FR3075136B1 - METHOD AND DEVICE FOR ASSISTING DRIVING A PARTIALLY AUTOMATED DRIVING VEHICLE BY COMPARING GLOBAL STATES - Google Patents

Abstract

A method assists driving a driver of a partially self-driving vehicle, acquiring information representing the environment, and having a current dynamic defined by magnitudes and a current light state defined by operating states of functions photometric. This method comprises: a first step (10) in which this information is analyzed in order to estimate a state of the environment; a second step (20) in which an overall reference state is determined which the vehicle should exhibit in presence of the magnitudes and state of the environment, and a current global state of the vehicle depending on the magnitudes, state of the environment and operating states, and - a third step (30-40) where a level of difference between these current and reference global states, and the driver is alerted and / or at least partial control of the driving at this level of difference.

Description

ASSISTANCE METHOD AND DEVICE FOR DRIVING A PARTIALLY AUTOMATED DRIVING VEHICLE, BY COMPARISON OF GLOBAL STATES

The invention relates to vehicles with partially automated (or autonomous) driving, and more specifically the assistance of drivers in driving such vehicles.

Here, the term "partially automated (or autonomous) driving vehicle" means a vehicle which can be driven manually by its driver during manual driving phases and automatically (without any intervention from the driver) during autonomous driving phases. which may be limited to the sole result of the invention or be the result of a specific functionality.

Driving a vehicle can be a complex task as it takes place in an environment that is almost constantly changing. This environment is made up of traffic lanes with variable traffic characteristics and rules, in the presence of other mobile or temporarily fixed vehicles, pedestrians, animals, objects possibly placed in abnormal places, and conditions potentially variable weather.

Because of this complexity, it sometimes happens that the driving of a vehicle by its driver is not optimally adapted, or even unsuitable, to the environment in which the vehicle operates. This may for example be the result of inexperience and / or fatigue and / or inattention and / or incapacity and / or discomfort and / or a second state.

Some vehicles offer driving assistance options, such as speed and / or distance control compared to the previous vehicle, or speed regulation according to the speed limit, or assistance positioning in the lane, or control of the alertness level and / or the driver's health and / or mood. But these assistances each act only on a part of the driving of the vehicle or on the driver according to a very limited part of the external or internal environment, and not in a global way. Consequently, these assistances do not make it possible to detect whether the driver's driving is unsuitable for the environment, and were not designed to alert the driver or take complete control of the vehicle in place of the driver when driving this the latter is unsuitable.

One of the object of the invention is to improve the situation.

To this end, it notably proposes an assistance method making it possible to assist the driving of a driver of a partially autonomous driving vehicle, moving on a traffic lane, comprising acquisition means acquiring information representative of the environment around the vehicle, and having a current dynamic defined by quantities and a current light state defined by operating states of photometric functions.

This assistance process includes:

- a first step in which we analyze this environmental information to estimate a state of the environment,

- a second step in which, on the one hand, a global reference state which the vehicle should present in the presence of the quantities and of this state of the estimated environment is determined, and, on the other hand, a global state in progress of the vehicle as a function of the sizes, the estimated state of the environment and the operating states, and

a third step in which a level of difference is determined between this overall current state determined and this overall reference state determined, and the driver is alerted to a problem related to his driving of the vehicle and / or control is taken at less partial driving of the vehicle in place of the driver depending on the level of difference determined.

This protects the passengers of the vehicle and other road users and its surroundings against a failure or loss of capacity of the driver.

The assistance method according to the invention may include other characteristics which can be taken separately or in combination, and in particular:

- in its first stage, the state of the environment can be estimated as a function of information acquired and representative of objects detected in the environment of the vehicle and / or the taxiway and / or meteorology and / or exterior brightness;

- in its second stage, it is possible to use at least one quantity chosen from an acceleration, a speed, a steering angle of the steering wheel or of the wheels, a roll angle, a pitch angle, a yaw angle as well as time derivatives of the previous quantities;

- in its second stage, it is possible to use operating states of at least two photometric functions chosen from a stop light function, a main beam function, a low beam function, a fog light function, a function position light, daytime running light function, hazard warning light function, and direction change indicator function;

- in its second step, the global reference state can be determined using at least one mathematical model chosen from a neural network, a Kalman filter, an expert system, fuzzy logic, and a forest of decision trees ;

- in its third stage, it is possible to alert the driver of a problem related to his driving of the vehicle and / or take at least partial control of the driving of the vehicle in place of the driver according to an estimate of a temporal evolution of the level of difference determined from the level of difference just determined and from at least one other level of difference determined previously;

- in its third step, the driver can be alerted of a problem related to his driving of the vehicle and / or take control at least partially of the driving of the vehicle in place of the driver according to an estimate of a level of vigilance of the driver.

