WO2016162434A1

WO2016162434A1 - Embedded system on board a motor vehicle for a change of lane functionality, and associated method of control

Info

Publication number: WO2016162434A1
Application number: PCT/EP2016/057646
Authority: WO
Inventors: Caroline Robert; Michel LEEMAN
Original assignee: Valeo Schalter Und Sensoren Gmbh
Priority date: 2015-04-07
Filing date: 2016-04-07
Publication date: 2016-10-13
Also published as: EP3307597A1; FR3034738A1; FR3034738B1

Abstract

The invention relates to a system able to be embedded on board a motor vehicle (1) for an automatic or semi-automatic change of lane functionality. The system according to the invention comprises: to observe a zone ahead of the motor vehicle (1) with a view to detecting an obstacle ahead, at least two and at most three front sets of sensors of different technologies (5₃, 4₁, 4₂, 3₁); to observe a zone behind the motor vehicle (1) with a view to detecting an obstacle behind (V_R), at least two and at most three rear sets of sensors of different technologies (5₄, 4₃, 4₄, 3₂); to observe each lateral zone of the motor vehicle with a view to detecting a lateral obstacle, at most two lateral sets of sensors (4- | -44, 5-i, 52) of different technologies; and a means (8) of control able to generate temporal extrapolations of information delivered by the rear sets and the front sets, and to combine the said temporal extrapolations with information delivered by the lateral set with a view to authorising or otherwise a change of lane of the motor vehicle (1).

Description

On-board vehicle system for lane change functionality, and associated control method

The present invention generally relates to motor vehicles equipped with automatic and semi-automatic systems for driving assistance, and more specifically the systems allowing the automatic or semi-automatic lane change.

The automation of driving advances to meet many issues such as safety, mobility, eco-driving, and accessibility for all to driving. Today, it is possible to have a fully automated vehicle without driver, on dedicated areas. This is not the case for automated road vehicle projects where many problems, particularly in the legal and security fields, remain to be solved before seeing such vehicles on sale. . In particular, in the case of an automated vehicle on the road in the presence of a driver, the Vienna Convention, in Article 8.5, stipulates that the driver must at all times be able to control his vehicle.

The Society of Automotive Engineer (SAE) Automated Road Vehicle Standards Committee (ASEC) recently released a new report providing a classification of automated driving levels, ("Taxonomy and Definitions for Terms Related to On-Road Motor"). Vehicle Automated Driving Systems ", Standard J3016, January 16, 2014), with, for each classification level, the rules for sharing the driving supervision between the driver and the automation system (s). More precisely, this report defines six levels, ranging from level 0 for a vehicle without any automation system, level 5 for a fully automated vehicle, to various degrees of automation for which the part of the vehicle is increased every time. automation and reduces the driver's share of responsibility. So :

Level 1 corresponds to the driver assistance systems relating to either the longitudinal control of the vehicle only (for example the systems known under the acronym ACC or Autonomous Cruise Control), or only lateral control of the vehicle (for example, a system for assisting the maintenance of a lane on a lane, or a lane change aid system used in case of overtaking, overturning after overtaking, or an avoidance procedure );

level 2 corresponds to a so-called "partial automation" level in which the driver assistance system or systems can combine the lateral and longitudinal control of the vehicle;

- level 3 corresponds to a so-called "conditional automation" level in which the driver is allowed, for a certain period of time, and on certain types of road (for example a motorway), not to be attentive to driving . The related automated systems then supervise the lateral and / or longitudinal control of the vehicle, but must make the driver liable in the event of a problem;

- level 4 corresponds to a complete automation of the vehicle, with the possibility for the driver to delegate driving in any situation, and to come back whenever he wants, regardless of the duration and any specific area;

- Level 5 corresponds to a completely automated vehicle, with no possibility for the driver to interact.

The levels 0 to 5 according to the SAE substantially correspond to the levels 0 to 4 of the Federal Agency NHTSA ("National Highway Traffic Safety Administration") in charge of road safety.

