Driver assistance methods using virtual targets for adaptive cruise control

Abstract

This invention relates to a driver assistance method for a so-called EGO (Electronic Vehicle) traveling in a driving lane, the driver assistance method comprising: a first step: identifying traffic around the EGO in the same driving lane as the EGO and in adjacent parallel lanes in the same direction of travel; a second step: determining a virtual center of gravity target (G), including calculating the position, velocity, and acceleration of the virtual center of gravity target; and a third step: calculating a longitudinal speed setpoint, an acceleration setpoint, and a torque setpoint of the EGO, the longitudinal speed setpoint being a function of the position, velocity, and acceleration of the virtual center of gravity target (G).

Description

Technical Field

[0001] This invention relates to a driver assistance method employing a virtual target for adaptive cruise control. The invention is advantageously applied, in the form of a driver assistance method for this vehicle, to motor vehicles equipped with such a driver assistance system.

[0002] The present invention also relates to a computer program product whose instructions are adapted to implement this method.

[0003] The present invention further relates to a driver assistance system for the present motor vehicle.

[0004] The present invention also relates to a motor vehicle including a transmission system, a braking device, and such an auxiliary system. Background Technology

[0005] Driver assistance technologies are becoming increasingly widespread and are no longer limited to high-end vehicles.

[0006] These technologies can simplify the driving of motor vehicles and / or make the behavior of vehicle drivers more reliable.

[0007] Modern vehicles are typically equipped with adaptive cruise control systems, which generally operate based on adjusting the distance between the equipped vehicle (also known as the vehicle itself) and the vehicle ahead in its lane (also known as the target vehicle or simply the target).

[0008] As is well known, adaptive cruise control (ACC) systems are used to equip motor vehicles.

[0009] This adaptive cruise control system includes devices (such as radar) for detecting the vehicle's environment, and is therefore able to detect other vehicles or objects on the road, and in particular, it is able to detect vehicles in front of the vehicle on the road.

[0010] These systems are designed, for example, to control a vehicle so that its speed is equal to a set point set by the driver, unless an event occurs on the road that requires the vehicle to slow down (following a vehicle whose speed is different from the driver's set point, traffic jams, traffic lights, etc.), in which case the vehicle speed is controlled accordingly.

[0011] Subsequently, the speed of a vehicle equipped with this adaptive cruise control system can be adjusted to maintain a substantially constant safe distance from the vehicle in front. Therefore, this system interacts with the engine's control system and / or braking system to accelerate or decelerate the vehicle.

[0012] Adaptive cruise control systems known in the prior art are sometimes unreliable. Delayed speed correction can cause sudden acceleration or deceleration, leading to discomfort and a lack of security. In particular, this can occur when a target vehicle changes lanes or a vehicle merges into the vehicle's lane.

[0013] An embodiment of a vehicle equipped with a controller employing a classic adaptive cruise control method is described below with reference to FIG1.

[0014] Motor vehicle 10 is any type of motor vehicle, particularly a passenger vehicle or a multi-purpose vehicle. In this document, the vehicle including the means for implementing the invention is referred to as "this" vehicle. This name is only used to distinguish it from other vehicles in the vicinity and does not in itself impose any technical limitations on motor vehicle 10.

[0015] The motor vehicle 10 or the vehicle itself includes a conventional adaptive cruise control device (DISP) employing a classic adaptive cruise control method, specifically in the form of a controller.

[0016] The classic adaptive cruise control device DISP for motor vehicles can be part of a more comprehensive driver assistance system.

[0017] In the remainder of this document, the term "target vehicle" refers to a vehicle located in the surrounding traffic of this vehicle 10, whose kinematic properties (including position, velocity, and acceleration) are taken into account when calculating the longitudinal velocity setpoint of this vehicle.

[0018] The target vehicle can be any type of motor vehicle, especially a passenger vehicle, a multi-purpose vehicle, or a motorcycle.

