Vehicle interested target confidence coefficient determination method and device and intelligent driving vehicle

Abstract

The invention provides a vehicle interested target confidence coefficient determination method and apparatus, and an intelligent driving vehicle. The method comprises the steps of obtaining a current position of an interested target in a navigation area of a vehicle; according to a preset navigation area confidence coefficient and the current position, determining an area confidence coefficient of the interested target; according to the current position and the navigation path of the interested target, the close degree of the interested target and the navigation path and the probability that the interested target invades the target area are respectively determined; and carrying out weighted summation on the region confidence of the interested target, the close degree of the interested target and the navigation path, and the probability that the interested target invades the target region to obtain the fusion confidence of the interested target. The final confidence degree of the interested target is determined through fusion calculation of the region confidence degree, the close degree of the region confidence degree and the navigation path and the probability of invading the target region, and the key target can be obtained more accurately.

Description

Technical Field

[0001] The present application relates to the field of vehicle intelligent driving technology, and in particular to a method and device for determining the confidence level of a vehicle target of interest, and an intelligent driving vehicle. Background Art

[0002] The complex and ever-changing environment in which intelligent driving vehicles travel poses a huge challenge to perception technology. How to correctly evaluate the targets of interest in the environment has also become a problem that intelligent driving vehicles must solve. The existing solution to this problem is to use the navigation area of ​​the vehicle to make judgments. The targets in the navigation area have high confidence, and the targets can be divided into two categories: credible and untrustworthy. However, as the distance of the target becomes farther, the accuracy of the navigation area will also decrease. Therefore, a single evidence navigation area is used to solve the confidence. The solution result will be affected by environmental changes, sensor noise, and tracking errors, reducing accuracy. Summary of the invention

[0003] The purpose of this application is to provide a method, device and intelligent driving vehicle for determining the confidence of a vehicle target of interest. The final confidence of the target of interest is determined by a fusion calculation of three aspects: regional confidence, closeness to the navigation path, and probability of intrusion into the target area, so that key targets can be obtained more accurately.

[0004] In a first aspect, the present application provides a method for determining the confidence of a target of interest of a vehicle, the method comprising: obtaining the current position of a target of interest within a navigation area of ​​the vehicle; the navigation area is obtained by expanding the navigation path through a lateral safety distance; determining the regional confidence of the target of interest based on a preset navigation area confidence and the current position; the preset navigation area confidence comprises: preset confidences corresponding to a plurality of sub-areas determined by dividing the navigation area; determining the degree of proximity of the target of interest to the navigation path and the probability of the target of interest invading a target area based on the current position of the target of interest and the navigation path; the target area is obtained by expanding the navigation path through a preset distance; performing weighted summation of the regional confidence of the target of interest, the degree of proximity of the target of interest to the navigation path, and the probability of the target of interest invading the target area to obtain a fused confidence of the target of interest.

[0005] Furthermore, the above step of determining the regional confidence of the target of interest based on the preset navigation area confidence and the current position includes: determining the target sub-area to which the current position belongs in the navigation area; and using the preset confidence corresponding to the target sub-area as the regional confidence of the target of interest.

[0006] Furthermore, the confidence of the above-mentioned preset navigation area is determined as follows: the navigation path is a curve with the current position of the own vehicle as the starting point and the radius as the turning radius of the own vehicle; it is assumed that the preset vehicle travels along the navigation path and the own vehicle travels along the specified straight line; the specified straight line is the tangent of the navigation path at the starting point; according to the turning radius of the own vehicle, the width of the own vehicle, and the width of the preset vehicle, the first intersection position and the second intersection position are calculated; wherein; the first intersection position is: when the upper right corner of the own vehicle collides with the rear center of the preset vehicle, the intersection position of the turning radius at the rear center of the preset vehicle and the specified straight line; the second intersection position is: when the upper right corner of the own vehicle collides with the rear center of the preset vehicle, the intersection position of the turning radius at the rear center of the preset vehicle and the specified straight line; When the lower left corners of the preset vehicles collide, the intersection position of the turning radius of the rear center position of the preset vehicle and the specified straight line is preset; starting from the starting point, the area in the navigation area is determined when the width of the vehicle is expanded along the navigation path to the first intersection position as the first sub-area; the area determined when the sum of the width of the vehicle and the preset vehicle width is expanded along the navigation path to the second intersection position is obtained by removing the first sub-area; the other remaining areas are taken as the third sub-area; a corresponding preset confidence is set for each sub-area; the preset confidences corresponding to the first sub-area, the second sub-area, and the third sub-area are reduced in sequence.

[0007] Further, the intersection position is characterized by a distance from the intersection to the starting point; and according to the turning radius of the vehicle, the width of the vehicle, and the preset vehicle width, the first intersection position and the second intersection position are calculated, including: calculating the distance from the first intersection to the starting point according to the following first specified formula:

[0008] D1=sqrt(2*R2*W1+W1 2 );

[0009] Where D1 represents the distance from the first intersection point to the starting point; R2 represents the turning radius of the vehicle; W1 represents half of the width of the vehicle;

[0010] Calculate the distance from the second intersection point to the starting point using the following second specified formula:

[0011] D2=sqrt(2*R2*W3+W3 2 ); W3 = W1 + W2;

[0012] Wherein, D2 represents the distance from the second intersection point to the starting point; W2 represents half of the preset vehicle width.

