Methods for avoiding obstacles

Abstract

The invention relates to a method for avoiding objects (C1, C2, C3, C4) by a motor vehicle (10), comprising the steps of detecting the objects located in the vicinity of the motor vehicle, obtaining data characterizing the position and / or movement of each object, then, if several objects are detected, verifying whether at least one criterion of proximity between at least two of the detected objects is fulfilled, and, if so, combining the two objects into a group (G2), calculating data characterizing the position and / or movement of the group, and activating a system for obstacle avoidance and / or determining an avoidance trajectory according to the data characterizing the position and / or movement of the group.

Description

[Technical Field]

[0001] This invention generally relates to means of assisting automobile driving.

[0002] The present invention relates, more particularly, to a method for avoiding obstacles.

[0003] The present invention also relates to an automobile equipped with a computer designed to carry out this method. [Background technology]

[0004] With the aim of increasing vehicle safety, automobiles today are equipped with driver assistance systems or autonomous driving systems.

[0005] These systems are known to include, in particular, automatic emergency braking (AEB) systems designed to avoid any collision with an obstacle located within the lane in which the vehicle is traveling, simply by acting on the vehicle's conventional braking system.

[0006] However, there are some situations in which these emergency braking systems do not enable the avoidance of a collision, or are not usable (for example, if another vehicle is following directly behind the vehicle).

[0007] In response to these situations, advanced steering avoidance or automatic steering avoidance (AES) systems have been developed, which enable the vehicle to avoid obstacles by acting on the vehicle's steering to divert it from its trajectory.

[0008] To enable this AES function, it is necessary to reliably detect the parts of the vehicle's surroundings that are relevant to the calculation of the optimal avoidance maneuver.

[0009] One parameter commonly used for this purpose is called the time to collision (TTC). A second parameter is formed by the deviation taken to pass a detected obstacle without colliding with it. Thus, each potential obstacle is considered separately in order to determine the most dangerous of the obstacles and to estimate the optimal avoidance trajectory from that obstacle.

[0010] Therefore, relying on these parameters, it can be seen that in certain dangerous situations where it is preferable to leave the avoidance to the driver, the AES collision avoidance function will be activated in autonomous mode (without driver intervention). However, there may also be cases where the collision avoidance function is not activated due to an excessive number of obstacles to be avoided within the same area, thereby making the function relatively unavailable. [Overview of the project]

[0011] To correct the aforementioned shortcomings of the prior art, the present invention proposes processing detected objects not individually, but by grouping them together whenever possible.

[0012] More specifically, the present invention provides a method for avoiding an object for an automobile, wherein the object is initially considered to be a potential obstacle. This method is - A step of detecting an object located around the vehicle, - A step of obtaining data characterizing the position and / or movement of each detected object, and then, if multiple objects are detected, - A step of checking whether at least one criterion for proximity between at least two of the detected objects is met, and if it is met, - The step of combining two objects into one group, - A step of obtaining data that characterizes the position and / or movement of the above group, - The steps include activating an obstacle avoidance system and / or determining an avoidance maneuver according to data characterizing the position and / or movement of the group described above.

[0013] There are numerous advantages to the methods used to group targets.

[0014] The main advantage of this method is that it means the AES function can be activated in a greater number of situations than it would be under general rules.

[0015] In fact, when there are an excessive number of obstacles to process, it is generally stipulated that the AES function be left inactive, while in this case, it is possible to process a large number of separate objects by grouping them together.

[0016] Similarly, when a vehicle is traveling on a two-lane road and one or more objects are located within each lane, it is generally stipulated that the AES function should not be activated. In contrast, in this invention, only deviations between various objects or groups of objects are considered to check whether it is possible to activate the AES function.

[0017] Furthermore, it is generally required to detect the traffic lane in which each object is located before activating the AES function. When an object is located across two traffic lanes, the gap for passing over this object is more limited than when the object is located in the center of the lane. Therefore, a larger safety margin is generally considered before activating the AES function. In contrast, in this invention, each target group defines an area to be avoided, and its location does not depend on the location of the traffic lane. Therefore, it is possible to limit the safety margin to be considered, thereby enabling the activation of the AES function in a wider range of situations.

[0018] More generally, this method allows us to ignore the concept of traffic lanes when calculating avoidance trajectories and prefers to treat the surroundings as a single space outside the object.

[0019] It should also be noted that this invention makes it possible to simplify calculations.

[0020] Therefore, the present invention enables the simultaneous management of a larger number of targets, which proves particularly useful when detecting groups of cyclists.

[0021] It should also be noted that the present invention prefers to ignore the class of the detected object (e.g., cyclist, car) and to group objects of any class according to only one or more proximity criteria.

[0022] Other advantageous and non-limiting features of the methods according to the present invention, which may be employed individually or in any technically achievable combination, are as follows: - Data characterizing the position and / or movement of the above group is calculated based on data characterizing each individual object in the group. Next, these data characterizing the individual objects in the group are no longer considered, the obstacle avoidance system is activated, and / or, in particular, the avoidance trajectory is determined regardless of the data characterizing the individual objects in the group. - The proximity criterion is related to the lateral distance between the two objects (along the axis perpendicular to the tangent to the road at the height of the objects). - In the acquisition step, one of the data points is the lateral trajectory deviation that the car must take to avoid each object. - The proximity criterion involves checking whether the difference between the lateral deviation that one object must take to avoid the first object on the side facing the second object and the lateral deviation that one object must take to avoid the second object on the side facing the first object is greater than or equal to a predetermined threshold. - This threshold is greater than or equal to 0. - If at least three objects are detected, it is stipulated that the objects be ranked in a continuous order from one edge of the road to the other, and then it is checked whether the proximity criterion is met between each pair of objects that are continuous in the above continuous order. - It is stipulated that the relative lateral velocity is calculated based on the deviation between the relative lateral velocity on the vehicle with respect to the road in a first reference coordinate system oriented along the tangent to the road at the vehicle's height, and the lateral velocity of the object with respect to the road in a second reference coordinate system oriented along the tangent to the road at the object's height, and then each lateral trajectory deviation is determined based on the relative lateral velocity. - The above proximity criterion relates to the longitudinal distance between two objects. - In the acquisition step, one of the data points relates to the remaining time before the car collides with each object. - To check whether the above proximity criteria are met, it is checked whether the deviation in the time between the two objects until collision is below a threshold. - This threshold is greater than or equal to 0. - In the acquisition step, one of the data is the lateral trajectory deviation that each object should take to avoid it on the exact same left or right side, and in the calculation step, one of the data that characterizes the group is selected such that it is equal to the largest lateral trajectory deviation that the group's objects should take to avoid it on the exact same side. - In the acquisition step, one of the data points characterizing each object is the remaining time before the car collides with each object, and in the calculation step, one of the data points characterizing the group is selected such that it is equal to the minimum remaining time before the car collides with one of the objects in the group.

[0023] The present invention also relates to an automobile comprising at least one steering wheel and a steering system for each steering wheel, which is designed to be operated by a computer-controlled actuator, the computer being designed to implement the triggering method described above.

[0024] Of course, the various features, variations, and embodiments of the present invention may be combined with each other in various combinations, provided that they are not incompatible or mutually exclusive.