[0019] The adaptive cruise control (DISP) system for motor vehicles requires perception of vehicles in the surrounding traffic, specifically through sensors C, such as perception sensors and position sensors. In fact, to accurately locate a vehicle, the kinematic properties Att_Target need to be determined via sensors C; these properties represent the relative dynamic presence, position, speed, and acceleration of surrounding vehicles within the lane. Specifically, data from the perception sensors is fused to identify surrounding objects and characterize them using their kinematic properties. These perception sensors can employ various technologies, such as ultrasonic, radar, lidar, and cameras, while position sensors are specifically inertial or odometry units, or GPS global positioning satellite systems. In particular, object characterization allows for understanding the environment around the vehicle and classifying objects by type, such as vehicles (motorcycles, cars, trucks, bicycles, etc.), pedestrians, infrastructure, and traffic lights, enabling the consideration of vehicles as the sole target. Data from these sensors (particularly correlated with maps) also enables the identification of road geometry (slopes, curves, etc.), and furthermore, navigation information allows for the provision of contextual information based on the perceived vehicle's position: region type (urban, suburban, rural), road type (highway, town road, intercity road, etc.), speed limits, all of which can be fused with sign recognition, road traffic information, and geometric information (gradient, curves, number of lanes, etc.). The kinematic properties of the identified target, Att_Target, are then transmitted to the distance control unit CD, which also takes driver data DC as input. This driver data consists of the control speed selected by the driver and the predetermined follow time selected by the driver (default 2 seconds). This driver data is also converted into a predetermined follow distance setpoint selected by the driver based on the vehicle's speed, for example, via a table.

[0020] Based on this information, the distance control unit CD generates a longitudinal speed setpoint Vc as output, enabling the vehicle to automatically maintain a predetermined following distance setpoint relative to the vehicle ahead in the same lane, and automatically adjust its speed to maintain this distance. This speed setpoint Vc, along with the vehicle's speed Vm obtained from measurements taken from sensor C (particularly wheel speed sensors, whose measurements are averaged and correlated with a Kalman filter), is then sent to the input of the speed control unit CV. This CV generates an acceleration setpoint Ac as output, causing the vehicle to automatically increase its acceleration to the driver-selected control speed once the lane ahead becomes clear after a cycle of acceleration data Am obtained from sensor C (particularly from an inertial measurement unit, etc.) or based on measured wheel speeds. This assists the driver in completing their driving tasks. The torque of actuators A (engine, brakes, etc.) is controlled by a torque setpoint Cc generated at the output of the torque control unit CC, which is a function of the combined acceleration command at the end of the cycle. However, the device only considers one target at a time, namely the target in front of the vehicle in its own lane. This makes the device very sensitive, especially to vehicle merging between the vehicle and the vehicle in front of it, as well as to lane changes of the target vehicle, thus reducing the smoothness of control in the event of sudden irregularities at the set point.

[0021] It is also known that document FR 2912981 relates to an autonomous driving method for motor vehicles including an ACC system, the purpose of which is to improve the smoothness of vehicle behavior to improve user comfort; however, in the case of multiple objectives, this method consumes a lot of energy because it requires a lot of computation to determine the dynamic components that the ACC system should track. Summary of the Invention

[0022] One object of the present invention is to overcome at least some of the disadvantages of the prior art by providing a driver assistance method for a vehicle traveling in a driving lane, the driver assistance method comprising:

[0023] - First step: Identify the traffic around the vehicle in its driving lane and in at least one adjacent parallel lane in the same direction of traffic;

[0024] - Second step: Determine the virtual center of gravity target, including calculating the position, velocity, and acceleration of the virtual center of gravity target;

[0025] - Third step: Calculate the longitudinal speed setpoint, acceleration setpoint and torque setpoint of the vehicle. The longitudinal speed setpoint is a function of the position of the virtual center of gravity target, the velocity of the virtual center of gravity target and the acceleration of the virtual center of gravity target.

[0026] Thanks to this invention, the trajectory of a vehicle in the surrounding traffic can be predicted without modifying the ACC control loop itself, only by modifying its input, and multiple targets can be taken into account, including the case where there are no targets in the vehicle's lane, while improving the smoothness of the ACC control loop.

[0027] According to the advantageous feature, the surrounding traffic includes at least two target vehicles traveling in front of the vehicle in the vehicle's lane or in an adjacent parallel lane in the same direction of traffic, which makes it possible to consider not only vehicles to the side, but also vehicles that will slow down in front of the vehicle.

[0028] According to another advantageous feature, the first identification step includes a sub-step of detecting each of the at least two target vehicles, wherein, for each target vehicle, the position of the target vehicle relative to the vehicle itself, the speed of the target vehicle, and the acceleration of the target vehicle are determined as outputs, and in particular, the trajectory of the target vehicle is determined, which allows a safe distance to be predicted by taking into account the predicted trajectory, regardless of whether the position is relative or absolute.

