METHOD AND DEVICE FOR MAINTAINING THE SELECTION OF A TARGET VEHICLE BY AN ON-BOARD DRIVING ASSISTANCE SYSTEM

Description

[0001] The invention is in the field of autonomous vehicle driver assistance systems. In particular, the invention relates to maintaining the selection of a target vehicle by a driver assistance system embedded in an autonomous vehicle when said autonomous vehicle is moving laterally.

[0002] The "left", "right", "front" and "rear" of a vehicle are determined in relation to the normal direction of travel of the vehicle.

[0003] The term "vehicle" refers to any type of vehicle, such as a car, moped, motorcycle, warehouse robot, etc. "Autonomous driving" of an "autonomous vehicle" refers to any process capable of assisting the driving of the vehicle. This process may consist of partially or fully controlling the vehicle or providing any type of assistance to a person driving the vehicle. The process thus covers all autonomous driving, from level 0 to level 5 in the OICA (International Organization of Motor Vehicle Manufacturers) scale.

[0004] The systems designed to assist vehicle driving are also called ADAS (Advanced Driver Assistance Systems), ADAS systems, or driver assistance systems. Currently, a commercially available vehicle includes several ADAS systems, such as at least cruise control and lane keeping assist.

[0005] A cruise control system on a personal electric vehicle (PEV), particularly so-called intelligent cruise control systems also known as ACC (Auto Cruise Control), is capable of regulating the PEV's speed based on the relative distance between the PEV and a target vehicle traveling ahead. This regulation also depends on instructions from the PEV's driver, such as a speed and a time interval (distance to the target vehicle, which depends on the PEV's instantaneous speed). A typical measurement of the relative distance is obtained using LiDAR. When a target vehicle is detected, it is selected by the ACC to adjust the PEV's speed. For example, on the motorway the driver set a speed limit of 130 km / h and an inter-vehicle time of 2 seconds.If no target vehicle is detected, the cruise control maintains the set speed of 130 km / h. When approaching a slowdown, upon detection of a target vehicle traveling, for example, at 60 km / h, the cruise control will reduce the ego-vehicle's speed to 60 km / h while maintaining a 2-second travel time between the ego-vehicle and the target vehicle, as long as the target vehicle is selected. If the target vehicle is lost from detection by the LiDAR, for example, when the target vehicle moves several meters laterally away from the ego-vehicle, and no other target vehicle is detected, the target vehicle's speed is no longer selected and the set speed returns to 130 km / h.

[0006] A lane keeping assist system, also known as LKA (from the acronym "Lane Keeping Assist"), is capable of keeping a vehicle in its lane if it detects an unintentional lane change. The intention is detected, for example, by the activation of a turn signal on the side of the detected lane change. A vehicle with LKA includes a front-facing camera capable of capturing images in front of the vehicle and processing these images. An LKA system can also detect a lane change by the vehicle by comparing an estimated trajectory of the road with an estimated trajectory of the vehicle. Upon detecting an unintentional lane change, the LKA system keeps the vehicle in its lane by modifying the vehicle's trajectory, for example, by applying torque or force to the steering system.A driver remains able to compensate for the cut or effort generated by the LKA system, for example if the driver wishes to travel on the same lane but laterally offset from a lane center in order to, for example, avoid an obstacle, let a motorcycle pass.

[0007] Currently, combining an ACC system with a LKA system creates problems. When the LKA system detects a lane change, the ACC system is notified and loses focus on the target vehicle; the target vehicle selection is no longer maintained. In the previous example, the speed setting then changes from 60 km / h to 130 km / h. If the driver remains in the lane, the target vehicle is selected again, and the speed setting drops from 130 km / h to 60 km / h. Finding a compromise in the LKA and ACC system settings is therefore difficult. Indeed, if a highly responsive and predictive lane change detection system is desired by the LKA system, then the ACC must be set to be less responsive, and vice versa.

[0008] In addition, the state of the art is known from documents DE102019203610A1, US2017341647A1, US2013085976A1 and the publication "Lane change intent prediction for driver assistance: On-road design and evaluation", Intelligent Vehicles Symposium (IV), 2011 IEEE, June 5, 2011, pp 895-901.

[0009] One object of the present invention is to remedy the aforementioned problem, in particular to avoid oscillations between the set speed of the speed regulator and to have a good compromise of adjustment between a speed regulation system and a lane keeping system.