The invention also provides an assistance device intended to assist the driving of a driver of a partially autonomous driving vehicle, moving on a traffic lane, comprising acquisition means acquiring information representative of the environment. around the vehicle, and having a current dynamic defined by quantities and a current light state defined by operating states of photometric functions.

This assistance device is characterized by the fact that it includes:

- means of analysis analyzing environmental information to estimate a state of the environment,

processing means determining, on the one hand, an overall reference state that the vehicle should present in the presence of the quantities and this estimated state of the environment, on the other hand, an overall state in progress of the vehicle function of the magnitudes, of the estimated state of the environment and of the operating states, and on the third hand, a level of difference between this global state in progress determined and this global state of reference determined, and

- control means triggering an alert to notify the driver of a problem related to his driving of the vehicle and / or taking at least partial control of the driving of the vehicle in place of the driver according to this determined level of difference.

For example, the control means can trigger the alert and / or take at least partial control of the driving of the vehicle in place of the driver according to an estimate of a level of vigilance of the driver.

The invention also provides a vehicle, possibly of the automobile type, with partially autonomous driving, capable of moving on a traffic lane, having a current dynamic defined by magnitudes and a current light state defined by operating states of photometric functions, and comprising acquisition means acquiring information representative of the environment around the vehicle and an assistance device of the type presented above.

Other characteristics and advantages of the invention will appear on examining the detailed description below, and the attached drawings, in which:

FIG. 1 schematically illustrates, in a top view, an example of a motor vehicle comprising an exemplary embodiment of an assistance device according to the invention, and

FIG. 2 schematically illustrates an example of an algorithm implementing an assistance method according to the invention.

The object of the invention is in particular to propose an assistance method, and an associated assistance device DA, intended to make it possible to assist the driving of a CV driver of a vehicle V with partially automated (or autonomous) driving. .

It is recalled that the term “vehicle with partially automated (or autonomous) driving” is understood here to mean a vehicle which can be driven manually by its driver during manual driving phases and automatically (without any intervention from the driver) during autonomous driving phases (which may, if necessary, be limited to the sole result of the invention).

In what follows, it is considered, by way of nonlimiting example, that the vehicle is of the automobile type. This is for example a car. However, the invention is not limited to this type of vehicle. It concerns in fact any type of vehicle which can be driven in a partially automated (or autonomous) manner and which can circulate on land traffic lanes.

There is diagrammatically illustrated in FIG. 1 a vehicle V comprising acquisition means MAQ, an assistance device DA and a passenger compartment H comprising in particular a driver seat SC on which a driver CV is installed and associated with a steering wheel VV and brake and accelerator pedals (not shown).

The MAQ acquisition means are responsible for acquiring information which is representative of the environment around the vehicle V (that is to say in front of, behind and on the two longitudinal sides). They can, for example, include at least one camera and / or at least one scanning laser and / or at least one radar or lidar and / or at least one ultrasound sensor and / or at least one communication module per channel wave. It will be noted that the wave communication module can communicate with a non-wired communication network and / or with communication modules by wave embedded in vehicles temporarily located in the environment close to the vehicle V and / or equipping information stations installed on the edge of the taxiway taken by vehicle V and / or on board satellites or other flying platforms.

At least part of these MAQ acquisition means may possibly be part of the DA assistance device. But it is not compulsory. In fact, they may be part, totally or partially, of at least one other piece of equipment on board the vehicle V. In the latter case, the data files defining the information acquired by the MAQ acquisition means can, for example, be transmitted to the assistance device DA via a communication network RC on board the vehicle V and possibly multiplexed.

It will be noted that the vehicle V has a current dynamic which is defined by quantities and a current light state which is defined by operating states of photometric functions.

For example, each dynamic variable can be chosen from an acceleration in progress, a speed in progress, a steering angle in progress for the steering wheel VV or the wheels, a rolling angle in progress, a pitch angle in progress, an angle of current yaw, as well as the time derivatives of these quantities. All these quantities are easily accessible in vehicle V, for example via the RC communication network.