[0029] According to another advantageous feature, the second step of determining the virtual center of gravity target uses a pre-selected control speed, a predetermined follow distance setpoint, and the result of the identification step as input. This allows the virtual center of gravity target to be constructed based on readily available information and takes into account the control speed and follow distance setpoint that can be customized by the driver.

[0030] Advantageously, the second step of determining the virtual center of gravity target includes a filtering step to select only certain targets based on their velocity or their time intervals.

[0031] According to another advantageous feature, at least one of these target vehicles is located in the vehicle's driving lane, which allows for consideration of the phenomenon of the vehicle ahead slowing down, as well as the phenomenon of one of these vehicles ahead of the vehicle leaving the lane.

[0032] An advantage associated with the characteristic that at least one of these target vehicles is located in an adjacent lane and attempts to merge into the vehicle's lane is that it makes it possible to anticipate and account for merging phenomena.

[0033] According to another advantageous feature, the second step in determining the virtual center of gravity target uses a target change prediction coefficient determined for each target, which allows for dynamic weighting of the center of gravity, thereby improving ACC fluency.

[0034] According to another advantageous feature, the target variation coefficient is a function of the relative lateral distance between the trajectory of the vehicle and at least one target vehicle, or a function of the relative lateral distance between the center of the lane in which the vehicle is traveling and at least one target vehicle, which makes it possible to ignore curves on the road.

[0035] According to another advantageous feature, the target variation coefficient is a function of the estimated time when the estimated trajectory of the target vehicle intersects with the estimated trajectory of the vehicle itself, which allows for the application of weighting using additional information provided by the module for identifying surrounding traffic.

[0036] Advantageously, the target change coefficient is a function of the following time of the vehicle in front of the target, which allows for a smooth consideration of the deceleration of vehicles traveling in front of the vehicle in front of the vehicle.

[0037] According to another advantageous feature, the target variation coefficient is a function of the stiffness coefficient, thus ensuring convenient updates, since the stiffness coefficient is unique for all incorporated targets.

[0038] The present invention also relates to a computer program product comprising program code instructions stored on a computer-readable medium, the program code instructions including instructions that cause the computer to perform the method of the present invention when the computer executes the program, the program product having similar advantages to the method, and the program product being easy to install in a motor vehicle computer.

[0039] The present invention also relates to a system for the present motor vehicle moving in a driving lane, the system comprising:

[0040] - A module for identifying traffic around the vehicle in its driving lane and in at least one adjacent parallel lane in the same direction of traffic.

[0041] - A module for determining a virtual center of gravity target, wherein the position, velocity, and acceleration of the virtual center of gravity target are calculated.

[0042] - A module for calculating the longitudinal speed setpoint, acceleration setpoint, and torque setpoint of the vehicle, wherein the longitudinal speed setpoint is a function of the position, velocity, and acceleration of the virtual center of gravity target. This system has similar advantages to methods using devices commonly used on vehicles (such as radar and cameras).

[0043] The present invention also relates to a motor vehicle comprising a drivetrain, acceleration and braking devices, and a driver assistance system according to the invention, which enables easy integration into vehicles equipped with ACC (whether autonomous or non-autonomous). Attached Figure Description

[0044] Other objects, features, and advantages of the invention will become clear from the following description, given only by way of non-limiting examples and with reference to the accompanying drawings, in which:

[0045] [Figure 1] Figure 1 (already mentioned) schematically illustrates a classic existing technology ACC device, and

[0046] [ Figure 2 ] Figure 2 The diagram schematically illustrates a driver assistance system according to the present invention, and

[0047] [ Figure 3 ] Figure 3 The application of the method according to the present invention is demonstrated, and

[0048] [ Figure 4 ] Figure 4 A diagram illustrating the relative lateral distance in the curves in one application of the invention, and

[0049] [ Figure 5 ] Figure 5 An illustration showing trajectory prediction according to the present invention in one usage scenario, and

[0050] [ Figure 6 ] Figure 6 This illustrates another use of the invention. Detailed Implementation

[0051] Throughout this text, direction and orientation are specified with reference to the right-handed coordinate system XYZ, classically used in motor vehicle design, where X represents the longitudinal direction of the vehicle, pointing towards its forward motion; Y is transverse to the vehicle, pointing to the left; and Z is vertical, pointing upwards. The concepts of "front" and "rear" are indicated with reference to the normal forward motion of the vehicle. Throughout the description, the term "substantially" means that small differences are permissible relative to a defined nominal quantity; for example, "substantially constant" means that a difference of approximately 5% is permissible in the context of this invention. For clarity, identical or similar elements are denoted by the same reference numerals in all figures.