[0010] To this end, a first aspect of the invention relates to a method for maintaining the selection of a target vehicle by a driver assistance system embedded in a self-driving vehicle when said self-driving vehicle moves laterally on a lane called the self-driving lane, said driver assistance system comprising a camera capable of detecting said lateral movement, said driver assistance system further comprising a cruise control capable of regulating the speed of the self-driving vehicle according to a relative distance between said self-driving vehicle and said target vehicle, said target vehicle traveling on said self-driving lane in front of said self-driving vehicle, said method comprising the steps of: Detection, by the camera, of said lateral displacement of the ego-vehicle and determination of a lateral side of said lateral displacement of the ego-vehicle, said determined side is said to be the displacement side, a side opposite said displacement side is said to be the opposite side; Detection, by a sensor based on a measurement of a reflection of a wave emitted by said sensor, of an object located to the left or right of said ego-vehicle, said detection determines a probability of presence, determines a position relative to said ego-vehicle, and determines a side, displacement side or opposite side, where said detected object is located relative to the ego-vehicle; Determination of a relative lateral distance between said ego-vehicle and said detected object;If the probability of presence is greater than a predetermined threshold, then if the detected object is on the moving side, as long as the relative lateral distance is greater than a first threshold, the selection of the target vehicle is maintained, and if the detected object is on the opposite side, as long as the relative lateral distance is less than a second threshold, the selection of the target vehicle is maintained, the first and / or second threshold being a function of a determination of the width of the lane.

[0011] Thus, the absence of an intention to change lanes is clearly detected. The cruise control continues to regulate the speed of the e-vehicle based on its relative distance to the target vehicle. The speed variation of the vehicle remains roughly equal to the speed variation of the target vehicle. This speed variation is therefore easily understood by the occupants of the e-vehicle and is thus comfortable. There is no abrupt change in the regulated speed, especially if the speed of the target vehicle is significantly different, by 10% or more, from a target speed previously set by an occupant of the e-vehicle. Furthermore, the system settings are simplified.

[0012] Advantageously, the detection of said lateral displacement of the ego-vehicle is a function of a parametric identification of a curve of a line of the ego-lane.

[0013] Advantageously, the detection of an object and the determination of the relative lateral distance are determined from a processing of information from at least one radar on board said ego-vehicle.

[0014] Advantageously, the detection of an object and the determination of the relative lateral distance are determined from a processing of information from at least one ultrasonic sensor on board said ego-vehicle.

[0015] Advantageously, the first threshold and / or the second threshold is a function of the detected object.

[0016] Thus, the absence of an intention to change lanes is clearly detected and is more reliable. Indeed, the detected object can be fixed, such as a guardrail at the side of the road, or moving, such as a vehicle traveling alongside the ego-vehicle.

[0017] Thus, the absence of an intention to change lanes is clearly detected and is more reliable and is a function, for example, of a type of road (such as a departmental road, a motorway, ...).

[0018] A second aspect of the invention relates to a device comprising a memory associated with at least one processor configured to implement the method according to the first aspect of the invention.

[0019] The invention also relates to a vehicle incorporating the device.

[0020] The invention also relates to a computer program comprising instructions adapted for the execution of the steps of the process, according to the first aspect of the invention, when said program is executed by at least one processor.

[0021] Other features and advantages of the invention will become apparent from the description of the non-limiting embodiments of the invention below, with reference to the accompanying figures, in which: [ Fig. 1 [ ] schematically illustrates a device, according to a particular embodiment of the present invention. ] Fig. 2 ] schematically illustrates a method of maintaining a selection of a target vehicle by a driver assistance system embedded in an ego-vehicle, according to a particular embodiment of the present invention.

[0022] The invention is described below in its non-limiting application to the case of an autonomous motor vehicle traveling on a road or traffic lane. Other applications, such as a robot in a warehouse or a motorcycle on a country road, are also conceivable.

[0023] There figure 1 represents an example of a device 101 included in the vehicle, in a network ("cloud") or in a server. This device 101 can be used as a centralized device responsible for at least some steps of the process described below with reference to the figure 2 In one embodiment, it corresponds to an autonomous driving computer.

[0024] In the present invention, the device 101 is included in the vehicle.

[0025] This 101 device can take the form of a case containing printed circuits, any type of computer or even a mobile phone (“smartphone”).

[0026] The device 101 includes a random access memory 102 for storing instructions for the implementation by a processor 103 of at least one step of the process as described above. The device also includes a mass storage 104 for storing data intended to be retained after the implementation of the process.