Due to the aforementioned Article of the Vienna Convention, level 3 to 5 vehicles are not currently permitted by law. Levels 0 to 2 are allowed, however, because the driver remains the only supervisor of driving.

In the future, we are interested in the projects of future level 3 automated vehicles according to the SAE or NHTSA Standards. The acceptance of a modification of the Vienna Convention, which imposes today a permanent responsibility of the driver, passes at the very least by the implementation of a certain number of procedures of safety of operations and strategies to allow the driver return to driving supervision in the event of system failures. In particular, if a failure is detected, the automated driving system level 3 must certainly give the hand to the driver, but must nevertheless provide functionality for a short period of time, typically of the order of ten seconds , to allow the driver to actually resume control of the driving.

These requirements in terms of dependability necessarily involve using an architecture with a lot of redundancy, which comes to burden the cost of these systems. Thus, in accordance with the international standard ISO 26262 which defines in particular a classification of the criticality of the failures according to four levels known as "ASIL A, ASIL B, ASIL C and ASIL D" (English initials set for Automotive Safety Integrity Level), a Level 3 system should be ASIL D, which in particular involves providing on a vehicle at least three different technologies of sensors to observe the same area of the vehicle environment. This prevents an external jammer come simultaneously inefficient three types of sensors.

FIG. 1 illustrates, by way of nonlimiting example, a known architecture for an automated level 3 system (SAE / NHTSA) for maintaining trajectory on a vehicle lane 1. Functionally, such a system track must be able to control what is happening in front of the vehicle, to control the braking system and / or the engine control system to adjust the speed accordingly, and to control the electronic system. control of the direction to stay on the same way of driving. In the known architecture shown in FIG. 1, the automated level 3 system (SAE / NHTSA) comprises:

- A front camera 2, for example in the middle of the front windshield, and the associated observation area A ₂ has an angle of about 40 ° over a range of about 100 meters;

a laser sensor 3, located for example in the middle of the front bumper, and whose associated observation area A ₃ has an angle of about 145 ° over a range of about 150 meters; two radar sensors 4 ₁ and 4 ₂ placed for example on the left and right front fenders of the vehicle 1, with a range of the order of a hundred meters.

These last two radar sensors 4i and 4 _Σ are generally part of a more complete system, also equipped with similar radar sensors 4 ₃ and 4 ₄ placed on the left and right rear wings of the vehicle 1, all of these four radar sensors. covering an observation area A ₄ surrounding the vehicle 1.

Four 5-i, 5 _Σ , 53 and 5 ₄ cameras, respectively preferably located at the left and right mirrors, the front bumper and the rear bumper, cover an observation area A ₅ also surrounding the vehicle. 1, used in particular also to give the driver a 360 ° representation of the vehicle environment.

The system of FIG. 1 may also comprise means making it possible to control that the driver is in a good state to take back the hand if the situation requires it, for example a camera 6 pointed at the driver (observation zone A ₆ ) and sensors (not shown) for detecting that the hands of the driver are on the steering wheel (systems called "Hands-on" in English terminology).

Optionally, the system also uses the data of a navigation receiver 7, for example of the GPS type, so as to anticipate, for example, motorway changes and therefore the changes of lines that would be necessary.

Other architectures exist for level 3 automated systems (SAE / NHTSA) for maintaining trajectory on a lane of a vehicle.

Such an architecture satisfies the requirements of Level 3 (SAE / NHTSA) for the functionality of trajectory maintenance on a lane since the front of the vehicle 1 is well observed on three partially overlapping zones (A ₂ , A ₃ and A). ₄ ), using three different technologies, as prescribed for ASIL level D. It is more difficult to achieve at least cost a function of automated lane change ("full automatic lane change" in English terminology) in level 3. Indeed, in this case, the system must, in addition to the features described below. above for maintenance in a lane, detecting, in the context of overtaking without driver intervention, that it is possible to change lanes by analyzing in particular the rear and sides of the vehicle, activate the direction indicator to warn third parties of an imminent lane change, order the electronic steering control system for lane change. Once exceeded, the system must also fully supervise the procedure of folding the vehicle.