[0052] Figure 2The diagram schematically illustrates one embodiment of an adaptive cruise control system 1 for a motor vehicle EGO, which is part of a more general driver assistance system according to one aspect of the invention. Elements of system 1 that are identical to those constituting the device DISP of FIG1 have the same reference numerals.

[0053] Sensors C, such as perception sensors capable of measuring the vehicle's dynamics and sensing the environment, are found in the driver assistance system according to the invention. These sensors C can provide information not only about the vehicle's speed and acceleration, and the position, speed, and acceleration of objects in the environment, but also provide trajectory predictions for these objects. In fact, to accurately predict the vehicle's trajectory, sensors C need to determine environmental information Env, including kinematic properties such as the presence, dynamic relative position, speed, and acceleration of surrounding vehicles in the lane. This environmental information is given, for example, in the vehicle's frame of reference, which is positioned at the level of the vehicle's rear axle, but any other position of the frame of reference is possible. Similarly, environmental information about the positions of other vehicles is typically based on detection of the rear axle. All this environmental information Env, obtained by fusing data from sensors C, is indexed object-by-object and generated at the output of what is referred to herein as an identification module (not shown) (which identifies traffic around the vehicle in its lane and in adjacent parallel lanes in the same direction of traffic), and sent as input to module CBV to determine a virtual center of gravity target. The identification module fuses the aforementioned information from sensors C with information from surrounding vehicles, such as turn indicators and other traffic lights. Then, a change in the position reference frame can be performed upstream or in the CBV module to determine that the distance is no longer axle-to-axle but bumper-to-bumper. Figure 3 As shown in the accompanying figures.

[0054] The CBV module also uses driver data DC, consisting of the driver-selected control speed and the driver-selected predetermined follow time, as input. This driver data is also converted into a driver-selected predetermined follow distance setpoint. As a function of these inputs, the CBV module determines a virtual center of gravity target in the vehicle's lane and calculates the position, velocity, and acceleration of this virtual center of gravity target. This CBV module allows information from various objects to be considered, selecting objects upstream of the vehicle (whether in its lane or to the side) as targets, and inferring the target vehicle whose predicted trajectory has been predetermined, so as to be able to predict the motion of the target vehicle and thus react like a human driver. To filter out small oscillations associated with measurement errors, filters can be added to the CBV module to consider only targets whose relative lateral velocity to the vehicle exceeds a threshold. The CBV module is placed upstream of the control loop of the actuator A subsystem (engine, brakes, etc.) and provides a control setpoint like that of classic ACC targets without modifying the ACC logic, thus facilitating its integration. In fact, the kinematic properties of the virtual center of gravity target, Att_CBV—namely, the virtual center of gravity target's presence on the lane, dynamic relative position, velocity, and acceleration—are generated as the output of the CBV module used to determine the virtual center of gravity target. These are then sent as input to the loop BCD to calculate dynamic setpoints. These dynamic setpoints include the vehicle's longitudinal velocity setpoint Vc and acceleration setpoint Ac, where the longitudinal velocity setpoint is a function of the virtual target's position, velocity, and acceleration. The dynamic setpoint calculation loop BCD incorporates:

[0055] The distance control unit CD takes as input the kinematic properties Att_CBV of the virtual center of gravity target, as well as driver data DC consisting of the control speed and the predetermined follow time selected by the driver. This driver data is also converted into a predetermined follow distance setpoint selected by the driver and outputs a longitudinal speed setpoint Vc, which corresponds to the required speed control amplitude, thereby meeting the driver's expectations in terms of safe distance and overall smoothness.

[0056] - The speed control unit CV, as mentioned earlier, takes the vehicle's speed setpoint Vc and speed Vm as inputs and generates an acceleration setpoint Ac as output.

[0057] - The torque control unit CC, as previously described, uses a combined acceleration command as input when leaving the cycle and generates a torque setpoint Cc for the control actuator A as output. This torque setpoint Cc allows the wheels to be controlled to track the speed setpoint Vc.

[0058] The CBV module for determining a single virtual center of gravity target allows for the consideration of surrounding vehicles that may merge into and become targets of ACC tracking in the near future. However, the CBV module also guarantees normal ACC control when no vehicle is detected in front or to the side, or when there is only one target vehicle in the lane where a vehicle is detected. Advantageously, this method can also be applied to vehicles V0 leaving their lanes to change lanes, a process known as lane departure. Once the vehicle enters a lane to the side of the vehicle, it becomes vehicle Vx, or it will cease indexing if it leaves the proximity range of the vehicle determined by the sensor's sensing distance and / or a predetermined distance from the vehicle (particularly as a function of a follow distance setpoint that can be customized by the driver).