[0013] Furthermore, the above-mentioned determining the degree of proximity of the target of interest to the navigation path based on the current position of the target of interest and the navigation path includes: determining the vertical distance from the current position of the target of interest to the navigation path; determining the degree of proximity of the target of interest to the navigation path based on the lateral safety distance and the vertical distance from the current position of the target of interest to the navigation path.

[0014] Further, the determining of the closeness of the target of interest to the navigation path according to the lateral safety distance and the vertical distance from the current position of the target of interest to the navigation path includes: determining the closeness β of the target of interest to the navigation path according to the following third specified formula:

[0015] β=max((W4-y) / W4,0);

[0016] Wherein, W4 represents the lateral safety distance reserved for the vehicle to travel along the pilot path; y represents the vertical distance from the current position of the target of interest to the pilot path.

[0017] Furthermore, the preset distance is determined by the width of the vehicle, the width of the target of interest and the reserved safety distance; the probability of the target of interest invading the target area is determined according to the current position and the navigation path, including: determining the left and right boundaries of the target area, and the vertical distance from the current position of the target of interest to the navigation path; respectively calculating the difference between the left and right boundaries and the vertical distance; substituting the difference and the tracking standard deviation of the target of interest into the complementary error function, and calculating the first probability that the target of interest is on the left side of the right boundary and the second probability that it is on the left side of the left boundary; and calculating the difference between the first probability and the second probability to obtain the probability of the target of interest invading the target area.

[0018] Furthermore, the above method also includes: determining whether the fusion confidence of the target of interest exceeds a preset threshold; if so, determining the target of interest as a key target; if there are multiple key targets determined, selecting the target of interest closest to the vehicle as the key target.

[0019] In a second aspect, the present application also provides a device for determining the confidence of a vehicle target of interest, the device comprising: a position acquisition module, used to obtain the current position of the target of interest within the navigation area of ​​the vehicle; the navigation area is obtained by expanding the navigation path through a lateral safety distance; a regional confidence determination module, used to determine the regional confidence of the target of interest based on a preset navigation area confidence and the current position; the preset navigation area confidence includes: preset confidences corresponding to multiple sub-areas determined by dividing the navigation area; a proximity and probability determination module, used to determine the proximity of the target of interest to the navigation path and the probability of the target of interest invading the target area according to the current position of the target of interest and the navigation path; the target area is obtained by expanding the navigation path through a preset distance; a fusion confidence determination module, used to perform weighted summation of the regional confidence of the target of interest, the proximity of the target of interest to the navigation path, and the probability of the target of interest invading the target area, to obtain the fusion confidence of the target of interest.

[0020] In a third aspect, the present application also provides an intelligent driving vehicle, comprising a processor and a memory, wherein the memory stores computer executable instructions that can be executed by the processor, and the processor executes the computer executable instructions to implement the vehicle object of interest confidence determination method described in any one of the first aspects.

[0021] In the method, device and intelligent driving vehicle for determining the confidence of the target of interest of the vehicle provided by the present application, the current position of the target of interest in the navigation area of ​​the vehicle is first obtained; the navigation area is obtained by expanding the navigation path through the lateral safety distance; then the regional confidence of the target of interest is determined according to the preset navigation area confidence and the current position; the preset navigation area confidence includes: the preset confidence corresponding to the multiple sub-areas determined by dividing the navigation area; then according to the current position of the target of interest and the navigation path, the degree of closeness of the target of interest to the navigation path and the probability of the target of interest invading the target area are determined respectively; the target area is obtained by expanding the navigation path through the preset distance; finally, the regional confidence of the target of interest, the degree of closeness of the target of interest to the navigation path, and the probability of the target of interest invading the target area are weighted and summed to obtain the fusion confidence of the target of interest. This method determines the final confidence of the target of interest through the fusion calculation of the regional confidence, the degree of closeness to the navigation path, and the probability of invading the target area, so as to obtain the key target more accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0023] Figure 1 A flowchart of a method for determining the confidence level of a vehicle target of interest provided in an embodiment of the present application;

[0024] Figure 2 A schematic diagram of solving the confidence level of a navigation area provided in an embodiment of the present application;

[0025] Figure 3 A schematic diagram of another navigation area confidence solution provided in an embodiment of the present application;

[0026] Figure 4 A schematic diagram of solving the degree of closeness of a navigation path provided in an embodiment of the present application;

[0027] Figure 5 A schematic diagram of solving the probability of a target invading the left and right boundaries provided in an embodiment of the present application;

[0028] Figure 6 A structural block diagram of a device for determining confidence of a vehicle target of interest provided in an embodiment of the present application;

[0029] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0030] The technical solution of the present application will be described clearly and completely below in conjunction with the embodiments. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0031] At present, in the field of intelligent driving technology, the common way to evaluate the target of interest in the environment is to use the navigation area of ​​the vehicle to judge. The targets in the navigation area have high confidence, and the targets can be divided into two categories: credible and untrustworthy. However, as the distance of the target becomes farther, the accuracy of the navigation area will also decrease, so the credible targets also need to be evaluated for specific confidence; in addition, the single use of the navigation area will cause a waste of resources of the target attributes, and the combination of the target and the navigation path can also provide evidence for solving the target confidence; furthermore, the use of the left and right boundaries of the navigation area only considers the position attribute of the target, ignoring the noise of the sensor and the use of the standard deviation in the process of tracking the target, which will also cause a waste of target resources; finally, most of the existing technologies use a single piece of evidence to solve the confidence, and the solution result will be affected by environmental changes, sensor noise, and tracking errors, reducing accuracy.