[0027] Device 101 may further include a digital signal processor (DSP) 105. This DSP 105 receives data to shape, demodulate and amplify, in a manner known per se, this data.

[0028] Device 101 also includes an input interface 106 for receiving data implemented by the method according to the invention and an output interface 107 for transmitting data implemented by the method according to the invention.

[0029] There figure 2This schematically illustrates a method for maintaining the selection of a target vehicle by a driver assistance system embedded in a self-driving vehicle, according to a particular embodiment of the present invention. In this example, said self-driving vehicle moves laterally on a lane called the self-driving lane, said driver assistance system comprising a camera capable of detecting said lateral movement, said driver assistance system further comprising a cruise control capable of regulating the speed of the self-driving vehicle according to a relative distance between said self-driving vehicle and said target vehicle, and said target vehicle traveling on said self-driving lane in front of said self-driving vehicle.

[0030] Step 201, RefChg, is a detection step, by the camera, of said lateral displacement of the ego-vehicle and determination of a lateral side of said lateral displacement of the ego-vehicle, said determined side is said displacement side, a side opposite said displacement side is said opposite side.

[0031] Typically, the onboard camera is mounted at the top of the windshield of the electric vehicle and captures images of the front of the vehicle. Through image processing, this device is capable of modeling, via parametric identification, in real time: A trajectory of the ego-vehicle over several tens of meters; A trajectory of a line, for example a white line marked on the ground indicating a lateral boundary of the ego-lane; A trajectory of the ego-lane on which the ego-vehicle travels. This trajectory is determined from the trajectories of lines on the lane; A positioning of the ego-vehicle relative to the center of the ego-lane.

[0032] Advantageously, the detection of this lateral movement of the ego-vehicle depends on a parametric identification of a curve within a line of the ego-lane. This involves comparing the modeled trajectory of the ego-vehicle to the trajectory of a line of the ego-lane. Depending on how this difference changes, the device is able to detect a lateral movement of the ego-vehicle.

[0033] Step 202, Obj?, is a detection step, by a sensor based on a measurement of a reflection of a wave emitted by said sensor, of an object located to the left or right of said ego-vehicle, said detection determines a probability of presence, determines a position with respect to said ego-vehicle, and determines a side, displacement side or opposite side, where said detected object is located with respect to the ego-vehicle.

[0034] Advantageously, the detection of an object is determined from the processing of information from at least one radar on board said ego-vehicle and / or at least one ultrasonic sensor on board said ego-vehicle.

[0035] In a preferred operating mode, ultrasonic sensors used for parking maneuvers detect the presence of an object (vehicle, railing, etc.) located laterally, to the left or right, relative to the vehicle. Typically, these sensors are distributed around the vehicle. A device associated with these sensors is capable of measuring the distance of the detected object from a sensor that detects the object.

[0036] Advantageously, the probability of presence is conventionally calculated using detection redundancy (local and / or temporal redundancy: several sensors detect the same object and / or an object is detected for several instants). Advantageously, the position of an object relative to the ego-vehicle is determined by a time-of-flight calculation of wave reflections emitted by the sensor.

[0037] Step 203, CalcDist, is a step for determining a relative lateral distance between said ego-vehicle and said detected object.

[0038] Advantageously, the determination of the relative lateral distance is based on the processing of information from at least one radar unit mounted on said ego-vehicle and / or at least one ultrasonic sensor mounted on said ego-vehicle. Advantageously, the distance of an object from said ego-vehicle is determined by a calculation based on the time of flight of reflections of waves emitted by the sensor and the position of said sensor relative to a reference point on said ego-vehicle.

[0039] Step 204, TestDist?, is a test step. The test is positive if the probability of presence is greater than a predetermined threshold, if the relative lateral distance is greater than a first threshold (204) for an object detected on the direction of movement, and if the relative lateral distance is less than a second threshold (204) for an object detected on the opposite side. In the case of a positive test, the process proceeds to step 204 described below. Conversely, if the test is negative, the process proceeds to step 205 described below.

[0040] The test is negative if an object has been detected and the probability of presence is less than a threshold, for example 70%.

[0041] The test is also negative if an object is detected on the travel side and the relative lateral distance is less than or equal to a first threshold. This means the vehicle is approaching an object on the travel side in a significant way. The driver is clearly indicating their intention to change lanes.