[0025] The following description, provided with reference to the accompanying drawings as non-limiting examples, will give a good understanding of the concept of the present invention and the ways in which it may be carried out. [Brief explanation of the drawing]

[0026] [Figure 1] This is a schematic diagram of an automobile according to the present invention and two target automobiles traveling in two separate traffic lanes. [Figure 2] Figure 1 shows a schematic diagram of an automobile and one of two target automobiles. [Figure 3] This diagram, corresponding to the one shown in Figure 2, illustrates the second step in the process of determining the position of one of the target vehicles. [Figure 4] This figure, corresponding to Figure 2, shows four reference coordinate systems used as part of the present invention. [Figure 5] This is a representation of the four reference coordinate systems from Figure 4. [Figure 6] This is a schematic diagram of one of the target vehicles from Figure 1 and the traffic lane of that vehicle. [Figure 7] This diagram shows one exemplary configuration in which an automobile according to the present invention is positioned, and a schematic diagram of four automobiles traveling in close proximity to the automobile. [Figure 8] This diagram corresponds to Figure 7, showing the reordering of automobile standards. [Figure 9]This figure corresponds to Figure 8, where automobiles are grouped in the first step. [Figure 10] This figure corresponds to Figure 8, where automobiles are grouped for the second time. [Figure 11] This diagram shows another exemplary configuration in which an automobile according to the present invention is positioned, and a schematic diagram of a group consisting of three automobiles traveling in close proximity to the automobile. [Figure 12] This diagram shows another exemplary configuration in which an automobile according to the present invention is positioned, and a schematic diagram of three automobiles traveling in close proximity to the automobile. [Modes for carrying out the invention]

[0027] Figure 1 shows a car 10 traveling on a road where two “objects” that form potential obstacles to the car 10 are present. In this case, these two objects are formed by cars C1 and C2. In one variation, these two objects may be other types of objects (pedestrians, cyclists, etc.). The objects considered are preferably in motion.

[0028] For the remainder of this specification, the automobile 10 is an embodiment of the present invention and is referred to as "the vehicle 10".

[0029] This vehicle 10, as in the conventional model, comprises a chassis that divides the passenger compartment, wheels, two of which are steering wheels, a drivetrain, a braking system, and a conventional steering system that acts in accordance with the direction of the steering wheels.

[0030] In the example under consideration, the steering system is controlled by an assisting steering actuator that allows it to act on the orientation of the steering wheel based on the orientation of the steering wheel and / or, in some cases, on commands issued by computer C10.

[0031] Computer C10 comprises at least one processor, at least one memory, and various input and output interfaces.

[0032] Through the computer's input interface, the computer C10 can receive input signals from various sensors.

[0033] The following, for example, are defined between the sensors: - A device such as a front camera for identifying the vehicle's position relative to its lane. - A device such as a radar or lidar remote detector for detecting obstacles located on the track of the vehicle 10. - At least one lateral device, such as a radar or lidar remote detector, for observing the surrounding area to the side of the vehicle.

[0034] Therefore, the computer C10 receives data from multiple sensors relating to objects present around the vehicle 10. As before, this data is combined with each other to provide reliable fused data about each object.

[0035] The computer output interface allows the computer C10 to send commands to the assist steering actuator.

[0036] Therefore, it becomes possible to ensure that the vehicle follows an obstacle avoidance trajectory to the best of its ability, and where conditions permit.

[0037] Computer C10 stores data in its memory that will be used as part of the method described later.

[0038] Memory stores computer applications, in particular, computer programs that, through execution by the processor, enable the computer to perform the methods described below.

[0039] These programs include, in particular, an "AES system" designed to calculate an obstacle avoidance trajectory and control the vehicle 10 to follow that trajectory, or to assist the driver by controlling the vehicle 10 to follow that trajectory. The AES system has an autonomous mode in which the vehicle follows the trajectory without driver assistance, and a manual mode in which the AES system assists the driver by avoiding obstacles, with the driver remaining in control of the operating means. This AES system is well known to those skilled in the art and will not be described in detail here.

[0040] The computer program also includes startup software for activating the AES system, which determines whether the AES system should be activated (taking into account the vehicle's trajectory and the trajectories of objects surrounding the vehicle) and waits until the best moment to activate the AES system. More specifically, this startup software is particularly influenced by the present invention as described herein.

[0041] This software becomes active immediately when vehicle 10 moves.

[0042] This software runs in a loop with periodic time increments.

[0043] This software includes a preliminary step of acquiring data relating to the vehicle 10 and its surroundings, followed by nine main steps. These successive steps can then be described one by one.

[0044] In a preliminary step, computer C10 receives at least one image acquired by the vehicle's front camera. The computer also receives data from a remote detector. These images and data are then fused together.

[0045] Therefore, at this stage, the computer C10 holds an image of the road located in front of the host vehicle 10 and, in particular, fusion data detected around the host vehicle 10 and characterizing each surrounding object. These surroundings are here considered as the area located around the host vehicle, where the sensors of the vehicle are designed to acquire data.

[0046] In the example of FIG. 1, the host vehicle is traveling within the central traffic lane V C and on both sides of V C there are two other traffic lanes V R 、V L existing.

[0047] Next, the computer C10 seeks to determine the position and shape of the boundary lines NL, L, R, NL of these traffic lanes V C 、V R 、V L .

[0048] For this purpose, in this case, each of these lines is modeled by a polynomial. Here, the selected polynomial is cubic and can thus be written as follows. [Equation 1] yLine = d.x 3 + c.x 2 + b.x + a

[0049] In this equation, - the term yLine represents the abscissa of the lane boundary line being considered. - the term x represents the longitudinal coordinate of this line. - the terms a, b, c and d are the coefficients of the polynomial, determined based on the shape of the lane that the front camera of the host vehicle is looking at (or acquired by the computer C10 from a navigation system including a detailed map of the location where the host vehicle is moving).

[0050] In practice, these terms are provided by fusing data. These terms allow for the modeling of lane boundary lines up to a distance of approximately 100 meters, provided visibility conditions are good.

[0051] At this stage, it should be noted that in the remainder of this disclosure, the term “longitudinal” corresponds to the component of a vector along the transverse coordinate of the reference coordinate system under consideration, and the term “transverse” corresponds to the component of a vector along the y-coordinate of the reference coordinate system under consideration (the reference coordinate systems under consideration are always orthogonal).

[0052] Equation [Equation 1] is used herein to refer to the vehicle 10 and the reference coordinate system (X) shown in Figure 1. EGO ,Y EGO This is represented in ). This reference coordinate system is oriented such that the horizontal axis of the coordinate system extends along the longitudinal axis of the vehicle 10. The reference coordinate system is centered on the front radar of the vehicle 10.

[0053] As a variation, other simpler or more complex models of the lane boundary geometry may be used.

[0054] Once coefficients a, b, c, and d are determined for each lane boundary, computer C10 can perform nine steps in a manner that allows it to perceive the degree to which a detected object is hazardous to its own vehicle in order to trigger the AES obstacle avoidance system when necessary.

[0055] The first step is to determine the distance between the vehicle and the object under consideration (one of the vehicles C1 or C2).

[0056] The distance calculated here is not a Euclidean distance. In fact, it is desirable to take the shape of the road into account in order to determine the distance that the vehicle 10 and the object need to cover before colliding with each other.

[0057] Therefore, computer C10, where the arc distance L AB Calculate.