Also, for example, the photometric functions can be chosen from a brake light function, a high beam function, a low beam function, a fog light function, a position light function, a daytime running light (or DRL (for Daytime running Light (or Lamp)), a hazard warning function, and a direction change indicator function. All these photometric functions are ensured by optical units fitted to vehicle V and their operating states can be “activated” or “deactivated.” These operating states are easily accessible in vehicle V, for example via the RC communication network.

As mentioned above, the invention proposes in particular an assistance method intended to assist in driving the driver CV of vehicle V.

This method comprises three steps which can be implemented at least partially by a DA assistance device according to the invention comprising, as illustrated in FIG. 1, MAL analysis means, MT processing means and means of MC control. The steps of the method are described below with reference to the example algorithm illustrated in FIG. 2.

In a first step 10, the environmental information, which has been acquired by the MAQ acquisition means, is analyzed in order to estimate a state of this environment. This analysis is carried out by the MAL analysis means of the DA assistance device.

For example, in the first step 10 we (the MAL analysis means) can (Fri) t estimate the state of the environment as a function of information acquired by the MAQ acquisition means and representative of objects which have were detected in the environment of vehicle V and / or of the traffic lane on which vehicle V travels and / or of meteorology and / or exterior light. Among this information, we can notably cite a vehicle, a pedestrian, an animal, an object, a building, a wall, a fence, a relative position of one of these with respect to vehicle V, a relative speed of the one of these with respect to vehicle V, a trajectory of one of these, a direction change indicator activated, brake lights on, hazard lights on, a type of taxiway, a number of usable traffic lanes, the type of each line delimiting the traffic lane, an emergency stop lane, a parking lot, a rest area, an intersection, a traffic light, a road sign, a posted message on a screen, day, night, darkness, darkness, wind, rain, snow, sleet, and fog. It will be understood that the type of information and the number of information used to estimate the state of the environment depend on the types of the MAQ acquisition means on board the vehicle V.

For example, an environmental report may include a car detected 20 meters in front of vehicle V, having a relative speed of -5 km / h compared to vehicle V, a traffic lane of the national road type , a speed limit of 80 km / h, poor light due to a tunnel, and rain. Also by way of example, another state of the environment may include cars detected at a stop 200 meters in front of vehicle V, a traffic lane of the fast lane type, a speed limit of 110 km / h , and a slowdown signal (or traffic jam).

In a second step 20 of the method, on the one hand, a global state of reference egr is determined which the vehicle V should present in the presence of the available dynamic quantities of the vehicle V and the state of the environment estimated during the first step 10, and, on the other hand, an overall state in progress egc of the vehicle V which is a function of the available dynamic quantities of the vehicle V, of the state of the environment estimated during the first step 10, and of the states of operation during photometric functions.

The determinations of the global reference state egr and of the global state in progress egc are carried out by the processing means MT of the assistance device DA.

It will be noted that in the second step 20, preferably at least one dynamic quantity is used among those which have been mentioned above (acceleration, speed, steering angle, roll angle, pitch angle and yaw angle).

It will also be noted that in the second step 20, the operating states in progress of at least two photometric functions are preferably used among those which have been mentioned above.

For example, a global state of reference egr may include the speed and acceleration that vehicle V should have given the estimated state of the environment, the turning radius that vehicle V should have taken into account of the estimated state of the environment, the roll, pitch and yaw angles that vehicle V should have taking into account the estimated state of the environment, the time derivatives that these angles should have, the speed of rotation that should have the steering wheel, the position that vehicle V should have in relation to the boundaries of its lane, the position that vehicle V should have in relation to each "element" detected in its environment, and the operating states that should have the photometric functions of vehicle V taking into account the estimated state of the environment.

Also by way of example, a global state in progress egc may include the speed and acceleration in progress, the angles during roll, pitch and yaw and their current time derivatives, the speed of rotation of the flywheel VV in progress , the current turning radius, the current relative position of the vehicle V in relation to the limits of its lane, the position of the vehicle V in relation to each element detected in its environment, and the current operating states of the functions photometric of vehicle V.

Note that in the second step 20 we (the MT processing means) can (Fri) t determine the global reference state egr using at least one mathematical model of neural network type. But other mathematical models can be used, such as for example and without limitation a Kalman filter, an expert system, fuzzy logic, or a forest of decision trees (or "random forest").

In a third step 30-40 of the method, a level of difference is determined between the overall state in progress egc determined and the overall state of reference egr determined, and the driver CV is alerted to a problem related to his driving of the vehicle. V and / or at least partial control of the driving of the vehicle V is taken in place of the driver CV according to the level of difference determined in the second step 20.