[0042] The test is also negative if an object is detected on the opposite side of the movement and if the relative lateral distance is greater than or equal to a second threshold. This means that the e-vehicle is moving noticeably away from an object on the opposite side. The driver is clearly indicating their intention to change lanes. The first and second thresholds are distances in meters, ideally around 1 meter. The first and / or second thresholds are based on a determined e-vehicle lane width. These thresholds can also take other values ​​between 0 and 10 meters and / or be based on the e-vehicle width, and / or be based on the detected object.

[0043] Stage 205, MaintACC, is a stage where a driver assistance system in a self-driving vehicle maintains its selection of a target vehicle when the vehicle moves laterally within a designated lane. Reaching this stage indicates that the driver did not intentionally change lanes and wishes to remain within the lane. The system continues to monitor the target vehicle's speed. Therefore, the speed control remains dependent on the target vehicle, even if the self-driving vehicle makes small lateral movements. A small lateral movement corresponds to a steering angle of a few degrees, such as approximately 2° if the vehicle is traveling at 130 km / h. At lower speeds, this can be several tens of degrees of steering angle.

[0044] Step 206, DesACC, is a step to deselect the target vehicle even if the vehicle is visible to the LiDAR. This is because the driver has clearly indicated their intention to change lanes. Therefore, there is no longer any need to restrict the vehicle's speed relative to the target vehicle, which will no longer be in the new lane.

[0045] The present invention is not limited to the embodiments described above by way of example; it extends to other variants, within the limits of the scope of protection defined by the claims.

[0046] Thus, we described above an example of an implementation in which we determined the driver's lack of intention to change lanes. We have made the lane change detection by an onboard camera more robust to simplify and reduce the dependence on parameter settings of an ACC system coupled with a LKA-type system or a type of steering input on the ego-vehicle. Further improvements in robustness are possible by taking into account other measured or estimated values ​​representing a driver's intention to change lanes, such as turn signal activation, steering wheel and / or steering shaft rotation angle, steering rack movement, pneumatic forces or torque, longitudinal and lateral acceleration, yaw rate and / or acceleration, etc.

Claims

1. Method for maintaining a selection of a target vehicle by a driving aid system on board an eco-vehicle when said eco-vehicle is moving laterally on a track called eco-track, said driving aid system comprising a camera capable of detecting said lateral movement, said driving aid system further comprising a speed regulator capable of regulating the speed of the eco-vehicle as a function of a relative distance between said eco-vehicle and said target vehicle, said target vehicle travelling on said eco-track in front of said eco-vehicle, said method comprising the steps of: • Detection (201), by the camera, of said lateral displacement of the ego-vehicle and determination of a lateral side of said lateral displacement of the ego-vehicle, said determined side is said displacement side, a side opposite said displacement side is said opposite side; • Detection (202), by a sensor based on a measurement of a wave reflection emitted by said sensor, of an object located to the left or right by report to said ego-vehicle, said detection determines a probability of presence, determines a position by report to said ego-vehicle, and determines a side, side of movement or opposite side, where said object detected by report to the ego-vehicle is located; • Determination (203) of a relative lateral distance between said eco-vehicle and said detected object; • If said probability of presence is greater than a predetermined threshold, then if the detected object is of the side displacement, as long as said relative lateral distance is greater than a first threshold (204) then the selection of said target vehicle is retained (205), and if the detected object is of the opposite side, as long as said relative lateral distance is less than a second threshold (204) then the selection of the target vehicle is retained (205), the first threshold and / or the second threshold being a function of a determination of the width of the ego-track.

2. Method according to claim 1, wherein the detection of said lateral displacement of the ego-track is a function of a parametric identification of a curve of a line of the ego-track.

3. Method as claimed in claim 1, wherein the detection of an object and the determination of the relative lateral distance is determined from a processing of information coming from at least one radar on board said eco-vehicle.

4. Method as claimed in claim 1, wherein the detection of an object and the determination of the relative lateral distance is determined from a processing of information coming from at least one ultrasound sensor on board said eco-vehicle.

5. Method according to claim 1, wherein the first threshold and / or the second threshold is a function of the detected object.

6. Device (101) comprising a memory (102) associated with at least one processor (103) configured to implement the method according to one of the previous claims.

7. Vehicle comprising the device according to the previous claim.

8. Computer plan comprising instructions adapted for the execution of the steps of the method according to one of claims 1 to 5 when said plan is executed by at least one processor (103).