[0032] Among them, the above-mentioned targets of interest refer to targets that may affect the driving behavior of the vehicle during the autonomous driving process, that is, targets that may cause the autonomous driving vehicle to take actions such as acceleration, deceleration, steering and following. There can be zero or more targets. The key target refers to the target that is most likely to affect the driving behavior of the autonomous driving vehicle among all the targets of interest. The navigation area is the area covered by the path that may be predicted in the future based on the current vehicle motion state and driving environment; the confidence is the target reliability degree 0-1, and the larger the value, the higher the accuracy; evidence refers to various parameters used in calculating confidence, such as the navigation area, navigation path, target attributes, etc.

[0033] In summary, the existing technology has the following shortcomings: first, the single confidence division of the navigation area cannot provide evidence for solving the confidence of the target of interest; second, the attributes of the target of interest are not fully utilized; third, the probability of solving the target invading the left and right boundaries does not consider the target noise and standard error; fourth, the low accuracy caused by the single-dimensional evidence confidence solution of the target of interest.

[0034] Based on this, the embodiments of the present application provide a method, device and intelligent driving vehicle for determining the confidence of a vehicle target of interest. The final confidence of the target of interest is determined by a fusion calculation of three aspects: regional confidence, closeness to the navigation path, and probability of intrusion into the target area, so that key targets can be obtained more accurately.

[0035] To facilitate understanding of this embodiment, a method for determining the confidence level of a vehicle target of interest disclosed in an embodiment of the present application is first introduced in detail.

[0036] Figure 1 A flow chart of a method for determining the confidence level of a vehicle target of interest provided in an embodiment of the present application, the method specifically comprising the following steps:

[0037] Step S102, obtain the current position of the target of interest in the navigation area of ​​the vehicle; the navigation area is obtained by expanding the navigation path through the lateral safety distance; the navigation path refers to the predicted future travel path of the vehicle, which can be given by the path planning software or calculated based on the vehicle status such as speed, yaw rate, steering wheel angle and other information. The embodiment of the present application does not specifically expand on it, nor does it limit it. The navigation area is obtained by expanding the navigation path through the lateral safety distance. The specific expansion distance is determined based on the width of the vehicle, the preset width of the target of interest (the same as the preset vehicle width described later) and the reserved safety distance. For example, the lateral safety distance can be half of the width of the vehicle or a certain set value, and the specific value is not limited.

[0038] Here, there may be one or more targets of interest in the navigation area. If there are more than one target of interest, the following calculation steps are performed for each target of interest to determine the fusion confidence corresponding to each target of interest.

[0039] Step S104, according to the preset navigation area confidence and the current position, the regional confidence of the target of interest is determined; the preset navigation area confidence includes: preset confidences corresponding to multiple sub-areas determined by dividing the navigation area; the multiple sub-areas are determined according to the position change relationship between the vehicle and the preset vehicle, and the preset vehicle has a greater impact on the driving of the vehicle in which area, and the preset confidence of the area is high. The preset confidence corresponding to the area where the current position of the target of interest is located is determined as the regional confidence of the target of interest.

[0040] Step S106, according to the current position of the target of interest and the navigation path, respectively determine the proximity of the target of interest to the navigation path and the probability of the target of interest invading the target area; the target area is obtained by expanding the navigation path by a preset distance.

[0041] The above preset distance is usually smaller than the lateral safety distance, that is, the target area is smaller than the navigation area. The probability of the target of interest invading the target area and the degree of proximity of the target of interest to the navigation path are also used as evidence to judge the confidence of the target of interest, which can further improve the accuracy of the confidence.

[0042] Step S108 , performing weighted summation of the regional confidence of the target of interest, the closeness of the target of interest to the navigation path, and the probability of the target of interest invading the target area, to obtain the fusion confidence of the target of interest.

[0043] In the above weighted summation, the weights can be different weights preset based on different sub-areas in the navigation area. For example, the weight values ​​corresponding to the regional confidence, the degree of proximity, and the probability of invading the target area can be pre-set in each sub-area in the navigation area. Then, when specifically calculating the fusion confidence of the target of interest, first determine the sub-area to which the current position of the target of interest belongs, and then determine the weight values ​​corresponding to the regional confidence, the degree of proximity, and the probability of invading the target area in the sub-area, and then multiply and sum them with the regional confidence of the target vehicle, the degree of proximity between the target of interest and the navigation path, and the probability of the target of interest invading the target area, to obtain the final fusion confidence.

[0044] The method for determining the confidence of a target of interest of a vehicle provided in an embodiment of the present application can calculate the fusion confidence of the vehicle through multi-dimensional confidence evidence, thereby improving the accuracy of the confidence of the target of interest and obtaining key targets more accurately.

[0045] The embodiment of the present application also provides another method for determining the confidence of a vehicle target of interest. The method is implemented on the basis of the previous embodiment. This embodiment focuses on describing the calculation process of three types of confidence evidence and the division method of the navigation area.