We will understand that depending on the level of difference, we decide either to alert the CV driver so that he focuses more on driving (with possibly at least one driving tip), or only to take at least partial control of driving vehicle V in place of driver CV, that is to say both alerting driver CV and taking at least partial control of driving vehicle V. Of course, if the level of difference is small (for example example below a predefined threshold), an alert is not triggered and the driving of the vehicle V is not taken, even partially.

Each interval of difference levels can, for example, correspond to a specific decision. Thus, we can, for example, do nothing when the difference level belongs to a first interval [0, n1 [(with n1> 0), then we can generate an alarm when the difference level belongs to a second interval [ n1, n2 [(with n2> n1), then you can generate an alarm and take partial control of the vehicle V when the difference level belongs to a third interval [n2, n3 [(with n3> n2), one can take a total control of the driving of the vehicle V when the level of difference belongs to a fourth interval [n3, n4 [(with n4> n3).

The takeover may, for example, be intended to park vehicle V as quickly as possible, for example on a rest area, an emergency stop strip or a parking space or in an after-sales service, or to better take into account one or more constraints imposed by the environment than does the CV driver.

The alert can be made by emission of an audible message by at least one speaker of vehicle V and / or by display of a text message and / or of thumbnail (s) on at least one screen of vehicle V (such as that of the central handset installed in or on the dashboard, or that of the dashboard (or a type of display called head-up)).

The determination of the level of difference (substep 30 of FIG. 2) is carried out by the processing means MT of the assistance device DA. The decision to alert and / or take control (substep 40 of FIG. 2) is made by the control means MC of the assistance device DA.

During an at least partial takeover, the control means MC determine the trajectory that vehicle V will have to follow, and commands for organs involved in the movements of this vehicle V, capable of causing the latter to follow ( V) this trajectory determined by constantly controlling its speed. These bodies are part of an assembly comprising the powertrain (or GMP), the braking system, the electric power steering and the gear change means (for example an automatic or robotized gearbox). This takeover can also take place at the photometric functions of vehicle V.

Furthermore, during a partial takeover, the CV driver continues to partially drive his vehicle V, for example, by controlling his steering wheel VV, possibly by asking him to follow as closely as possible a trajectory displayed on a screen of the vehicle V , and the control means MC control the other driving parameters.

It will be noted that in the third step 30-40 it is advantageous to take a decision according to an estimate of a temporal evolution of the level of difference determined from the level of difference having just been determined and from at least one other level of difference determined previously. This indeed makes it possible to avoid making an “instantaneous” decision, potentially useless, following the determination of a suddenly abnormal level of difference or significantly different from at least the previous level of difference, possibly following a calculation error. or an incorrect measurement of a sensor.

It will also be noted that in the third step 30-40, it is possible to alert the driver CV (of a problem linked to his driving of the vehicle V) and / or take at least partial control of the driving of the vehicle V in place of the driver CV also according to an estimated level of vigilance of the CV driver. This makes it possible to take a decision quickly adapted to the level of vigilance of the CV driver. This is particularly useful in the event of drowsiness or malaise or frequent inattention or incapacity or distraction by frequent rotation of the head towards another passenger in vehicle V.

This option requires that the DA assistance device be supplied with information representative of the driver’s level of vigilance CV. This information can be directly of the levels of difference determined by a monitoring device on board the passenger compartment of the vehicle V. In this case, the monitoring device determines the levels of difference from images of the driver's head CV acquired by a CNS digital camera installed in the passenger compartment H (for example fixedly secured to the internal mirror RI, as shown without limitation), but which may possibly be part of other equipment on board the vehicle V. As a variant, this information can be deduced, for example by the means of analysis MAL, of images of the head of the driver CV acquired by the aforesaid digital camera CNS. This CNS digital camera can therefore possibly be part of the DA assistance device.

It will also be noted that the control means MC can also be arranged so as to alert road users and those around it, for example by activating the hazard warning lights, of a problem detected in vehicle V and requiring a external assistance and / or going to impose an unusual behavior of the vehicle V, and / or to initiate a communication by means of waves to alert a rescue service.

It will also be noted that in the example illustrated without limitation in FIG. 1, the assistance device DA (and in particular its analysis means MAL, treatment means MT and control means MC) is (are) part of a CA computer, possibly ensuring at least one other function within vehicle V, such as for example the supervision of the transmission chain). But the DA assistance device could be an electronic device on board vehicle V and having its own computer. Consequently, the assistance device DA can be produced in the form of software module (s) (or computer (s) or even “software”), or else of a combination of electronic circuits (or “hardware”). ") And software module (s).