[0058] For this purpose, the computer uses the following formula, as detailed in, for example, reference FR3077547. [Formula 2] TIFF0007863569000001.tif11170

[0059] During the ceremony, - L AB is the arc distance between two points A and B (corresponding to the positions of the vehicle and the object being considered). - x A This represents the longitudinal position of the vehicle (at radar height). - x B The reference coordinate system (X EGO ,Y EGO This is the longitudinal position of the object being considered, as indicated in parentheses.

[0060] The second step involves determining the position of each detected object relative to the road's traffic lanes, taking into account the equations and fused data of each lane boundary.

[0061] Computer C10 calculates the reference coordinate system (X) associated with the vehicle for each detected object's feature point (hereinafter referred to as "anchor point"). EGO ,Y EGO The coordinates within the parentheses are known. This feature point is typically the center of the object as seen by the front camera or radar remote detector. Here, the center is considered to be the center of the radiator grille of vehicles C1 and C2.

[0062] In the example in Figure 1, where two objects (two cars) are detected, the coordinates of the anchor points are referred to as (X_rel1, Y_rel1) and (X_rel2, Y_rel2), respectively.

[0063] Figure 1 also shows the reference coordinate system (X EGO ,YEGO The following values ​​on the vertical axis of ) are also shown. - Y_road_NL_1 is the value of the term yLine from the equation [Equation 1] for the lane boundary line NL at the horizontal coordinate point X_rel1. - Y_road_NL_2 is the value of the term yLine from the equation [Equation 1] for the lane boundary line NL at the horizontal coordinate point X_rel2. - Y_road_L_1 is the value of the term yLine from the equation [Equation 1] for the lane boundary line L at the horizontal coordinate point X_rel1. - Y_road_L_2 is the value of the term yLine from the equation [Equation 1] for the lane boundary line L at the horizontal coordinate point X_rel2. - Y_road_R_1 is the value of the term yLine from the equation [Equation 1] for the lane boundary line R at the horizontal coordinate point X_rel1. - Y_road_R_2 is the value of the term yLine from the equation [Equation 1] for the lane boundary line R at the horizontal coordinate point X_rel2. - Y_road_NR_1 is the value of the term yLine from the equation [Equation 1] for the lane boundary line NR at the horizontal coordinate point X_rel1. - Y_road_NR_2 is the value of the term yLine from the equation [Equation 1] for the lane boundary line NR at the horizontal coordinate point X_rel2.

[0064] Next, by comparing these values ​​with the horizontal coordinates Y_rel1 and Y_rel2 of cars C1 and C2, it is possible to determine the traffic lane in which each of the two cars is located.

[0065] For example, the horizontal coordinate Y_rel1 of car C1 here lies between the values ​​Y_road_R_1 and Y_road_L_1, which means that this car is located between the lane boundary lines L and R.

[0066] Therefore, at this stage, computer C1 determines the traffic lane V in which each detected object is located. L , V C , V R This can be confirmed.

[0067] The third step aims to determine the parameters that characterize the kinematic behavior of each object relative to the lane boundary line.

[0068] To clarify this disclosure, the remainder of this step's description will refer to only one of these objects (automobile C1).

[0069] The third step includes a first substep in which the computer C10 determines the position of the object relative to one of the lane boundary lines. The lane boundary line under consideration is preferably one that separates the central traffic lane from the traffic lane on which the object under consideration is located.

[0070] As one variant, the lane boundary line being considered may be another line, such as a lane edge line (see Figures 2 and 3) when no lane is detected between the traffic lane of an object and the traffic lane of the vehicle 10.

[0071] The idea is to discretize the intervals of these lane boundaries into a finite number of N points, and then select the ones closest to the object under consideration. This process is performed multiple times, by discretizing the lane boundaries located on either side of the selected points over intervals that are reduced each time, in order to ultimately obtain a good estimate of the traffic lane point closest to the object under consideration.

[0072] In practice, as shown in Figure 2, the computer uses N coordinate points (X) within the reference coordinate system of the vehicle 10 to determine the lane boundary line R. i ,Y i It begins by discretizing the coordinates (in fact, along the X axis). These coordinate points are regularly distributed along this line (actually, along the X axis). EGO The distance between two consecutive points along the same axis is always the same. The first point is either at the same height as the vehicle (having a horizontal coordinate of 0) or at a first predetermined distance from the vehicle, and the last point is at a second predetermined distance from the vehicle.

[0073] Next, here (Xrel ;Y rel Given the coordinates of the anchor point of car C1, as shown below, the computer uses the following formula to calculate the Euclidean distance Bird between each discretized point of the lane boundary line R and the anchor point of car C1. Distance It is possible to infer this. [Formula 3] TIFF0007863569000002.tif8170

[0074] Euclidean distance (Bird) Distance The discretized point where is minimized is the point closest to car C1. Therefore, this coordinate point (X s ,Y s ) is selected.

[0075] Next, as shown in Figure 3, this discretization operation is repeated over smaller intervals with finer discretization. The interval boundaries are preferably coordinate points (X s-1 ,Y s-1 ) and (X s+1 ,Y s+1 ) is formed by the new coordinate point (X s ,Y s This allows you to select ).

[0076] After a certain number of loops (e.g., 10), or when the distance between two discretization points becomes sufficiently small (e.g., less than 10 cm), the computer stops repeating these looped operations.

[0077] The last selected point is called "projection point F". Projection point F will be considered a good approximation of the boundary point closest to car C1.

[0078] The horizontal coordinate X of projection point F s The value is Distance Xproj It is called [name].

[0079] Euclidean distance between projection point F and car C1: Bird Distance The value is Dist Target2Lane It is called [name].

[0080] The second substep involves determining, for computer C10, the speed of the vehicle itself in a reference coordinate system that is connected to the road and located at the same height as the vehicle itself, and the speed of vehicle C1 in a reference coordinate system that is connected to the road and located at the same height as the vehicle itself.

[0081] In this substep, it is assumed that the road follows the tangent line at projection point F. Therefore, it can be considered a straight line starting from car C1.

[0082] To provide a good understanding of the calculations, Figure 4 shows the four reference coordinate systems used in the remainder of this disclosure.

[0083] The first reference coordinate system is the previously presented reference coordinate system (X) which is linked to the vehicle itself. EGO ,Y EGO )

[0084] It should be noted that this reference coordinate system moves simultaneously with the vehicle 10. Therefore, although it coincides with the first reference coordinate system at the time of measurement, the absolute reference coordinate system (X) is considered to be fixed. abs ,Y abs ) is also shown.

[0085] Another reference coordinate system is (X lineEGO ,Y lineEGO This is shown as ), and this reference coordinate is linked to the lane boundary line R, and the lateral coordinate of this reference coordinate system is oriented so that it is tangent to this lane boundary line R, and is centered on the vehicle's radar (the lateral coordinate of this radar is 0 in the second reference coordinate system).

[0086] Furthermore, another reference coordinate system is (X obj ,Y objThis is shown as ), and this reference coordinate system is linked to vehicle C1, and the horizontal coordinate of this reference coordinate system is oriented so as to be aligned with the direction of movement of vehicle C1, and is centered on the anchor point of vehicle C1.

[0087] The last reference coordinate system is (X lineObj ,Y lineObj This is shown as ), and this reference coordinate system is linked to the lane boundary line R, and the lateral coordinate of this reference coordinate system is oriented so that it is tangent to this lane boundary line R, and is centered on the anchor point of vehicle C1.