[0046] The method for determining the confidence level of the above-mentioned preset pilot area (i.e., the method for dividing the pilot area into multiple confidence levels) is as follows:

[0047] (1) The pilot path is a curve with the current position of the ego vehicle as the starting point and a radius equal to the turning radius of the ego vehicle; it is assumed that the preset vehicle is traveling along the pilot path and the ego vehicle is traveling along a specified straight line; the specified straight line is the tangent of the pilot path at the starting point;

[0048] (2) Calculate the first intersection position and the second intersection position according to the turning radius of the vehicle, the width of the vehicle, and the width of the preset vehicle; wherein the first intersection position is: when the upper right corner of the vehicle collides with the rear center of the preset vehicle, the turning radius of the rear center of the preset vehicle and the intersection position of the specified straight line; the second intersection position is: when the upper right corner of the vehicle collides with the lower left corner of the preset vehicle, the turning radius of the rear center of the preset vehicle and the intersection position of the specified straight line;

[0049] The above intersection position is characterized by the distance from the intersection to the starting point; according to the turning radius of the vehicle, the width of the vehicle, and the preset vehicle width, the first intersection position and the second intersection position are calculated, including:

[0050] Calculate the distance from the first intersection point to the starting point according to the following first specified formula:

[0051] D1=sqrt(2*R2*W1+W1 2 );

[0052] Where D1 represents the distance from the first intersection point to the starting point; R2 represents the turning radius of the vehicle; W1 represents half of the width of the vehicle;

[0053] Calculate the distance from the second intersection point to the starting point using the following second specified formula:

[0054] D2=sqrt(2*R2*W3+W3 2 ); W3 = W1 + W2;

[0055] Wherein, D2 represents the distance from the second intersection point to the starting point; W2 represents half of the preset vehicle width.

[0056] (3) In the navigation area, the area determined when the width of the vehicle is extended along the navigation path to the first intersection from the starting point is taken as the first sub-area; the area determined when the width of the vehicle and the preset vehicle width are extended along the navigation path to the second intersection is taken as the sum of the width of the vehicle and the preset vehicle width, and the first sub-area is removed to obtain the second sub-area; and the remaining area is taken as the third sub-area;

[0057] (4) A corresponding preset confidence level is set for each sub-region; the preset confidence levels corresponding to the first sub-region, the second sub-region, and the third sub-region decrease in sequence.

[0058] like Figure 2 As shown in the figure, taking the preset vehicle B driving in the curve of the pilot path as an example, the pilot area is expanded from the pilot path through the lateral safety distance, and the specific expansion distance is determined according to the vehicle width, target width and reserved safety distance. The pilot area is in the shape of a semi-circle.

[0059] When dividing the confidence of the pilot area, it is assumed that the preset vehicle B (the target of interest) stops along the pilot path and can be anywhere on this path. The ego vehicle A starts from the starting point O1. It is assumed that the ego vehicle does not follow the pilot path but drives in a straight line. The distance between the starting point O1 and the first intersection O3 is D1. The solution process of D1 is as follows: the turning radius R2 of the ego vehicle and the half ego vehicle width W1 are known;

[0060] D1=sqrt(R1 2 -R2 2 );

[0061] R1=R2+W1;

[0062] The above two equations can be simplified to:

[0063] D1=sqrt(2*R2*W1+W1 2 );

[0064] Regardless of whether vehicle A is driving along the pilot path or along a straight line, it will collide with vehicle B. At this time, the collision area is the width of the vehicle. The confidence of the pilot area at this distance is 1.0. That is, the confidence of the target in the minimum dotted line area Area1 constructed by D1 and the width of the vehicle is 1.0;

[0065] like Figure 3 As shown, vehicle A is traveling in a straight line and passing Figure 2 After the first segment, assuming that the ego vehicle intersection point O3 still does not follow the pilot path and continues to drive in a straight line, the distance between the starting point O1 and the second intersection point O6 is D2. The solution process of D2 is as follows: Given the ego vehicle turning radius R2, half ego vehicle width W1, and half target width W2, then;

[0066] D2=sqrt(R3 2 -R4 2 );

[0067] W3=W1+W2;

[0068] R3=R2+W3;

[0069] The above three equations can be simplified to:

[0070] D2=sqrt(2*R2*W3+W3 2 );

[0071] The positions of the first intersection point and the second intersection point can be determined by the distances D1 and D2.

[0072] Whether vehicle A is driving along the pilot path or in a straight line, there is a risk of collision with vehicle B. The collision area at this time is the vehicle width plus the preset vehicle width. Area 2 is obtained by subtracting area 1 from the area constructed by D2 and the vehicle width + preset vehicle width. The built-in confidence in area 2 is 0.75, and the confidence of the remaining pilot area is 0.5.

[0073] The following are three evidence determination processes:

[0074] 1. The process of determining the regional confidence of the target of interest:

[0075] The above-mentioned regional confidence settings can be used to determine the regional confidence of the target of interest as one of the evidences for the final confidence solution. Specifically, first determine the target sub-region to which the current position of the target of interest belongs in the navigation region; then use the preset confidence corresponding to the target sub-region as the regional confidence of the target of interest. For example, the current position of the target of interest is in the first sub-region, and its regional confidence is 1.0.

[0076] 2. The process of determining the degree of closeness to the pilot path:

[0077] (1) Determine the vertical distance from the current position of the target of interest to the pilot path;

[0078] (2) Determine the proximity of the target of interest to the navigation path based on the lateral safety distance and the vertical distance from the current position of the target of interest to the navigation path.