Some solutions proposed so far are to use three observation zones surrounding the vehicle 360 °, covered by three different technologies of sensors. With reference to the architecture shown in FIG. 1, it is possible in particular to use observation zones A ₄ and A ₅ related to radar and camera technologies. For the third 360 ° observation zone, related to the laser technology, at least three additional laser sensors are required with respect to the laser sensor 3 of FIG. 1, placed in particular on the sides and at the rear of the vehicle. These solutions are thus very expensive.

The invention aims in particular to provide an affordable architecture for a completely automated level 3, or semi-automated level 2, lane change system.

To do this, the subject of the present invention is a system that can be embarked on a motor vehicle for an automatic or semi-automatic lane change function, the system comprising:

to observe a front zone of the motor vehicle for the purpose of detecting a front obstacle, at least two and at most three front assemblies of sensors of different technologies, of which a first front assembly using at least one sensor of a first technology, a second together before using at least one sensor of a second technology, and, where optionally, a third front assembly comprising a single sensor of a third technology;

to observe a rear area of the motor vehicle for the purpose of detecting a rear obstacle, at least two and at most three rear sets of sensors of different technologies, of which a first rear assembly using at least one sensor of a first technology, a second rear assembly using at least one sensor of a second technology, and, where appropriate, a third rear assembly comprising a single sensor of a third technology;

to observe each lateral zone of the motor vehicle with a view to detecting a lateral obstacle, at most two lateral assemblies of sensors of different technologies; and

control means capable of generating temporal extrapolations of information delivered by the rear assemblies and the front assemblies, and combining said temporal extrapolations with information supplied by the lateral assemblies with a view to authorizing or not allowing a change of direction of the motor vehicle.

According to other particularly advantageous characteristic features, taken alone or in combination:

the information delivered by the rear assemblies, the front assemblies and the lateral assemblies comprise information relating to a front, rear or lateral obstacle detected in the front zone, rear zone or in each lateral zone;

the information relating to a front, rear or lateral obstacle comprises a speed, an acceleration and a position of said obstacle, in other words, all the information necessary for establishing a trajectory followed by the obstacle;

the control means combine the time extrapolations with the information delivered by the lateral sets in a weighted linear combination. Thus, it is possible to associate in particular a higher index of confidence with the information coming from the lateral assemblies, than with extrapolated information coming from the front or rear assemblies; a decreasing weighting coefficient over time can advantageously be assigned to time extrapolations;

- The sensor of the third front assembly, respectively rear, is adapted to be arranged for example on a front bumper, respectively rear of the motor vehicle;

said first front and rear assemblies each comprise a single sensor of first technology capable of being arranged for example on a front bumper, respectively rear of the motor vehicle;

said second front and rear assemblies each comprise two sensors of second technology, able to be arranged for example on the front and respectively rear lateral wings of the motor vehicle;

the first technology sensors are preferably image sensors;

the second technology sensors are preferably radar sensors;

the third technology sensors are preferably laser sensors;

in a possible implementation, the system comprises exactly three front and rear sets of sensors of different technologies, and two lateral sets of sensors of different technologies for each lateral zone of the vehicle. This implementation makes it possible to guarantee, at lower cost, a fully automatic level 3 lane change functionality, with an ASIL D safety level;

in another implementation, the system comprises exactly two front and rear sets of sensors of different technologies, and two lateral sets of sensors of different technologies for each lateral zone of the vehicle. This implementation makes it possible to guarantee, at a lower cost, a level 2 semi-automatic lane change functionality.