[0088] Figure 5 shows the angles that separate these reference coordinate systems. - Angle lineEGO / EGO The reference coordinate system (X EGO ,Y EGO ) from the reference coordinate system (X lineEGO ,Y lineEGO This makes it possible to change to ). - Angle lineObj / EGO The reference coordinate system (X EGO ,Y EGO ) from the reference coordinate system (X lineObj ,Y lineObj This makes it possible to change to ). - Angle Obj / EGO The reference coordinate system (X EGO ,Y EGO ) from the reference coordinate system (X obj ,Y obj This makes it possible to change to ). - Angle Obj / LineObj The reference coordinate system (X lineObj ,Y lineObj ) from the reference coordinate system (X obj ,Y obj This makes it possible to change to ).

[0089] Angle LineX / EGO More generally, the reference coordinate system (X EGO ,Y EGO The horizontal coordinate of ) and (reference coordinate system (X EGO ,Y EGO) It is the name given to the angle that separates the tangent to the lane boundary line R at the abscissa point X (expressed within).

[0090] Therefore, it is possible to write as follows. [Equation 4] Angle LineX / EGO = arctan(d(yLine(x)) / dx)

[0091] Here, [Equation 5] TIFF0007863569000003.tif10170

[0092] [[ID=Q22]]Therefore, for x = 0, it is possible to write as follows. [Equation 6] Angle LineX / EGO = Angle lineEGO / EGO = arctan(b)

[0093] For the abscissa point x = Distance Xproj it is possible to write as follows. [Equation 7] Angle Obj / LineObj = Angle<I Obj / EGO - Angle lineObj / EGO

[0094] The computer can calculate the longitudinal component Vx lineEGO , Y lineEGO ) of the speed of the host vehicle 10 within the reference coordinate system (X EGO / LineEGO [[ID=QSS]]and the lateral component Vy EGO / LineEGO by the following equations. [Equation 8] TIFF0007863569000004.tif8170[Equation 9] .. TIFF0007863569000005.tif8170

[0095] In these equations, - V EGO / abs is measured by a sensor located, for example, on the axle of the vehicle, in the absolute reference coordinate system (Xabs , Y abs is the speed of the host vehicle 10 within it. - Angle VEgo / Ego is the angle of the speed vector of the host vehicle 10 with respect to the abscissa of the reference coordinate system (X EGO , Y EGO ). This angle is assumed to be 0 here.

[0096] The computer can also calculate the longitudinal component Vx Obj / abs and the lateral component Vy Obj / abs of the speed V of the automobile C1 in the absolute reference coordinate system. For this purpose, the computer uses the following equations. Obj / abs [Equation 10] Vx Obj / abs = Vx Obj / EGO + Vx Ego / abs [Equation 11] Vy Obj / abs = Vy Obj / EGO + Vy Ego / abs

[0097] In these equations, - Vx EGO / abs and Vy EGO / abs are the components of the speed of the host vehicle 10 along the abscissa and ordinate in the absolute reference coordinate system (X abs , Y abs ). - Vx Obj / EGO and Vy Obj / EGO are the components of the speed of the automobile C1 with respect to the host vehicle 10 along the abscissa and ordinate in the reference coordinate system (X EGO , Y EGO ).

[0098] Therefore, it can be written as follows. [Equation 12] TIFF0007863569000006.tif13170

[0099] As shown by the following two equations, the angles Angle lineObj / EGO and Angle lineEGO / EGOBased on this, the component Vx of the relative velocity of car C1 "along the lane boundary line R" with respect to the lane boundary line R at projection point F. Obj / LineObj , Vy Obj / LineObj This makes it possible to determine this, thereby enabling better representativeness of the information. [Formula 13] TIFF0007863569000007.tif9170[Formula 14] TIFF0007863569000008.tif9170

[0100] In these two equations, Angle VObj / Obj This is the angle of the velocity vector of car C1 in the reference coordinate system associated with car C1, and Angle Obj / EGO The reference coordinate system (X EGO ,Y EGO This is the angle of the direction of travel of the vehicle within the parentheses.

[0101] In reality, the velocity vector of the object is collinear with the angle of the object's direction of motion, and as a result, the angle VObj / Obj It is assumed that this is 0.

[0102] A similar process is applied to determine the relative speed "along the lane boundary line R" between the vehicle and the lane boundary line R at the 0 lateral coordinate point, and between the vehicle C1 and the lane boundary line R at the projection point F.

[0103] Therefore, the computer calculates the reference coordinate system (X) using the following equation. lineEGO ,Y lineEGO The longitudinal component Ax of the acceleration of the vehicle 10 within ) EGO / LineEGO and the lateral component Ay EGO / LineEGO It is possible to calculate this. [Formula 15] TIFF0007863569000009.tif9170[Formula 16] TIFF0007863569000010.tif8170

[0104] The computer also calculates the reference coordinate system (X) using the following equation. lineObj ,YlineObj The longitudinal component Ax of the acceleration of car C1 inside ) Obj / LineObj and the lateral component Ay Obj / LineObj It can also be calculated. [Formula 17] TIFF0007863569000011.tif8170[Formula 18] TIFF0007863569000012.tif8170

[0105] In these formulas, - A EGO / abs This is the absolute acceleration of the vehicle 10 within the absolute reference coordinate system. - A Obj / abs This is the absolute acceleration of car C1 in the absolute reference coordinate system.

[0106] Next, using the four equations defined below, we obtain the longitudinal components VRelRoute of the relative velocity and relative acceleration of the vehicle and vehicle C1 related to the road being traveled. Longi ARelRoute Longi and the lateral component VRelRoute Lat ARelRoute Lat To obtain this, it is possible to combine the calculated velocity and calculated acceleration.

[0107] In reality, the longitudinal component of the relative velocity between your vehicle and car C1 is VRelRoute. Longi This refers to a reference coordinate system (X) that is linked to the traffic lane at the height of the vehicle. lineEGO ,Y lineEGO The longitudinal component of the vehicle's speed expressed within (X), and the reference coordinate system (X) connected to the traffic lane at the height of vehicle C1. lineObj ,Y lineObj This can be considered equivalent to the deviation between the longitudinal component of the velocity of automobile C1 expressed within the given space.

[0108] Similarly, the lateral component of the relative velocity between your vehicle and car C1, VRelRoute LatThis refers to a reference coordinate system (X) that is linked to the traffic lane at the height of the vehicle. lineEGO ,Y lineEGO The lateral component of the vehicle's speed expressed within (X) and the reference coordinate system (X) linked to the traffic lane at the height of vehicle C1 on the other side. lineObj ,Y lineObj This can be considered equivalent to the deviation between the lateral component of the velocity of car C1 expressed within the given space.

[0109] Therefore, it is possible to write it as follows: [Formula 19] VRelRoute Longi =Vx Obj / LineObj -Vx EGO / LineEGO [Formula 20] VRelRoute Lat =Vy Obj / LineObj -Vy EGO / LineEGO

[0110] It is possible to calculate the components of acceleration in a similar manner. [Formula 21] ARelRoute Longi =Ax Obj / LineObj -Ax EGO / LineEGO [Formula 22] ARelRoute Lat =Ay Obj / LineObj -Ay EGO / LineEGO

[0111] As will be revealed in detail in the remainder of this disclosure, using relative velocity makes it possible to provide an indicator of collision risk that would be difficult to obtain in other ways.