[0059] like Figure 3 As shown, on a straight line, the vehicle's EGO senses two target vehicles, as determined by the module used to identify surrounding traffic: one is vehicle V0 in front of it in its lane, and the other is vehicle V1 slightly ahead and to its left, in the left lane, whose right turn indicator is activated to indicate that it will enter the vehicle's lane. Three lanes in the same traffic direction are represented by short dashed lines, and solid lines represent their separation from potential lanes in the opposite direction. Although in this example, target vehicle V1 is to the left of the vehicle's EGO in the direction of forward movement, alternatively, target vehicle V1 could be to the right of the vehicle's EGO and attempt to enter the vehicle's lane, for example, by activating its left turn indicator. Therefore, the virtual center of gravity target G here corresponds to the system's center of gravity G(A,a), (B,b), where a+b≠0, a and b are weighting coefficients, and for any point O as the origin, we have:

[0060] [Math.1]

[0061]

[0062] The module used to identify surrounding traffic supplies geometric references corresponding to the positions of vehicle V0 and V1, i.e., points A and B, to the module CBV used to determine the virtual center of gravity target G in a reference frame associated with the vehicle (here centered on its rear axle, but an alternative can be made). The module CBV then performs a change in the reference frame to associate the new reference frame of the vehicle with the front of the vehicle's bumper.

[0063] X0 is the distance of the first target vehicle, specifically its rear bumper, within the range of the direct approach of the vehicle in a relative frame of reference associated with this vehicle.

[0064] X1 is the distance in a relative frame of reference to this vehicle at which a second target vehicle, particularly its front bumper, is at risk of entering the approach range of this vehicle.

[0065] and are the relative speeds of the first vehicle and the second vehicle respectively.

[0066] Similarly, and are the relative accelerations of the first vehicle and the second vehicle respectively.

[0067] The purpose of the virtual center of gravity target method is to predict target changes by adding a dynamic offset o_d to the following distance setpoint d_s_c supplied by the driver and applied to the target vehicle V0 on the lane. The sum of the dynamic offset distance o_d and the following distance setpoint d_s_c constitutes the distance between the front of the ego vehicle EGO sharing the same lane and the rear of the target vehicle V0 to be left. The offset o_d is obtained by projecting the virtual center of gravity target G onto the trajectory of the ego vehicle, and then the offset o_d will be transmitted to the module CD together with the position information of the virtual center of gravity target G and other kinematic attributes Att_CBV. The weighting coefficients a and b of the center of gravity are selected according to the target change prediction coefficient. These coefficients can be obtained in two different ways.

[0068] The first way is based on the lateral position of the merging vehicle V1. Accordingly, the closer the lateral target vehicle V1 is to the lane of the ego vehicle EGO, the more likely it is considered to merge into the lane. Preferably, it is not the ordinate Y1 of the vehicle, and preferably the lateral distance Y1' relative to the trajectory of the ego vehicle EGO and / or the center of the lane is used, which coincide here and are represented by long dashed lines. This makes it possible to avoid interference related to the orientation of the ego vehicle EGO in its lane, in the form of the tangent multiplied by the distance to the target (5° at 100 meters = 8 meters of lateral error). Therefore, this method can work in the case of curves.

[0069] Figure 4 Intuitively indicates the error that appears in the curve. The values of Y1 and Y2 correspond to the ordinates of the rear ends (e.g., the middle of the bumper) of the lateral target vehicles V1 and V2 in the relative reference frame related to the ego vehicle, and the values of Y1' and Y2' correspond to the lateral distances from the center of the lane of the ego vehicle EGO to the lateral target vehicles V1 and V2. It can be seen that, regarding the lateral distance, Y1' > Y2', while the ordinate Y1 < Y2. The lateral distance values are particularly supplied by the module for identifying the surrounding traffic under the same conditions as the target position. Therefore, the weighting coefficients a and b are obtained as functions of the lateral distance, particularly a as a function of Y1', and b as a function of the reciprocal of Y1'.

[0070] As Figure 5As shown, the second method involves using the estimated time of intersection with the trajectory of the vehicle's EGO as a target change prediction coefficient, here corresponding to 3 seconds, T+3. The estimated time of intersection with the trajectory of the vehicle's EGO is based on trajectory prediction, both supplied by a module for identifying surrounding traffic under the same conditions as the target location.