[0079] The degree of closeness β between the target of interest and the navigation path is determined according to the following third specified formula:

[0080] β=max((W4-y) / W4,0);

[0081] Wherein, W4 represents the lateral safety distance reserved for the vehicle to travel along the pilot path; y represents the vertical distance from the current position of the target of interest to the pilot path.

[0082] like Figure 4As shown in the figure, the target of interest C(x2, y2) is transformed into the expression of C′(x2′, y2′). C′ is the intersection point of the perpendicular line from C to the pilot path and the pilot path. x2′ is the radial distance from the intersection point to the origin, and y2′ is the length of the perpendicular line. The target of interest B(x1, y1) is also transformed in the same way to B′(x1′, y1′). The magnitudes of y1′ and y2′ characterize the degree of proximity of the target of interest to the pilot path, and the smaller the value, the closer it is to the pilot path.

[0083] In the embodiment of the present application, B′(x1′, y1′) and C′(x2′, y2′) can be obtained by using the vertical approximation algorithm.

[0084] The vehicle traveling along the pilot path reserves a lateral safety distance W4:

[0085] W4 = W1 + 0.2;

[0086] Then the path proximity degree of the target of interest B:

[0087] β 1 = max((W4 - y1′) / W4, 0);

[0088] Similarly, the path proximity degree of the target of interest C:

[0089] β 2 = max((W4 - y2′) / W4, 0);

[0090] At the same time, from Figure 4 it can be seen that y1′ < y2′, then β 1 > β 2 , then B is closer to the pilot path compared to C. When β is 1, the target is traveling on the pilot path, and when β is 0, the target has deviated from the pilot path.

[0091] 3. The determination process of the probability that the target of interest invades the target area:

[0092] The target area is determined by the above preset distance, the vehicle width, the target of interest width, and the reserved safety distance; the determination method of the probability that the target of interest invades the target area is as follows:

[0093] (1) Determine the left and right boundaries of the target area, and the vertical distance from the current position of the target of interest to the pilot path;

[0094] (2) Calculate the differences between the left and right boundaries and the vertical distance respectively;

[0095] (3) Substitute the differences and the target of interest tracking standard deviation into the complementary error function to calculate the first probability that the target of interest is on the left of the right boundary and the second probability that the target of interest is on the left of the left boundary;

[0096] (4) Calculate the difference between the first probability and the second probability to obtain the probability that the target of interest invades the target area.

[0097] See also Figure 5 As shown, the target position (x3, y3) and (v x ,v y ), we can find the straight line y1 along the direction of the resultant velocity:

[0098] y1=tan(v x / v y )*x+(y3-tan(v x / v y )*x3)(x3′,y3′).

[0099] According to the curve equation, the representation of the vertical point in the vehicle coordinate system (x31, y31) can be obtained.

[0100] Then, according to the curve equation and (x31, y31), the tangent y2 at this point can be obtained. Then the intersection point (x4, y4) of y1 and y2 can be obtained. Given the three vertices of the triangle, any angle of the triangle can be obtained, and the angle θ between the vertical line and the hypotenuse can be obtained. Further, the velocity component in the direction of the vertical line can be expressed as:

[0101]

[0102] Given the right boundary W5, the target (x′, y′) sought by the approximation algorithm, the target tracking standard deviation SD, and the probability γ to the left of the right boundary:

[0103]

[0104] Where ttc is the calculated longitudinal collision time, and its value range is 0.0≤ttc≤4.6 seconds. Outside the range, ttc is 0 seconds; y′ lr The target is represented by y′, which is y′ on the left side of the trajectory and -y′ on the right side of the trajectory.

[0105] Then we know the left boundary W6, and the probability σ that the target is to the left of the left boundary:

[0106]

[0107] The probability of the target invading the left and right boundaries is:

[0108] δ=γ-σ.

[0109] Finally, the greater the probability that the target intrudes into the left and right boundaries, the higher the confidence. The left and right boundaries are defined according to the specific working conditions, target type, etc., and the intrusion degree of the target is further determined according to the target δ. The larger the δ, the higher the intrusion degree. At the same time, the corresponding interaction strategy can be formulated according to different δ.

[0110] Finally, the fusion confidence τ of the target of interest based on multi-dimensional evidence fusion is solved:

[0111] The confidence calculation evidence of the three targets of interest is obtained from steps 1, 2, and 3 above. After regional fusion, the weighted sum can be calculated based on the preset weights of each sub-region. The preset weights of each sub-region are shown in Table 1:

[0112] Table 1

[0113] area ε ratio β ratio δ ratio Area1 0.9 0.03 0.07 Area2 0.2 0.3 0.5 Area3 0.0 0.3 0.7

[0114] Among them, the sum of the three proportions of ε, β, and δ in each row is 1.0.

[0115] The above method also includes: determining whether the fusion confidence of the target of interest exceeds a preset threshold; if so, determining the target of interest as a key target; if there are multiple key targets determined, selecting the target of interest closest to the vehicle as the key target. For example, a target with fusion confidence τ>0.6 can be used as a candidate key target. If there is only one, then this target is the key target; if there are multiple candidate key targets, then the target (x ′ ,y ′ ) ′ Sort and select the key target that is closest to the vehicle; if there is no candidate key target, there is no key target.