The invention offers several advantages, including:

- it protects the passengers of the vehicle as well as other road users and its surroundings against a failure or loss of capacity of the driver,

- it makes it possible to provide advice to (young) drivers in the driving learning phase and to secure their first driving experiences once the driving license has been obtained and / or to make up for any driving errors driver regardless of experience,

- it makes it possible to estimate the quality of the conduct of drivers, for example in order to reduce the insurance premiums of those who have a high quality of driving.

Claims (10)

1. Method for driving assistance for a driver (CV) of a vehicle (V) with partially autonomous driving, moving on a traffic lane, comprising acquisition means (MAQ) acquiring information representative of the environment around said vehicle (V), and having a current dynamic defined by quantities and a current light state defined by operating states of photometric functions, characterized in that it comprises i) a first step (10) in which one analyzes said information to estimate a state of said environment, ii) a second step (20) in which one determines, on the one hand, a global reference state which said vehicle (V) should present in the presence of said quantities and said estimated state of the environment, and, on the other hand, an overall current state of said vehicle (V) as a function of said quantities, of said estimated state of the environment and of said states of e operation, and iii) a third step (30-40) in which a level of difference is determined between said determined overall state in progress and said determined reference overall state, and said driver (CV) is alerted of a related problem when driving said vehicle (V) and / or at least partial control is taken of driving said vehicle (V) in place of said driver (CV) according to said determined level of difference. 2. Method according to claim 1, characterized in that in said first step (10) the said state of the environment is estimated as a function of information acquired and representative of objects detected in the environment of said vehicle (V) and / or said traffic lane and / or meteorology and / or exterior luminosity. 3. Method according to claim 1 or 2, characterized in that in said second step (20) at least one quantity chosen is chosen from an acceleration, a speed, a steering angle, a roll angle, a pitch angle, a yaw angle, and time derivatives of these quantities. 4. Method according to one of claims 1 to 3, characterized in that in said second step (20) using operating states of at least two photometric functions chosen from a stop light function, a light function main beam, a low beam function, a fog light function, a position light function, a daytime running light function, a hazard warning function, and a direction change indicator function. 5. Method according to one of claims 1 to 4, characterized in that in said second step (20) determining said global reference state by means of at least one mathematical model chosen from a neural network, a filter of Kalman, an expert system, fuzzy logic, and a forest of decision trees. 6. Method according to one of claims 1 to 5, characterized in that in said third step (30-40) alerting said driver (CV) of a problem related to his driving of said vehicle (V) and / or takes at least partial control of the driving of said vehicle (V) in place of said driver (CV) according to an estimate of a temporal evolution of said level of difference determined from the level of difference having just been determined and at least minus another level of difference previously determined. 7. Method according to one of claims 1 to 6, characterized in that in said third step (30-40) alerting said driver (CV) of a problem related to his driving of said vehicle (V) and / or takes at least partial control of the driving of said vehicle (V) in place of said driver (CV) also according to an estimate of a level of alertness of said driver (CV). 8. Device (DA) for driving assistance for a driver (CV) of a vehicle (V) with partially autonomous driving, moving on a traffic lane, comprising acquisition means (MAQ) acquiring information representative of the environment around said vehicle (V), and having a current dynamic defined by quantities and a current light state defined by operating states of photometric functions, characterized in that it comprises i) means d analysis (MAL) analyzing said information to estimate a state of said environment, ii) processing means (MT) determining, firstly, a global reference state which said vehicle (V) should have in the presence of said quantities and said estimated state of the environment, secondly, a current global state of said vehicle (V) as a function of said quantities, said estimated state of the environment and said states of f operation, and thirdly, a level of difference between said determined overall state in progress and said determined reference global state, and iii) control means (MC) triggering an alert to signal to said driver (CV) a problem linked to his driving of said vehicle (V) and / or taking at least partial control of the driving of said vehicle (V) in place of said driver (CV) according to said determined level of difference. 9. Device according to claim 8, characterized in that said control means (MC) trigger said alert and / or take at least partial control of the driving of said vehicle (V) instead of said driver (CV) also according to a estimated level of vigilance of said driver (CV). 10. Vehicle (V) with partially autonomous driving, capable of moving on a traffic lane, comprising acquisition means (MAQ) acquiring information representative of the environment around said vehicle (V), and having a dynamic in course defined by quantities and a light state in progress defined by operating states of photometric functions, characterized in that it further comprises an assistance device (DA) according to one of claims 8 and 9.