[0071] The following uses prediction coefficients based on the lateral position of the merging vehicle V1, but the application is the same as the second method.

[0072] The weighting coefficients a and b are preferably stiffness coefficients k or update coefficients, particularly in the following manner: a = Y1'; b = k / Y1'

[0073] Therefore, the stiffness coefficient k is placed at the level of amplitude X1, representing the amplitude of the vehicle V1. From a behavioral perspective, this coefficient allows for the definition of the strength of the target introduced in the virtual target calculation; the larger k is, the larger the vehicle is expected to be. Given that the problem is symmetric about the x-axis, the coefficient k for the right-hand target and the coefficient k for the left-hand target will be the same, which allows the method to be applied equally to right-hand and left-hand driving countries without any specific modifications. Choosing to use inversely proportional (a = 1 / b) weighting coefficients a and b preserves the proportion modulo the stiffness coefficient, which is crucial for center of gravity balance, its representativeness and proportionality, and facilitates management constraints. The equations for the position, velocity, and acceleration amplitude of the center of gravity G of the target vehicles V0 and V1 are then obtained:

[0074] [Math.2]

[0075]

[0076]

[0077] as well as

[0078]

[0079] These magnitudes are calculated at each time increment, so the center of gravity G is conditioned on the changes in distance to the target, velocity, and acceleration throughout the maneuver, and this is done dynamically. The values ​​of Y1' and k / Y1' are not considered in the derivation because these coefficients are assumed to be dimensionless, and if Y1' corresponds to the lateral distance, this can also be the incorporation time as previously described. X can also be calculated. GThen, the derivative is obtained to obtain the other magnitudes of velocity and acceleration. Furthermore, the fact that the weighting coefficients a and b are functions of the lateral position or merging time of the lateral target V1 allows merging dynamics to be considered. The more lanes the target vehicle V1 merges into the vehicle's EGO lane, the more magnitudes of merging vehicle V1 need to be considered. Advantageously, the three equations require only a single stiffness coefficient k, which greatly simplifies the update. This stiffness coefficient k can be adjusted according to the vehicle's speed to behave differently depending on the situation (free-flowing interstate highway or congested city ring road), and its value can be particularly in the range of [0; 10], and preferably is 1.

[0080] Advantageously, creating a virtual center of gravity target G with distance, velocity, and acceleration does not interfere with the conventional ACC cycle. Therefore, the method of the present invention does not modify the adjustment of the drivetrain (whether internal combustion or electric) control in any way, nor does it modify the adjustment of the braking device in any way. Furthermore, the method advantageously has a spatial representation that can be easily visualized during its updates, such as… Figure 3 As already shown, it graphically illustrates the calculations performed to determine the center of gravity G for generating the positions of the target vehicles V0 and V1. In physics, this is equivalent to adding an offset o_d to the following distance setpoint d_s_c.

[0081] Once vehicle V1 merges into the lane of vehicle EGO between vehicle V0 and vehicle V0, it is indexed as vehicle V0.

[0082] As already mentioned, this method can also advantageously consider targets located on the right or left, and in particular, allows for the consideration of multiple targets, regardless of whether they are located on the right and / or left in front of the vehicle EGO. In a manner similar to target vehicle V1 merging between the vehicle EGO and the vehicle V0 in front of it, the method is therefore applicable to three or more targets, such as... Figure 6 As shown. Then, the virtual centroid target G corresponds to the system's centroids G(A,a), (B,b), (C,c), where a+b+c≠0, a, b, and c are weighting coefficients, and for any point O as the origin, we have:

[0083] [Math.3]

[0084]

[0085] Therefore, we can conclude that:

[0086] [Math.4]

[0087]

[0088] NbTm is the number of vanishing lateral targets to which this method can be applied, including uniform dynamic behavior even when targets are vanished. Furthermore, default values ​​need to be defined for cases where one or more targets are vanished. Therefore, if there are no vehicles in the lane of this vehicle's EGO, X0 is set to equal the following distance setpoint d_s_c, and... Equal to the control speed selected by the driver, take Equal to 0, and similarly, in the absence of a lateral target Vi, take X. i , Equal to 0, and take Y i 'Equals 1. Therefore, if target vehicles V1 and V2 disappear, then X...' G =X0.