[0116] like Figure 5 In the example shown, there is a target that pre-invades the left and right safety boundaries of the ego vehicle. During its intrusion process, the intrusion probability δ can be graded: Level 1 intrusion, δ<0.3, the ego vehicle does not take any action; Level 2 intrusion, 0.3≤δ<0.5, the ego vehicle slowly decelerates and sounds the horn to alert the target ahead; Level 3 intrusion, δ≥0.5, is judged based on the value of the sub-regional fusion confidence τ. If τ>0.6, the target can be followed or braked as a candidate key target; if τ≤0.6, the level 2 intrusion strategy continues to be executed.

[0117] Those skilled in the art should understand that the specific parameters involved in the methods provided in the embodiments of the present application are empirical values ​​of the inventors, and actual applications need to be adjusted according to specific engineering differences.

[0118] The method for determining the confidence level of a vehicle target of interest provided in the embodiment of the present application has the following advantages:

[0119] First, a method for multi-segment confidence division of the pilot area is provided to solve the confidence solution of the pilot area as one of the evidences for the target confidence solution. Providing multiple confidence areas solves the problem that a single confidence division of the pilot area cannot give confidence to the target of interest.

[0120] Second, a verification method for the target attribute and the ego vehicle pilot path curve is provided to solve the degree of closeness of the pilot path as one of the evidences for solving the target confidence. The closeness of the pilot path is provided to solve the problem that the target attributes of interest are not fully utilized.

[0121] Third, a target area confidence algorithm is provided to solve the probability of the target invading the left and right boundaries as one of the evidences for solving the confidence of the target of interest. A target area confidence algorithm is provided to solve the problem that the target noise and standard error are not considered when solving the probability of the target invading the left and right boundaries.

[0122] Fourth, a multi-dimensional confidence fusion solution is provided to solve the confidence of the target of interest of the intelligent driving vehicle under multi-dimensional evidence. It provides a platform for the implementation and presentation of the entire solution. It provides a reliable confidence fusion solution, solves the problem of low accuracy caused by single-dimensional confidence solution, and can obtain key targets more accurately.

[0123] The method provided in the embodiment of the present application comprehensively considers the impact of each dimension of evidence on the confidence solution of the target of interest, provides three evidence calculation methods, and adopts a regional fusion method to solve the confidence of the target of interest. It makes full use of information such as the navigation area, target position, target standard deviation, and provides corresponding solutions, thereby providing multi-dimensional evidence. In the process of multi-dimensional evidence fusion, a multi-dimensional confidence fusion solution is provided according to the characteristics of each piece of evidence, which improves the accuracy of the confidence of the target of interest, and also provides a numerical basis for the screening of key targets.

[0124] Based on the above method embodiment, the present application embodiment also provides a vehicle interest target confidence determination device, see Figure 6As shown, the device includes: a position acquisition module 62, which is used to obtain the current position of the target of interest in the navigation area of ​​the vehicle; the navigation area is obtained by expanding the navigation path through a lateral safety distance; a regional confidence determination module 64, which is used to determine the regional confidence of the target of interest according to the preset navigation area confidence and the current position; the preset navigation area confidence includes: preset confidences corresponding to multiple sub-areas determined by dividing the navigation area; a proximity and probability determination module 66, which is used to determine the proximity of the target of interest to the navigation path and the probability of the target of interest invading the target area according to the current position of the target of interest and the navigation path; the target area is obtained by expanding the navigation path through a preset distance; a fusion confidence determination module 68, which is used to perform weighted summation of the regional confidence of the target of interest, the proximity of the target of interest to the navigation path, and the probability of the target of interest invading the target area to obtain the fusion confidence of the target of interest.

[0125] Furthermore, the above-mentioned regional confidence determination module 64 is used to: determine the target sub-region to which the current position belongs in the navigation region; and use the preset confidence corresponding to the target sub-region as the regional confidence of the target of interest.

[0126] Furthermore, the above-mentioned device also includes: an area division setting module, which is used to execute the process of determining the confidence of the preset navigation area: the navigation path is a curve with the current position of the self-vehicle as the starting point and the radius as the turning radius of the self-vehicle; it is assumed that the preset vehicle travels along the navigation path and the self-vehicle travels along the specified straight line; the specified straight line is the tangent of the navigation path at the starting point; according to the turning radius of the self-vehicle, the width of the self-vehicle, and the width of the preset vehicle, the first intersection position and the second intersection position are calculated; wherein; the first intersection position is: when the upper right corner of the self-vehicle collides with the rear center of the preset vehicle, the intersection position of the turning radius of the rear center position of the preset vehicle and the specified straight line; the second intersection position The intersection point is: when the upper right corner of the vehicle collides with the lower left corner of the preset vehicle, the turning radius of the rear center position of the preset vehicle and the specified straight line; starting from the starting point, the area determined when the width of the vehicle is extended along the navigation path to the first intersection point is taken as the first sub-area; the area determined when the sum of the width of the vehicle and the width of the preset vehicle is extended along the navigation path to the second intersection point is obtained by removing the first sub-area; the other remaining areas are taken as the third sub-area; a corresponding preset confidence is set for each sub-area; the preset confidences corresponding to the first sub-area, the second sub-area, and the third sub-area are reduced in sequence.