[0112] At this stage, computer C10 calculates the distance between the lane boundary line (at projection point F) and the anchor point of vehicle C1. Target2Lane It can be recalled that the value of [the value] can be determined.

[0113] In the third substep, computer C10 determines the projection point F and point P of car C1 that is closest to the lane boundary line R. prox distance between Lane DY Determine this (see Figure 6).

[0114] The computer calculates this distance using the following formula: [Formula 23] TIFF0007863569000013.tif18170

[0115] In this formula, the term Width corresponds to the width of the car C1.

[0116] Next, the set of calculations allows, in the fourth substep, to determine the time to collision (TTC) with the object under consideration (car C1), i.e., the time required for the vehicle to collide with car C1, assuming both maintain their speeds.

[0117] In fact, at this stage, the computer can determine from equation [equation 2] the length of the arc L separating its own vehicle 10 from automobile C1. AB This can be understood. The computer can also obtain from equation [Equation 19] the longitudinal component of the relative velocity between the vehicle 10 and the vehicle C1, which is referenced with respect to the shape of the road, VRelRoute. Longi This can also be understood. Finally, from equation [Equation 21], the computer can obtain the corresponding longitudinal component of acceleration, ARelRoute. Longi You can see that.

[0118] By using these longitudinal components, it becomes possible to obtain a good approximation of the collision time (TTC) when the road is curved and the vehicle does not have a parallel trajectory.

[0119] Here, computer C10 then determines the sought collision margin time TTC using the following formula: [Formula 24] TIFF0007863569000014.tif14170

[0120] It should be noted that two validity conditions for this equation must be satisfied beforehand. These conditions are as follows: [Formula 25] VRelRoute 2 Longi +2*ARelRoute Longi *L AB ≥0 and ARelRoute Longi ≠0

[0121] Conversely, the longitudinal component of the relativization rate is ARelRoute. Longi If it is 0, it can be written as follows: [Formula 26] TIFF0007863569000015.tif11170

[0122] As a variation, for example, it was possible to calculate the collision time limit (TTC) in a different way by assuming that the relative velocity and / or relativizing velocity are constant.

[0123] In summary, at this stage, the computer, by exploring the fused data, retains various parameters that characterize the various objects located in the surroundings and potential obstacles in the orbit. Specifically, for each object, the computer retains the following: - Collision margin time TTC (Equation [Equation 24]). - The position of the object on the road (determined in step 2). - Information confirming the existence of an object (provided in the data fusion step).

[0124] Next, in the fourth step, computer C10 performs a first filtering of the various detected objects based on parameters held by the computer, so as to retain only those related to the implementation of the AES function (i.e., those that form potential obstacles).

[0125] Therefore, the filtering operation relies on considering that the relevant object (hereinafter referred to as "target") is an object whose existence has been verified during data fusion, whose location may be dangerous (in the example of the present invention, this is equivalent to checking that the object is located within a single traffic lane), and whose collision time to cross (TTC) is below a predetermined threshold.

[0126] If multiple targets are detected within the exact same traffic lane, it is also possible to consider only a limited number of those targets (for example, 4), specifically those that are the closest to the vehicle itself.

[0127] The fifth step involves computer C10 identifying the lateral trajectory deviations (or overlaps) required to avoid each target or group of targets to the right and left, while simultaneously avoiding other objects present on the road.

[0128] This step is carried out in five substeps.

[0129] Prior to the first substep, computer C10 identifies each target through criteria specific to that target.

[0130] Figure 7 shows an example of a situation in which four targets are positioned around vehicle 10 in front of vehicle 10.

[0131] For example, each target is identified by a computer by a criterion written here in the form Cn, where n is a natural integer equal to 1, 2, 3, or 4.

[0132] The number of target Cn elements, n, is given randomly here.

[0133] The first substep involves considering each target Cn independently and calculating the deviation ovLn required to avoid this target Cn to the right and the deviation ovRn required to avoid this target Cn to the left.

[0134] Figure 7 shows the deviations ovL1 and ovR1 that should be taken to avoid target C1.

[0135] These right and left deviations ovLn and ovRn are determined by taking into account the trajectory and fusion data of the vehicle 10. In fact, data fusion provides kinematic information related to the target relative to the vehicle 10, which is used to calculate these deviations, along with the vehicle's trajectory. Thus, these deviations are dynamically calculated based on the movement of the vehicle 10, the movement of the target, and the shape of the lane.

[0136] The example in Figure 7 corresponds to a case where the vehicle 10 and the target are moving within a straight lane.

[0137] In this context, it is observed that the value of deviation ovRn is greater than 0 when a deviation should be taken to avoid the target to the right. Otherwise, this value is 0 or less. More specifically, if vehicle 10 does not need to change its trajectory to avoid the target while simultaneously passing as close to the target as possible, then deviation ovRn is equal to 0. In contrast, if vehicle 10 does not need to change its trajectory to avoid the target, but needs to change it if it desires to pass as close to the target as possible, then deviation ovRn is strictly less than 0.

[0138] Similarly, if a deviation should be taken to avoid the target to the left, the value of deviation ovLn is greater than 0. Otherwise, this value is 0 or less.

[0139] If the road is not straight, it is suggested that these deviation calculations be reviewed taking the following information into consideration. - The lateral component of the relative velocity between the vehicle and the target being considered, along the road being traveled (VRelRoute) Lat (Equation [Equation 20]). - AngleObj / LineObj. - Collision clearance time (TTC).

[0140] The first of these information items allows for the incorporation of the actual lateral speed between the vehicle 10 and the target, taking into account the shape of the traffic lane.

[0141] To gain a good understanding of the advantages of this parameter, one can consider an example where the vehicle and the target are traveling in opposite directions within two separate traffic lanes, precisely following the curvature of these two lanes. In this case, theoretically, the risk of an accident would be understood to be zero. In the example of the present invention, the lateral component of the relative velocity of the vehicle with respect to the target is VRelRoute. Lat This becomes 0, which means the calculated deviation is less than or equal to 0, which precisely represents the idea that the risk of collision is theoretically 0.

[0142] In other words, the lateral component VRelRoute Lat The collision margin time (TTC) allows for weighting of the influence of lateral and longitudinal relative velocities on the calculation of left and right deviations ovLn and ovRn.

[0143] Similarly, information Angle Obj / LineObj This allows for the determination of more precise values ​​for the impact surface of the target being considered by weighting the length and width of the vehicle.

[0144] These right and left deviations E right , E left Here, the calculation is performed using a method corresponding to that described in reference FR1907351, but these calculations differ in that they take into account the three information items mentioned above.

[0145] Therefore, half the width of the target is the information angle. Obj / LineObj It is calculated based on the following. This half-width is then used to calculate a preliminary value for each deviation that does not take into account the safety radius for avoiding the target. Lateral component VRelRoute LatThat portion is then multiplied by the collision margin time TTC so that it is added to this pre-deviation to obtain the desired deviation.

[0146] In other words, when considering reference FR1907351 to calculate the deviation, the lateral component VRelRoute is used in the calculation of the horizontal coordinate dVy. Lat It is necessary to use the product of the collision margin time TTC. The coordinate Y used in the relevant document A (Here, coordinate Y rel (corresponding to) the coordinates resulting from data fusion for that part, and the length and angle of the target. Obj / LineObj It can be considered to be equal to the sum of terms that are equal to the product of with the cosine.