[0089] The ACC target problem is adhering to the following time relative to the target in front of it. Therefore, if the following time between two vehicles in front of it in its lane is dangerous, maintaining an additional safe distance is also in the vehicle's interest. Thus, this method can monitor not only the merging of lateral target vehicles but also the behavior of vehicles in front of the vehicle it is following. In fact, this method can apply not only to multiple targets V1, V2 attempting to merge in front of the vehicle's EGO but also to vehicles VP in front of the vehicle V0 following the vehicle's EGO. Indeed, as... Figure 6 As shown, in its lane, this vehicle EGO follows vehicle V0, which is itself following vehicle VP (referred to as the target vehicle in front), and in the transverse lane, there are vehicles V1 to the left front of this vehicle and V2 to the right front of this vehicle. First, if we only focus on this vehicle EGO, the vehicle V0 in front of it, and the target vehicle VP in front of it, we can use X... P Write down the distance between the vehicle's EGO and the target vehicle's VP, as its derivative with respect to time. and the stiffness coefficient k as the condition before the target is reached P and use and Write down the weighting coefficients and the individual stiffness coefficient k. P Placed in amplitude X P Level:

[0090] [Math.5]

[0091]

[0092] The weighting coefficients a and b here depend on the following time of the target vehicle VP, which physically represents the time when the trajectories intersect in the longitudinal coordinate. Therefore, both targets V0 and VP are always considered, and to ensure the vehicle is unaffected by every speed change of the target vehicle VP, filtering with respect to a threshold speed is preferably performed so that the target vehicle VP is considered only if its speed relative to V0 is below a threshold. Thus, the target vehicle VP is considered only when the difference between the target vehicle speed and the vehicle speed is below a predetermined threshold, such as a few km / h, to filter out small oscillations and only consider approaching target vehicles. This threshold will be an updated parameter, which may eventually take into account the relative distance of the targets to assimilate it into a following time threshold; that is, filtering is applied by considering the target vehicle VP based on its speed and the following time adjustment made by the driver, so that the target vehicle VP is considered only if the time between VP and EGO is below a predetermined threshold. Stiffness coefficient k P The value of is preferably in the range of [0; 10]. It is known that the higher it is, the faster the braking. Therefore, its adjustment can be a function of the mode selected by the driver, for example, the value in Sport mode is higher than the value in City mode. Furthermore, to account for the case where there is no target vehicle VP, the index F of the target vehicle VP is used as before. Pabs When this indicator has a value of 1, there is no vehicle ahead of the target; when this indicator has a value of 0, there is a vehicle ahead of the target. A default value is used so that if the vehicle ahead of the target disappears: X G =X0:

[0093] [Math.6]

[0094]

[0095]

[0096]

[0097] X G Preferably, the velocity and acceleration magnitude are calculated and then differentiated to obtain other velocities and acceleration magnitudes. The method can also consider four (or more) targets, such as... Figure 6 As shown: V1, V2, V0, and VP, with a default value X0, are equal to the following distance setpoint d_s_c. Equal to the control speed selected by the driver, take Equal to 0, and similarly, in the absence of a lateral target Vi, take X. i , The value equals 0, where NbTm is the number of vanished lateral targets, and F is the value when there are no vehicles in front of the target. Pabs =1, therefore, when the lateral target vehicle disappears, we obtain X.G =X0:

[0098] [Math.7]

[0099]

[0100]

[0101]

[0102] X G Preferably, the velocity and acceleration magnitude are calculated and then differentiated to obtain other velocities and acceleration magnitudes. If the sensing sensors allow, more targets can be considered, especially lateral targets.

[0103] Because of this method, the ACC system will exhibit more comfortable behavior when the target changes, and safety is improved because the movement of the target vehicle is anticipated through this method. Therefore, the vehicle's EGO behavior will closely resemble that of a human driver and will be smoother in traffic.

Claims

1. A driver assistance method for a ego vehicle (EGO) traveling on a traffic lane, characterized in that, The driver assistance method comprises: - a first step of identifying the surrounding traffic of the host vehicle on its lane and on at least one adjacent parallel lane in the same traffic direction; - a second step of determining a virtual barycenter target (G) of target vehicles included in said surrounding traffic, wherein the position of the virtual barycenter target is calculated as a barycenter of the positions of the target vehicles, the speed of the virtual barycenter target is calculated as a barycenter of the speeds of the target vehicles, and the acceleration of the virtual barycenter target is calculated as a barycenter of the accelerations of the target vehicles; - a third step of calculating a longitudinal speed setpoint (Vc), an acceleration setpoint (Ac) and a torque setpoint (Cc) of the host vehicle, said longitudinal speed setpoint (Vc) being a function of the position of the virtual barycenter target (G), the speed of the virtual barycenter target (G) and the acceleration of the virtual barycenter target (G).