[0127] Furthermore, the position of the intersection is characterized by the distance from the intersection to the starting point; the area division setting module is used to: calculate the distance from the first intersection to the starting point according to the following first specified formula:

[0128] D1=sqrt(2*R2*W1+W12 );

[0129] Where D1 represents the distance from the first intersection point to the starting point; R2 represents the turning radius of the vehicle; W1 represents half of the width of the vehicle;

[0130] Calculate the distance from the second intersection point to the starting point using the following second specified formula:

[0131] D2=sqrt(2*R2*W3+W3 2 ); W3 = W1 + W2;

[0132] Wherein, D2 represents the distance from the second intersection point to the starting point; W2 represents half of the preset vehicle width.

[0133] Furthermore, the proximity and probability determination module 66 is used to: determine the vertical distance from the current position of the target of interest to the navigation path; and determine the proximity of the target of interest to the navigation path based on the lateral safety distance and the vertical distance from the current position of the target of interest to the navigation path.

[0134] Furthermore, the closeness and probability determination module 66 is used to determine the closeness β between the target of interest and the navigation path according to the following third specified formula:

[0135] β=max((W4-y) / W4,0);

[0136] Wherein, W4 represents the lateral safety distance reserved for the vehicle to travel along the pilot path; y represents the vertical distance from the current position of the target of interest to the pilot path.

[0137] Furthermore, the preset distance is determined by the width of the vehicle, the width of the target of interest and the reserved safety distance; the proximity and probability determination module 66 is used to: determine the left and right boundaries of the target area, and the vertical distance from the current position of the target of interest to the navigation path; respectively calculate the difference between the left and right boundaries and the vertical distance; substitute the difference and the tracking standard deviation of the target of interest into the complementary error function, and calculate the first probability that the target of interest is on the left side of the right boundary and the second probability that it is on the left side of the left boundary; calculate the difference between the first probability and the second probability to obtain the probability that the target of interest invades the target area.

[0138] The above-mentioned device also includes: a key target screening module, which is used to: determine whether the fusion confidence of the target of interest exceeds a preset threshold; if so, determine the target of interest as a key target; if there are multiple key targets determined, select the target of interest closest to the vehicle as the key target.

[0139] The device provided in the embodiment of the present application has the same implementation principle and technical effects as those of the aforementioned method embodiment. For the sake of brief description, for matters not mentioned in the embodiment of the device, reference may be made to the corresponding contents in the aforementioned method embodiment.

[0140] Based on the above method embodiment, the embodiment of the present application also provides an intelligent driving vehicle, including a processor and a memory, the memory stores computer executable instructions that can be executed by the processor, and the processor executes the computer executable instructions to implement the above vehicle object of interest confidence determination method.

[0141] The intelligent driving vehicle provided in the embodiment of the present application has the same implementation principle and technical effects as those of the aforementioned method embodiment. For the sake of brief description, for parts not mentioned in the embodiment of the intelligent driving vehicle, reference may be made to the corresponding contents in the aforementioned method embodiment.

[0142] The present application also provides an electronic device, such as Figure 7 As shown, it is a schematic diagram of the structure of the electronic device, wherein the electronic device includes a processor 71 and a memory 70, the memory 70 stores computer executable instructions that can be executed by the processor 71, and the processor 71 executes the computer executable instructions to implement the above method.

[0143] exist Figure 7 In the illustrated embodiment, the electronic device further includes a bus 72 and a communication interface 73 , wherein the processor 71 , the communication interface 73 and the memory 70 are connected via the bus 72 .

[0144] Among them, the memory 70 may include a high-speed random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage. The communication connection between the system network element and at least one other network element is realized through at least one communication interface 73 (which can be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. can be used. The bus 72 can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The bus 72 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 7 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0145] The processor 71 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the processor 71. The above processor 71 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in the embodiment of the present application can be directly embodied as a hardware decoding processor to execute, or the hardware and software modules in the decoding processor can be executed. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor 71 reads the information in the memory and completes the steps of the method of the above embodiment in combination with its hardware.

[0146] An embodiment of the present application also provides a computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are called and executed by a processor, the computer-executable instructions prompt the processor to implement the above method. The specific implementation can be found in the aforementioned method embodiment, which will not be repeated here.

[0147] The computer program products of the methods, devices, and electronic devices provided in the embodiments of the present application include a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the previous method embodiments. The specific implementation can be found in the method embodiments, which will not be repeated here.

[0148] Unless otherwise specifically stated, the relative steps, numerical expressions and values ​​of the components and steps set forth in these embodiments do not limit the scope of the present application.

[0149] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0150] In the description of the present application, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present application. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0151] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The protection scope of the present application is not limited thereto. Although the present application is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed in the present application, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A method for determining the confidence of a vehicle target of interest, characterized in that: The method comprises: Obtaining the current position of the target of interest within the navigation area of ​​the vehicle; the navigation area is obtained by expanding the navigation path through a lateral safety distance; Determine the regional confidence of the target of interest according to the preset navigation area confidence and the current position; the preset navigation area confidence includes: preset confidences corresponding to a plurality of sub-areas determined by dividing the navigation area; According to the current position of the target of interest and the navigation path, respectively determining the proximity of the target of interest to the navigation path and the probability of the target of interest invading a target area; the target area is obtained by expanding the navigation path by a preset distance; The regional confidence of the target of interest, the closeness of the target of interest to the navigation path, and the probability of the target of interest invading the target area are weightedly summed to obtain the fusion confidence of the target of interest.