[0147] Therefore, it is possible to write it as follows: [Formula 27] TIFF0007863569000016.tif7170

[0148] In this equation, er is a term used to compensate for lateral measurement errors, and Long is the length of the target. [Formula 28] TIFF0007863569000017.tif9170

[0149] When data fusion is used, term er also takes into account the errors arising from data fusion. This term is predetermined and stored in the computer's memory.

[0150] The second substep involves sorting the targets in order to rank them in a sequence that depends on their position on the road, and more precisely, on their deviation from one of the road's edges.

[0151] Here, this operation is performed based on the calculated left deviation ovLn in descending order. As a variation, it is of course possible to apply a different sorting method.

[0152] The advantage of the method used here is that the deviation calculation takes into account the relative motion of the scene, based not only on the lateral relative velocities of the vehicle 10 and the target, but also on the calculated collision timeout (TTC), which incorporates the concept of predicting the relative positions of the vehicle and the target.

[0153] Here, as shown in Figure 8, the target previously referred to as Cn is now referred to as C n It is referred to as such.

[0154] The following can be seen in Figures 7 and 8. - Target C1 becomes target C2. - Target C2 becomes target C4. - Target C3 becomes target C3. - Target C4 becomes target C1.

[0155] Similarly, the deviations that have been referred to as ovLn and ovRn are now referred to as ovL n , ovR n It is referred to as such.

[0156] Therefore, the following applies: - Deviations ovL1 and ovR1 become ovL2 and ovR2. - Deviations ovL2 and ovR2 become ovL4 and ovR4. - Deviations ovL3 and ovR3 become ovL3 and ovR3. - Deviations ovL4 and ovR4 become ovL1 and ovR1.

[0157] This classification of targets allows for ranking the targets from left to right relative to the vehicle 10.

[0158] From then on, target C n To show the calculated collision margin for TTC, n Use this.

[0159] Taking into account each target that is consecutive in the order thus determined, the computer C10 can then determine whether it is possible to move to the right or left of this target.

[0160] For this purpose, in the third substep, the computer solves the following equation (for all values ​​of n ranging from 1 to 4 in the example of this invention): [Formula 29] Gap Left_n =-(ovL n +ovR n-1 )-dSafe [Formula 30] Gap Right_n =-(ovR n +ovL n+1 )-dSafe

[0161] In these formulas, the parameter dSafe has a strictly positive value and corresponds to the safety distance that is desired to be formed around the target to avoid moving too close to it. The value of dSafe may vary based, for example, on the target's speed and the vehicle's or traffic conditions (such as weather). This value is at least equal to the width of the vehicle.

[0162] From now on, the parameter called the left gap Left_n This corresponds to the width needed to move to the left of the target, taking other targets into consideration.

[0163] From now on, the parameter called the right gap Right_n This corresponds to the width needed to move to the right of the target, taking other targets into consideration.

[0164] Here, we know that it is possible to move to the left of target C1, so the left gap Left_n It should be noted that, consequently, the calculation is not performed for n that is equal to 1.

[0165] Similarly, since it is known that it is possible to move to the left of target C4, the right gap Right_nNote that for n equal to 4, it is not necessarily calculated.

[0166] At this stage, when the left gap Gap Left_n is 0 or more and only in that case, the host vehicle 10 can be considered to be able to move to the left of the target C n under consideration.

[0167] Similarly, when the right gap Gap Right_n is 0 or more and only in that case, the host vehicle 10 can be considered to be able to move to the right of the target C n under consideration.

[0168] Therefore, in the example of FIG. 8, the following is obtained. - Gap Right_1 ≧0. - Gap Left_2 ≧0. - Gap Right_2 <0. [[ID=(34)]]- Gap Left_3 <0. - Gap Right_3 <0. - Gap Left_4 <0.

[0169] This situation is shown in FIG. 9.

[0170] The values of Gap Right_2 , Gap Left_3 , Gap Right_3 , Gap Left_4 are all observed to be strictly less than 0, which means that it is not possible to move between the targets C2, C3, and C4 a priori.

[0171] The values of Gap Right_4 and Gap Left_2 are also observed to be 0 or more, which means that it is possible to move to either side of this group consisting of the targets C2, C3, and C4. Since the term Gap Right_1 is 0 or more, it is also possible to move to the right of the target C1.

[0172] Therefore, at this stage, the computer can group targets when it is not possible to move between them a priori.

[0173] For this purpose, at least one criterion of proximity between targets is used.

[0174] The first proximity criterion relates to the lateral distance between targets.

[0175] Therefore, in order to form a group, computer C10 considers target C n and adjacent target C n+1 Gap between (in a sequence of events) Right_n or Gap Left_n+1 Identify whether the value is strictly less than 0. If so, the two targets are grouped together.

[0176] Therefore, in the example in Figure 9, this gives us a group consisting of three targets, which will hereafter be called preliminary group G1.

[0177] The process of combining a target into a group consisting of multiple targets can be stopped there.

[0178] However, it is now stipulated that a second proximity criterion be considered for forming a group. This second proximity criterion relates to the longitudinal distance between targets.

[0179] Specifically, this idea is to form a group only if it is not possible for the vehicle 10 to fit among some of the targets of the reserve group G1.

[0180] For this purpose, computer C10 is programmed to divide one or more preliminary groups G1 when the preliminary groups include targets that are longitudinally separated from each other.

[0181] This sorting of targets within each group is here based on the collision time to collision (TTC) n for each target C n and is performed by comparing the TTCs. As a variant, this sorting may be performed using other parameters, in particular the arc distance L AB . The advantage of using this parameter TTC n is that it takes into account the longitudinal relative speed and acceleration of target C n .

[0182] Computer C10 acts in the same way on each group of targets consisting of a plurality of targets.

[0183] Computer C10 ranks the targets of this group, for example in descending order, based on the collision time to collision of the targets.

[0184] In the example of FIG. 10, the targets are then ranked in the order C4, C2, C3.

[0185] The computer then calculates the deviation between each pair of consecutive targets (in the determined order).

[0186] Thus, in the example of FIG. 10, the computer calculates the following: - Δ between the collision time to collision TTC4 associated with target C4 and the collision time to collision TTC2 associated with target C2 4-2 . - Δ between the collision time to collision TTC2 associated with target C2 and the collision time to collision TTC3 associated with target C3 2-3 . [[ID=…]]

[0187] [[ID=…]] [[ID=…]] The computer then compares each of these deviations Δ 4-2 , Δ 2-3 with a threshold value Sx.

[0188] This threshold value may be a constant. However, preferably, the threshold value is selected based at least on the longitudinal speed of the host vehicle 10. Note: In the original text, there are some tags like

[0182] , ,

[0183] , , ,

[0185] , ,

[0186] , ,

[0187] , ,

[0188] , which seem to be incomplete or have some internal reference issues. I've left them as they are in the translation with appropriate ellipsis where necessary. If there's more context available for these tags, the translation could be more precise.

[0189] If the deviation between the collision margin time associated with two consecutive targets is greater than this threshold, the computer splits the reserve group into two groups.

[0190] In the example shown in Figure 10, at the end of this operation, targets C4 and C2 subsequently form the first group G2, while target C3 is separated.