The invention also relates to a method for controlling a system that can be loaded onto a motor vehicle for a function of automatic or semi-automatic lane change, characterized in that, the system comprising:

to observe a front zone of the motor vehicle for the purpose of detecting a front obstacle, at least two and at most three front assemblies of sensors of different technologies, of which a first front assembly using at least one sensor of a first technology, a second a front assembly using at least one sensor of a second technology, and a third front assembly comprising a single sensor of a third technology;

to observe a rear area of the motor vehicle for the purpose of detecting a rear obstacle, at least two and at most three rear sets of sensors of different technologies, of which a first rear assembly using at least one sensor of a first technology, a second rear assembly using at least one sensor of a second technology, and a third rear assembly comprising a single sensor of a third technology; and

to observe each lateral zone of the motor vehicle with a view to detecting a lateral obstacle, at most two lateral assemblies of sensors of different technologies;

the method comprises a step of generating time extrapolations of information delivered by the rear assemblies and the front sets, and a step of combining said time extrapolations with information provided by the side sets with a view to authorizing or not allowing a change of track of the motor vehicle.

The invention and the various advantages that it provides will be better understood from the following description, made with reference to the appended figures, in which:

FIG. 1, already described above, schematically illustrates an example of known architecture for a trajectory maintenance system on a fully automated level 3 track (SAE / NHTSA) fitted to a motor vehicle; - Figure 2 schematically illustrates an example of architecture according to the invention for a fully automated level 3 lane change system (SAE / NHTSA) fitted to a motor vehicle;

FIG. 3 schematically represents an exemplary simulation of a vehicle equipped with the lane change system of FIG.

2, driving on a stretch of highway.

In the rest of the description, and unless otherwise stated, the elements common to all the figures bear the same references.

Throughout the briefing, "Automatic lane change functionality" is known as Level 3 (SAE / NHTSA) functionality, and "Semi-automatic lane change functionality" (Tier 2) functionality (SAE / NHTSA ).

To achieve an automatic lane change functionality, the invention proposes here to complete the architecture of FIG. 1 by a single laser sensor placed at the rear of the vehicle, for example at the level in the middle of the rear bumper. The architecture thus obtained is shown schematically in FIG. 2, similar to FIG. 1, except that:

the laser sensor placed at the front of the vehicle 1 here bears the reference number ₃ with an associated front observation zone A ₃ i;

a laser sensor 3 ₂ similar to the front laser sensor 3-i, with an associated observation zone A ₃₂ , has been added to the rear of the vehicle 1. Such an architecture is in accordance with the ASIL D level:

at the front of the vehicle since it offers three observation zones (portion of A ₄ , portion of A ₅ and A ₃ i) covered by three different sensor technologies; and

- At the rear of the vehicle since it offers three observation areas (portion of A ₄ , portion of A ₅ and A ₃₂ ) covered by three different sensor technology. Note that the sensors used here for the front and rear of the vehicle can detect obstacles located at a minimum distance, depending on the speed of use of the lane change feature, and therefore do not include the ultrasonic sensors of short range used for example for parking assistance.

Nevertheless, this architecture is not sufficient on its own to guarantee the ASIL D level since only two left and right lateral observation zones of the vehicle, covered by two different sensor technologies (left and right lateral portions of A ₄ and A ₅ ), are present.

To overcome this shortcoming, the invention starts from the observation that, in most cases encountered on the highway, an obstacle, in particular another vehicle at the rear, respectively at the front of the vehicle concerned, and whose proximity would prohibit lane change, only appear in a lateral observation zone after first appearing in the rear observation areas, respectively before.

This is particularly visible in FIG. 3 which illustrates an exemplary simulation of the vehicle 1 of FIG. 2, traveling on a three-lane highway Li, L ₂ , L ₃ in the direction indicated by an arrow F. The vehicle 1 moves on the rightmost Li lane, following another vehicle V _F. Figure 3 also shows that a vehicle V _R rolls to the rear of the vehicle 1, in the middle lane. In order not to overload FIG. 3, only the important observation areas used in an automatic override procedure, in particular the zones A ₂ , A ₃ , A ₃₂ , A ₄ and A ₅ , have been reproduced. Given the width of each path Li, L ₂ , L ₃ , typically of the order of 3.50 meters in France, and the sizes of the different observation zones of the system equipping the vehicle 1, the system is able to detecting, in level ASIL D, the presence of the vehicle V _F at the front of the vehicle 1, and the presence of the vehicle V _R at the rear of the vehicle 1. In particular, the system is able to:

- decide that a change of lane to overtake the vehicle V _F is necessary if the VF vehicle moves at a higher speed weaker than that of the vehicle 1 and / or that the safety distance between the two vehicles 1 and VF is about to be no longer respected;

- prevent, or at least postpone, this lane change if the vehicle V _R at the rear is too close and / or ready to overtake the vehicle 1.

To do this, and thus guarantee an ASIL D level also on the vehicle sides 1, the present invention proposes to use the information provided by the rear assemblies and before assemblies, to extrapolate this information temporally, then to combine the time extrapolations. with the information provided by the side assemblies to allow or not a lane change of the motor vehicle.

In other words, the triple redundant rear observation zones will be used to detect any vehicle VR ready to overtake the vehicle 1, and prohibit disengagement if necessary, but also, in accordance with the invention, to extrapolate the information related to this vehicle V _R in the lateral zone, here left, of the vehicle 1, and postpone the lane change decision until the vehicle V _R has not finished passing. Similarly, when the vehicle 1 will be passed on the lane L ₂ in order to pass the vehicle V _F , the triple redundancy front observation zones will be used in accordance with the invention to extrapolate the information related to this vehicle. vehicle V _F in the lateral zone, right here, of the vehicle 1, and thus manage the return on the track Li of the vehicle 1 after passing the vehicle V _F.

In all cases, the information relating to a front, rear or lateral obstacle comprises a speed, an acceleration and a position of said obstacle, in other words, all the information necessary to restore a trajectory followed by the obstacle.

System control means, schematically represented as 8 in FIG. 2, combine the time extrapolations with the information provided by the side sets, preferably in a weighted linear combination. We can thus associate an index of higher confidence in information from side sets, than extrapolated information from front or rear sets.

It is also possible to provide a weighting coefficient assigned to temporal extrapolations that decreases over time.

In other words, the invention proposes a cost-optimized architecture (triple redundancy only on the front and rear areas) which makes it possible to offer a completely automated lane-change functionality that meets the safety requirements of the ASIL D level. .

Moreover, although the sensors have been described in the diagrams as image sensors for sensors 5 _3, 5 _4, the sensors under the radar technology for 4- _\ sensors 4 ₄ and falling sensor laser technology for 3i sensors and _Σ 3, the principles of the invention are applicable regardless of the combination of technologies (or types) different use.

Finally, the invention also finds an interesting application and also at a lower cost for a level 2 semi-automatic lane-changing functionality. In this case, only two sets of sensors of different technologies are necessary on the one hand for 'before, and secondly at the back, and only one sensor is needed to observe each side area. According to the principle of the invention, temporal extrapolations of information delivered by the rear assemblies and the front assemblies are realized, then combined with information supplied by the lateral assemblies in order to authorize or not a change of lane of the motor vehicle. . In this particular case, however, the authorization is only a proposition of the system made to the driver to change lanes, since the semi-automatic functionality implies a supervision of the driver.

Claims

A vehicle-mountable system (1) for automatic or semi-automatic lane-change functionality, the system comprising:

to observe a front zone of the motor vehicle (1) in order to detect a front obstacle (V _F ), at least two and at most three front assemblies of sensors of different technologies, including a first front assembly using at least one sensor a first technology (5 ₃ ), a second front assembly using at least one sensor of a second technology (4- \, 4 ₂ ), and, if appropriate, a third front assembly comprising a single sensor of a third technology (3i);

to observe a rear area of the motor vehicle (1) for the purpose of detecting a rear obstacle (VR), at least two and at most three rear assemblies of sensors of different technologies, including a first rear assembly using at least one sensor; a first technology (5 ₄ ), a second rear assembly using at least one sensor of a second technology (4 ₃ , 4 ₄ ), and, if appropriate, a third rear assembly comprising a single sensor of a third technology ( 3 ₂ );

- to observe each lateral region of the motor vehicle in order to detect a lateral obstacle, at the two lateral sets of sensors (4- | -4 _4, 5-i, 5 ₂₎ of different technologies; and

control means (8) able to generate temporal extrapolations of information delivered by the rear assemblies and the front assemblies, and to combine said temporal extrapolations with information delivered by the lateral sets in order to authorize or not a change track of the motor vehicle (1).

