ranked after the first target as corresponding invalid targets; the target value is determined based on the total number of targets matched by adjacent lanes; if the index value of the first target is less than or equal to the target value, determining other targets whose index values ​​are ranked before the first target as corresponding invalid targets; determining valid targets other than the invalid targets from multiple targets matched by adjacent lanes; A density determination unit is used to determine the traffic flow density of the adjacent lanes based on the perception data of the effective target. The congestion determination unit is used to determine the traffic congestion status of the adjacent lanes based on the traffic flow density.

Images