[0191] Therefore, at this stage, the computer considers each group as an equally isolated target. Thus, the computer considers the left deviation ovL i Right deviation ovR i , and collision margin time TTC i Associate it with a group (i is the index of the group being considered).

[0192] At this time, these parameters are calculated as follows:

[0193] For easier understanding, an example in Figure 11 may be referenced, where the group G3 under consideration includes three targets C1, C2, and C3.

[0194] Group collision buffer time TTC i This refers to the collision margin time (TTC) for group targets C1, C2, and C3. n The minimum collision time (TTC) among them n It is selected so as to be equal to [the given value].

[0195] Group left deviation ovL i This is the left deviation of the group's target ovL n The largest left deviation among them (ovL) n It is selected so as to be equal to [the given value].

[0196] Group right deviation ovR i This is the right deviation of the group's target ovR n From among them, the largest rightward deviation ovR n It is selected so as to be equal to [the given value].

[0197] For the remainder of this disclosure, the general term “target” is used to specify both a group of multiple targets and a single target that does not form part of any group.

[0198] The sixth step involves performing a second filtering among targets in order to distinguish critical targets from other targets.

[0199] A target is considered critical if it requires the activation of AES (Autonomous Anti-Screen) functions to be evaded. The most critical targets (target MCTs) are those that require the earliest possible activation of AES functions.

[0200] A target is described as "medium risk" if its location should be taken into consideration when determining the evasive trajectory to follow. Therefore, medium-risk targets tend to hinder the activation of AES functions.

[0201] The idea is to consider each detected target sequentially and independently (and therefore independently of its surroundings).

[0202] Initially, computer C10 determines that all targets located within the traffic lane of its own vehicle 10 are in danger.

[0203] With regard to targets located in lanes adjacent to the lane in which vehicle 10 is traveling, the computer checks whether those targets comply with additional criteria.

[0204] Here, these criteria relate to the following parameters. - arc distance L AB . - The distance between the projection point F and the target point closest to the lane boundary line. DY (Equation [Equation 23]). - The lateral component Vy of the target's velocity relative to the lane boundary line at projection point F Obj / lineObjLane: The distance between the nearest target point and the target point. DY (Equation [Equation 14]).

[0205] Figure 12 shows three targets C4, C5, and C6 located within two lanes adjacent to the lane in which the vehicle 10 is traveling.

[0206] To determine whether each target is critical, the computer checks whether the following two conditions i) and ii) are met.

[0207] To check the first condition i), the computer calculates the distance between the target and the lane boundary line being considered. DY We begin by calculating (Equation [Equation 23]). This makes it possible to determine whether the target is relatively close to or relatively far from the lane boundary line being considered.

[0208] Computer C10, from that distance, Vy thresholdMin We estimate the minimum lateral velocity threshold, which is shown as follows.

[0209] Next, component Vy Obj / lineObj The minimum threshold Vy thresholdMin If it exceeds this value, the first condition (for considering the target critical) is met. Otherwise, the target is simply considered medium risk.

[0210] Therefore, it should be noted that the threshold used is a variable that depends on the distance between the target and the lane boundary line being considered, thereby allowing us to take into account the fact that the smaller this distance, the greater the risk of collision.

[0211] The threshold used may further depend on the mode in which the target was characterized in the previous increment (critical or non-critical). Specifically, this idea assumes that the target's characterization does not change in each increment due to noise in the measurement of the data to be fused. For this purpose, the threshold used for the target to change from non-critical to critical is higher than the threshold used for the target to change from critical to non-critical (similar to a hysteresis function).

[0212] The second condition (ii) allows for the exclusion of targets resulting from possible perceptual errors, along with those having excessively large or abnormal lateral velocities. For this purpose, the computer considers the lateral component Vy Obj / lineObj The absolute value of a predetermined maximum threshold Vy thresholdMax This is compared to the other component. If this component is greater than the maximum threshold, the second condition is not met, and the target is considered to be of medium risk. The maximum threshold should be strict to avoid considering false positives, which sometimes have unreasonable values.

[0213] This maximum threshold is preferably greater than 2 m / s and preferably equal to 3 m / s.

[0214] Targets that do not meet either condition i) or ii) and / or the other are considered to be medium-risk targets. Other targets are considered to be critical.

[0215] Classifying perceived targets into critical and medium-risk targets reduces computation time. Furthermore, this classification of targets simplifies the decision-making process regarding whether to activate the AES system and allows for the justification of the decision made for human drivers.

[0216] In Figure 12, since neither of the other two targets satisfies condition i), it can be concluded that only target C4 satisfies both conditions i) and ii).

[0217] The seventh step involves determining a critical time Tcrit for each critical target for computer C10, which integrates two separate information items related to ensuring obstacle avoidance while minimizing the intrusive nature of the AES function on the driver.

[0218] The advantage of calculating this parameter is that it makes it possible to determine which of the critical targets is the most critical target MCT.

[0219] Before explaining how this parameter is obtained, it should be noted that the AES avoidance system may operate fully automatically (in which case the assist steering actuator autonomously follows the avoidance trajectory) or semi-automatically (in which case the avoidance is performed manually by the driver, and the assist steering actuator is controlled to assist the driver by following the avoidance trajectory when the driver triggers the avoidance). In the remainder of this specification, the terms autonomous mode and manual mode are used to specify these two methods, respectively.

[0220] Avoidance in manual mode may be less effective than avoidance in autonomous mode. Therefore, the avoidance trajectories calculated by the AES system will not be the same in manual and autonomous modes. It should be noted that the calculation of this avoidance trajectory (in clothoid form) is not the subject of this invention. Simply put, the shape of this trajectory is calculated based on the vehicle's dynamic performance and (in manual mode) the driver's ability.

[0221] In any case, the calculated deviation ovL n , ovR n Taking this into consideration, it is possible to construct four evasion trajectories (four clothoids) to avoid critical targets to the right and left in both manual and autonomous modes.

[0222] These clothoids have a curvature that depends on the vehicle's maximum dynamic capability. The shape of these clothoids is then retrieved from a database based on the vehicle's speed.

[0223] Taking these deviations and determined avoidance trajectories into account, it is possible to estimate four operation time (TTS) (see Figure 7) corresponding to the time required to perform an avoidance maneuver to the left or right in each mode.

[0224] The manner in which these operation times (TTS) are obtained depends on the dynamic characteristics of the vehicle 10 and the driver's performance, and therefore will not be described here. In practice, these operation times may be read from a database established using a test battery.

[0225] Time to Control (TTC) for each critical target n And, knowing the operation time TTS, the computer can estimate the critical time Tcrit from those values ​​using the following formula. [Formula 31] Tcrit=TTC-TTS

[0226] Therefore, this critical time becomes 0 at the last moment when it is still possible to avoid a collision with the critical target by activating the AES function and evading the critical target on the considered side.

[0227] Therefore, this critical time Tcrit is calculated based on three essential pieces of information about the target being considered: the side that will avoid the target, the performance of the system (and driver), and the time to react before impact with the target. n To carry.

[0228] Next, the eighth step involves classifying the critical targets using the calculated critical time Tcrit and determining the most critical target MCT.

[0229] If the driver is in manual mode and unable to activate autonomous mode, the computer will select the critical target with the shortest critical time when evading to the right, and the critical target with the shortest critical time when evading to the left.

[0230] Next, as soon as computer C10 detects that the driver has initiated an evasive maneuver, the computer can decide to activate the AES system to assist the driver's actions based on two selected critical time zones.