2. System according to claim 1, characterized in that the information delivered by the rear assemblies, the front assemblies and the lateral assemblies, include information relating to a front (VF), rear (VR) OR lateral obstacle detected in the front zone, the rear zone or in each lateral zone respectively.

3. System according to claim 2, characterized in that the information relating to a front (V _F ), rear (V _R ) or lateral obstacle comprise a speed, an acceleration and a position of said obstacle.

4. System according to any one of the preceding claims, characterized in that said control means (8) combine the time extrapolations with the information delivered by the side sets according to a weighted linear combination.

5. System according to claim 4, characterized in that a weighting coefficient decreasing with time is assigned to the time extrapolations.

6. System according to any one of the preceding claims, characterized in that the sensor of the third front assembly, respectively rear, is adapted to be disposed on a front bumper respectively rear of the motor vehicle (1).

7. System according to any one of the preceding claims, characterized in that said first front and rear assemblies each comprise a single sensor of first technology (5 ₃ , 5 ₄ ) adapted to be disposed on a front bumper, respectively rear of the motor vehicle (1).

8. System according to any one of the preceding claims, characterized in that said second front and rear assemblies each comprise two second technology sensors (4- | -4 ₄ ), adapted to be arranged on the front and rear side wings respectively. of the motor vehicle (1).

9. System according to any one of the preceding claims, characterized in that the first technology sensors (5 ₃ , 5 ₄ ) are image sensors.

10. System according to any one of the preceding claims, characterized in that the second technology sensors (4i-4 ₄ ) are radar sensors.

11. System according to any one of the preceding claims, characterized in that the third technology sensors (3-i, 3 ₂ ) are laser sensors.

12. System according to any one of the preceding claims, having exactly three sets front, rear, respectively, of different technology sensors, and two lateral sets of sensors (4- | -4 _4, 5-i, 5 ₂₎ Technology different for each side area of the vehicle (1).

13. System according to any one of claims 1 to 11, having exactly two sets before, respectively rear, of sensors of different technologies, and two side sets of sensors (4- | -4 ₄ , 5i, 5 ₂ ) technologies. different for each side area of the vehicle (1).

14. A method for controlling a system suitable for being fitted to a motor vehicle (1) for an automatic or semi-automatic lane change feature, characterized in that the system comprising:

to observe a front zone of the motor vehicle (1) in order to detect a front obstacle (V _F ), at least two and at most three front assemblies of sensors of different technologies, including a first front assembly using at least one sensor a first technology (53), a second front assembly using at least one sensor of a second technology (4- \, 4 ₂ ), and, if appropriate, a third front assembly comprising a single sensor of a third technology (3i);

to observe a rear area of the motor vehicle (1) for the purpose of detecting a rear obstacle (V _R ), at least two and at most three rear assemblies of sensors of different technologies, of which one first rear assembly using at least one sensor of a first technology (5 ₄ ), a second rear assembly using at least one sensor of a second technology (4 ₃ , 4 ₄ ), and, if necessary, a third rear assembly comprising a single technology of a third sensor (3 _2); and

- to observe each side area of the motor vehicle for the purpose of detecting a lateral obstacle, at most two lateral sets of sensors (41-44, 51, 52) of different technologies;

the method comprises a step of generating time extrapolations of information delivered by the rear assemblies and the front sets, and a step of combining said time extrapolations with information provided by the side sets with a view to authorizing or not allowing a change of track of the motor vehicle (1).