2. Driver assistance method for a ego vehicle (EGO) driving on a traffic lane, according to the preceding claim, characterized in that, The surrounding traffic comprises at least two target vehicles (V0, V1, V2, VP) driving in front of the host vehicle (EGO) on its lane or on an adjacent parallel lane in the same traffic direction.

3. The driver assistance method for a ego vehicle (EGO) travelling on a traffic lane as claimed in the preceding claim, characterized in that, The first step comprises a sub-step of detecting each of the at least two target vehicles (V0, V1, V2, VP), wherein for each target vehicle, the position of the target vehicle relative to the host vehicle (EGO), the speed of the target vehicle, the acceleration of the target vehicle are determined as outputs.

4. The driver assistance method for a ego vehicle (EGO) travelling on a traffic lane as claimed in the preceding claim, characterized in that, The first step comprises determining a trajectory of each target vehicle.

5. The driver assistance method for a ego vehicle (EGO) travelling on a traffic lane according to any one of the preceding claims, characterized in that, The second step of determining the virtual barycenter target uses as inputs a pre-selected control speed, a pre-defined following distance setpoint (d_s_c) and the results of the first step.

6. The driver assistance method for a ego vehicle (EGO) travelling on a traffic lane according to any one of the preceding claims, characterized in that, The second step of determining the virtual barycenter target comprises a filtering step.

7. The driver assistance method for a ego vehicle (EGO) travelling on a traffic lane according to any one of claims 2 to 6, characterized in that, At least one of the target vehicles (V1, V2) is on an adjacent lane and tries to merge into the lane of the host vehicle (EGO).

8. The driver assistance method for a ego vehicle (EGO) travelling on a traffic lane according to any one of the preceding claims, characterized in that, The second step of determining the virtual barycenter target uses a target change prediction coefficient determined for each target.

9. The driver assistance method for a ego vehicle (EGO) travelling on a traffic lane as claimed in the preceding claim, characterized in that, The target change prediction coefficient is a function of the relative lateral distance (Y1', Y2') between the trajectory of the host vehicle (EGO) and at least one target vehicle, or a function of the relative lateral distance between the center of the lane on which the host vehicle (EGO) is driving and at least one target vehicle (V1, V2).

10. The driver assistance method for a host vehicle (EGO) traveling on a traffic lane according to claim 8, characterized by, The target change prediction coefficient is a function of the estimated time of intersection of the estimated trajectory of the target vehicle and the estimated trajectory of the host vehicle (EGO).

11. The driver assistance method for a host vehicle (EGO) travelling on a traffic lane as claimed in claim 8, characterized by, The target change prediction coefficient is a function of the following time of a target leading vehicle (VP).

12. The driver assistance method for a host vehicle (EGO) travelling on a traffic lane according to any one of claims 8 to 11, characterized in that, The target variation prediction coefficient is a function of the stiffness coefficient (k, k P ).

13. A driver assistance system (1) for a host motor vehicle (EGO) moving on a lane, characterized in that The driver assistance system comprises: - a module for identifying the surrounding traffic of the host motor vehicle on its lane and on at least one adjacent parallel lane in the same traffic direction; - a module (CBV) for determining a virtual barycenter target of a target vehicle included in the surrounding traffic, wherein a position of the virtual barycenter target, a velocity of the virtual barycenter target and an acceleration of the virtual barycenter target are calculated, the position of the virtual barycenter target corresponding to a barycenter of the position of the target vehicle, the velocity of the virtual barycenter target corresponding to a barycenter of the velocity of the target vehicle, the acceleration of the virtual barycenter target corresponding to a barycenter of the acceleration of the target vehicle; - a module (CD, CV, CC) for calculating a longitudinal velocity setpoint (Vc), an acceleration setpoint (Ac) and a torque setpoint (Cc) of the ego motor vehicle (EGO), the longitudinal velocity setpoint (Vc) being a function of the position of the virtual barycenter target (G), the velocity of the virtual barycenter target (G) and the acceleration of the virtual barycenter target (G).

14. A ego vehicle (EGO) comprising a drive train, and acceleration and braking means, characterised in that, The ego motor vehicle comprises a driver assistance system (1) as claimed in the preceding claim.

Images