[0231] To this end, the computer determines which side the driver is turning the steering wheel to when initiating an evasive maneuver (to the right or left), and then selects from two chosen critical moments the one corresponding to the side on which the driver initiated the evasive maneuver.

[0232] The most critically targeted MCTs are those that correspond to selected critical time periods.

[0233] In autonomous mode, the computer operates differently. In fact, it needs to determine which side the vehicle 10 should avoid when encountering one or more obstacles.

[0234] At this point, the idea is to determine the best course of action for each critical target to avoid it, and then determine the most critical target (MCT).

[0235] In practice, the computer selects the side with the highest critical time Tcrit for each critical target.

[0236] At this point, the computer selects the shortest critical time from among the selected critical times.

[0237] The computer then determines that the most critical target MCT is the one selected during critical time Tcrit.

[0238] The ninth step involves triggering the AES function on computer C10 as needed and at the optimal moment.

[0239] In this case, in autonomous mode, the AES function is triggered immediately when the selected critical time Tcrit falls below a predetermined threshold, for example, equal to 0 seconds.

[0240] In contrast, in manual mode, this AES function is triggered differently.

[0241] The idea is to trigger the AES function and have the driver take evasive action in a manner that is adapted to the situation.

[0242] In fact, if the driver starts turning the steering wheel too early, the situation does not warrant action from this type of driver at that point, and therefore those actions are considered non-urgent. In this case, the AES system will not be activated.

[0243] Similarly, if a collision is imminent and autonomous mode could not be activated beforehand, it may be too late for the AES system to provide optional assistance to the driver. In this case, the computer will hand over responsibility to another safety system to minimize the impact.

[0244] Therefore, in manual mode, it is necessary to determine the time interval at which the AES system is activated when an evasive action taken by the driver is detected.

[0245] This time interval is divided by two boundaries.

[0246] Starting from there, the first boundary at which the driver is assisted when initiating an evasive maneuver corresponds to a critical time Tcrit that is strictly greater than 0.

[0247] Starting from there, the second boundary at which the AES system is too late to trigger corresponds to a critical time Tcrit that is less than or equal to 0, and preferably strictly less than 0.

[0248] Therefore, in order to activate the AES function, the computer determines whether the selected critical time Tcrit lies between these two boundaries, and only activates the AES function if this is the case.

[0249] The present invention is by no means limited to the embodiments described and illustrated, and those skilled in the art will be able to see how to provide any variation of embodiments according to the present invention.

[0250] A single proximity criterion may be considered that takes into account only the deviation between targets, and not the lateral or longitudinal direction.

[0251] In another variation, the first step may involve grouping the vehicles according to their longitudinal deviation, and then dividing each thus obtained group according to the lateral deviation between the vehicles in the group, if necessary.

Claims

1. A method for avoiding objects (C1, C2, C3, C4) for the sake of an automobile (10), - A step of detecting objects (C1, C2, C3, C4) located around the automobile (10), - A step of obtaining data characterizing the position and / or movement of each detected object (C1, C2, C3, C4) and Includes, When multiple objects (C1, C2, C3, C4) are detected, - A step of checking whether at least one proximity criterion is met regarding the proximity between at least two of the detected objects (C1, C2, C3, C4), and if it is met, - The step of combining the two objects (C1, C2, C3, C4) into one group (G1, G2, G3), - A step of calculating data characterizing the position and / or movement of the group (G1, G2, G3), - The method is characterized by defining the steps of activating an obstacle avoidance system (AES) according to data characterizing the position and / or movement of the group (G1, G2, G3), and / or determining an avoidance trajectory (T1) according to data characterizing the position and / or movement of the group (G1, G2, G3), The proximity criterion relates to the lateral distance of the automobile (10) between the two objects (C1, C2, C3, C4), A method comprising the step of acquiring data, wherein one of the data characterizing the position and / or movement of each detected object (C1, C2, C3, C4) is the lateral trajectory deviation (E left, E right) that the automobile (10) must take to avoid each object (C1, C2, C3, C4), and the proximity criterion is checking whether the difference (Gap Left_n, Gap Right_n) between the lateral trajectory deviation (E left) that the automobile (10) must take to avoid the first object on the side facing the second object of the two objects, and the lateral trajectory deviation (E right) that the automobile (10) must take to avoid the second object on the side facing the first object, is greater than or equal to a predetermined threshold (d safe).

2. The avoidance method according to claim 1, wherein if at least three objects (C1, C2, C3, C4) are detected, the objects (C1, C2, C3, C4) are to be ranked in the order they are arranged from one edge of the road to the other edge, and it is checked whether the proximity criterion is met between each pair of adjacent objects in the order they are arranged.

3. On the other hand, a first reference coordinate system (X) oriented along the tangent to the road at the height of the automobile (10) ＬｉｎｅＥｇｏ , Y ＬｉｎｅＥｇｏ The lateral speed (Vy) of the vehicle (10) relative to the road on which the vehicle (10) is traveling within ) Ｅｇｏ／ＬｉｎｅＥｇｏ ) and on the other side, a second reference coordinate system (X) oriented along the tangent to the road at the height of the object (C1, C2) ＬｉｎｅＯｂｊ , Y ＬｉｎｅＯｂｊ Based on the deviation between the lateral velocity (VyObj / LineObj) of the objects (C1, C2) relative to the road within ) and the relative lateral velocity (VRelRoute) Ｌａｔ It is stipulated that the relative lateral velocity (VRelRoute) is calculated for each lateral trajectory deviation (Elevt, Eight). Ｌａｔ The avoidance method according to claim 1 or 2, determined based on ).

4. The avoidance method according to any one of claims 1 to 3, wherein the proximity criterion relates to the longitudinal distance of the automobile (10) between the two objects.

5. In the step of obtaining, one of the data relates to the remaining time (TTC) before the motor vehicle (10) collides with each object (C1, C2, C3, C4), and in order to check that the proximity criterion is satisfied, the deviation (Δ ２－３ , Δ ４－２ ) between the collision margin times (TTC) with the two objects is checked to be less than a threshold value. The avoidance method according to claim 4.

6. In the acquisition step, one of the data is the lateral trajectory deviation (E) that should be taken to avoid each object (C1, C2, C3, C4) on the same left or right side. ｌｅｆｔ , E ｒｉｇｈｔ ) and in the calculation step, one of the data characterizing the group is the lateral trajectory deviation (E) that should be taken to avoid each object (C1, C2, C3, C4) of the group (G1, G2, G3) on the same side. ｌｅｆｔ , E ｒｉｇｈｔ The avoidance method according to any one of claims 1 to 5, selected such that it is equal to the largest deviation among them.

7. The avoidance method according to any one of claims 1 to 6, wherein in the acquisition step, one of the data characterizing each object is the remaining time (TTC) before the automobile (10) collides with each object (C1, C2, C3, C4), and in the calculation step, one of the data characterizing the group is selected to be equal to the minimum time of the remaining time (TTC) before the automobile (10) collides with each object (C1, C2, C3, C4) of the group (G1, G2, G3).

8. Automobile (10) comprising at least one steering wheel and a steering system for each steering wheel, which is designed to be operated by an actuator controlled by a computer (C10), wherein the computer (C10) is designed to carry out the avoidance method according to any one of claims 1 to 7.

Images