2. The method according to claim 1, characterized in that The step of determining the area confidence of the target of interest according to the preset navigation area confidence and the current position includes: Determine the target sub-area to which the current position belongs in the navigation area; The preset confidence corresponding to the target sub-region is used as the regional confidence of the target of interest.

3. The method according to claim 1, characterized in that The confidence level of the preset navigation area is determined as follows: The pilot path is a curve with the current position of the vehicle as the starting point and a radius equal to the turning radius of the vehicle; assuming that a preset vehicle travels along the pilot path, the vehicle travels along a designated straight line; the designated straight line is a tangent line of the pilot path at the starting point; Calculating the first intersection position and the second intersection position according to the turning radius of the vehicle, the width of the vehicle, and the preset vehicle width; in; The first intersection position is: when the upper right corner of the vehicle collides with the rear center of the preset vehicle, the intersection position of the turning radius at the rear center of the preset vehicle and the designated straight line; the second intersection position is: when the upper right corner of the vehicle collides with the lower left corner of the preset vehicle, the intersection position of the turning radius at the rear center of the preset vehicle and the designated straight line; In the navigation area, the area determined when the width of the vehicle is extended along the navigation path to the first intersection from the starting point is used as the first sub-area; the area determined when the width of the vehicle and the preset vehicle width are extended along the navigation path to the second intersection is obtained by removing the first sub-area from the area determined when the width of the vehicle and the preset vehicle width are extended along the navigation path to the second intersection; and the remaining area is used as the third sub-area; A corresponding preset confidence level is set for each sub-region; the preset confidence levels corresponding to the first sub-region, the second sub-region, and the third sub-region are decreased in sequence.

4. The method according to claim 3, characterized in that The intersection position is characterized by the distance from the intersection to the starting point; the first intersection position and the second intersection position are calculated according to the turning radius of the vehicle, the width of the vehicle, and the preset vehicle width, including: The distance from the first intersection point to the starting point is calculated according to the following first specified formula: D1=sqrt(2*R2*W1+W1 2 ); Where D1 represents the distance from the first intersection point to the starting point; R2 represents the turning radius of the vehicle; W1 represents half of the width of the vehicle; The distance from the second intersection point to the starting point is calculated according to the following second specified formula: <h2 style=";text-align:left;direction:ltr">D2 = sqrt(2*R2*W3+W3<h2 style=";text-align:left;direction:ltr"> 2 <h2 style=";text-align:left;direction:ltr"> );W3=W1+W2; Wherein, D2 represents the distance from the second intersection point to the starting point; W2 represents half of the preset vehicle width.

5. The method according to claim 1, characterized in that Determining the proximity of the object of interest to the navigation path according to the current position of the object of interest and the navigation path includes: Determining a vertical distance from a current position of the object of interest to the navigation path; The degree of proximity between the object of interest and the navigation path is determined according to the lateral safety distance and the vertical distance from the current position of the object of interest to the navigation path.

6. The method according to claim 5, characterized in that Determining the proximity of the target of interest to the navigation path according to the lateral safety distance and the vertical distance from the current position of the target of interest to the navigation path includes: The degree of closeness β between the target of interest and the navigation path is determined according to the following third specified formula: β=max((W4-y) / W4,0); Wherein, W4 represents the lateral safety distance reserved for the vehicle to travel along the navigation path; y represents the vertical distance from the current position of the target of interest to the navigation path.

7. The method according to claim 1, characterized in that The preset distance is determined by the vehicle width, the target width and the reserved safety distance; and the probability of the target intruding into the target area is determined according to the current position and the navigation path, including: Determine the left and right boundaries of the target area, and the vertical distance from the current position of the target of interest to the navigation path; respectively calculating the difference between the left and right boundaries and the vertical distance; Substituting the difference and the tracking standard deviation of the target of interest into a complementary error function, and calculating a first probability that the target of interest is to the left of the right boundary and a second probability that the target of interest is to the left of the left boundary respectively; The difference between the first probability and the second probability is calculated to obtain the probability that the object of interest invades the target area.

8. The method according to claim 1, characterized in that The method further comprises: Determining whether the fusion confidence of the target of interest exceeds a preset threshold; If yes, identify said target of interest as a key target; If there are multiple key targets identified, the target of interest closest to the vehicle is selected as the key target.

9. A device for determining the confidence level of a vehicle target of interest, characterized in that: The device comprises: A position acquisition module is used to acquire the current position of the target of interest within the navigation area of ​​the vehicle; the navigation area is obtained by expanding the navigation path through the lateral safety distance; A regional confidence determination module, used to determine the regional confidence of the target of interest according to a preset navigation area confidence and the current position; the preset navigation area confidence includes: preset confidences corresponding to a plurality of sub-areas determined by dividing the navigation area; A proximity degree and probability determination module, used to determine the proximity degree of the target of interest to the pilot path and the probability of the target of interest invading a target area according to the current position of the target of interest and the pilot path; the target area is obtained by expanding the pilot path by a preset distance; The fusion confidence determination module is used to perform weighted summation of the regional confidence of the target of interest, the degree of proximity between the target of interest and the navigation path, and the probability of the target of interest invading the target area to obtain the fusion confidence of the target of interest.

10. An intelligent driving vehicle, characterized in that: The method comprises a processor and a memory, wherein the memory stores computer executable instructions that can be executed by the processor, and the processor executes the computer executable instructions to implement the vehicle interest target confidence determination method according to any one of claims 1 to 8.

Images