CONTROL METHOD OF A MOTOR VEHICLE, CORRESPONDING MOTOR VEHICLE

Description

[0001] The present invention relates generally to the automation of the control of motor vehicles.

[0002] It relates more particularly to a piloting method and a vehicle according to claims 1 and 9. STATE OF THE ART

[0003] It is now well known to equip a motor vehicle with an automated straight-line speed management system (in particular, we speak of an ACC system, from the English "Adaptive Cruise Control").

[0004] These systems are, for example, designed to control the vehicle so that its speed is equal to a setpoint given by the driver, except in the presence of an event on the road requiring the vehicle to slow down (traffic jams, traffic lights, etc.), in which case the vehicle's speed is controlled accordingly.

[0005] One type of event that requires special handling is curves such as bends or roundabouts. With this type of event, it is important to adjust the vehicle's speed to the curvature of the road to prevent the vehicle from skidding or the centrifugal force from being too great, given the desired level of comfort for the vehicle's passengers.

[0006] To handle this particular type of event, document WO2007070160 discloses a technical solution that is based on road curvature information stored in the vehicle's navigation system to construct a trajectory and then a speed profile that the vehicle must follow along this trajectory. Documents DE 102013210916A1 and US 2012 / 0209489 A1 present comparable control systems.

[0007] Although it allows for anticipation of curves, this technical solution is highly dependent on the information provided by the navigation system, which information is itself subject to significant uncertainties, which can be potentially dangerous for the vehicle's passengers. PRESENTATION OF THE INVENTION

[0008] In order to overcome the aforementioned drawback of the state of the art, the present invention proposes to measure or calculate in real time the lateral acceleration experienced by the vehicle, and to adjust the longitudinal speed setpoint of the vehicle as a function of this lateral acceleration.

[0009] More particularly, the invention proposes a control method as defined by claim 1.

[0010] Thus, thanks to the invention, it is possible to limit the impact of estimation errors from the navigation system and to correct, in real time, the speed of the vehicle in order to maintain a lateral acceleration which is acceptable to the vehicle's passengers and to the vehicle itself and which allows the vehicle not to slow down more than necessary.

[0011] This process does not require the addition of any particular sensor in the sense that vehicles currently on the market are already equipped with sensors allowing its implementation, so it can be implemented easily and at no additional cost on this type of vehicle.

[0012] This method can also be implemented in vehicles without a navigation system.

[0013] Finally, its development will prove particularly simple since it could consist of adjusting only two parameters (the regulator gains) in order to take into account the specificities of the range of vehicles on which this solution would be deployed.

[0014] Other advantageous and non-limiting characteristics of the control method according to the invention, taken individually or in all technically possible combinations, are the following: the longitudinal speed setpoint is determined such that the lateral acceleration experienced by the motor vehicle is less than or equal to a lateral acceleration setpoint, and preferably equal to said lateral acceleration setpoint; the longitudinal speed setpoint is determined such that the derivative of the longitudinal acceleration experienced by the motor vehicle is as close as possible to zero; the derivative of the longitudinal speed setpoint has a saturated value between two limits; the verification, determination and control steps are repeated in a loop, at a sampling frequency greater than one Hertz; during the verification step, the computer checks whether the distance between the vehicle and said curve is less than a distance threshold and whether the steering wheel tilt angle is greater than an angular threshold;the process is terminated when the steering wheel tilt angle has reached a maximum and then decreased by an angle greater than another angular threshold; the process is terminated when an event downstream of said curve is detected which requires braking of the motor vehicle.;

[0015] The invention also proposes a vehicle as defined in the introduction, the computer of which will be programmed to implement the aforementioned method.

[0016] Of course, the various features, variants and embodiments of the invention may be combined with each other in various combinations to the extent that they are not incompatible or mutually exclusive. DETAILED DESCRIPTION OF THE INVENTION

[0017] The description which follows with reference to the appended drawings, given as non-limiting examples, will make it clear what the invention consists of and how it can be implemented.

[0018] On the attached drawings: [ Fig. 1 ] is a schematic view of a vehicle and a road taken by that vehicle; [ Fig. 2 ] is a graph illustrating the variation in lateral acceleration experienced by the vehicle from the figure 1 depending on the radius of curvature of the road taken, for 15 distinct speed values; [ Fig. 3 ] is a representation of the steps of a control method in accordance with the invention; [ Fig. 4 ] is a graphical representation of a control law used to calculate a longitudinal speed setpoint for the vehicle of the figure 1 ; [ Fig. 5 ] is a graph illustrating the maximum steering wheel angle variation as a function of the longitudinal speed of the vehicle of the figure 1 ; [ Fig. 6 ] is a graph illustrating the variation of a GW variable as a function of the ratio between the current steering wheel angle and the maximum angle; [ Fig. 7 ] is a timing diagram illustrating a first example of variation of an activation indicator of the method according to the invention; [ Fig. 8 ] is a graph timed on the timeline of the figure 7 , illustrating the variations in the angle of the vehicle's steering wheel; [ Fig. 9 ] is a graph timed on the timeline of the figure 7 , illustrating the variations of a reference lateral acceleration and the lateral acceleration of the vehicle; [ Fig. 10 ] is a graph timed on the timeline of the figure 7 , illustrating the variations in the longitudinal acceleration of the vehicle; [ Fig. 11 ] is a graph timed on the timeline of the figure 7 , illustrating the variations in longitudinal speed of the vehicle; [ Fig. 12 ] is a timing diagram illustrating a second example of variation of an activation indicator of the method according to the invention; [ Fig. 13 ] is a graph timed on the timeline of the figure 12 , illustrating the variations in the angle of the vehicle's steering wheel; [ Fig. 14 ] is a graph timed on the timeline of the figure 12 , illustrating the variations of a reference lateral acceleration and the lateral acceleration of the vehicle; [ Fig. 15 ] is a graph timed on the timeline of the figure 12 , illustrating the variations in the longitudinal acceleration of the vehicle; [ Fig. 16 ] is a graph timed on the timeline of the figure 12 , illustrating the longitudinal speed variations of the vehicle.

[0019] On the figure 1 , a motor vehicle 10 is shown driving on a road 20.

[0020] We will consider here that this road comprises a straight section 21 followed by a first bend 22, itself possibly followed by a second bend 22.

[0021] Vehicle 10 is a car here, but it could be another type of vehicle (motorcycle, truck, etc.). It has a chassis that is supported by wheels and which itself supports various equipment, including a powertrain, braking systems, and a computer.

[0022] It could be a manually driven vehicle, in which case it would be equipped with driving assistance systems, or, preferably, an autonomous vehicle.

[0023] As shown in the figure 1 , we will consider here a reference (X, Y, Z) attached to the vehicle, whose longitudinal axis X will be oriented from the rear to the front of the vehicle, whose lateral axis Y will be oriented towards the left side of the vehicle, and whose vertical axis Z will be oriented upwards.

[0024] This vehicle 10 is preferably equipped with a navigation system comprising a mapping system listing roads, events and information relating to these roads and events, and a geolocation system (for example of the GPS type) adapted to position the vehicle 10 on these roads.

[0025] Here we will consider a particular type of event, namely curved areas of the road (bends, roundabouts, etc.). In the example illustrated on the figure 1 , we will consider turns 22, 23 in particular. The navigation system will be able to determine the position of the vehicle in relation to turns 22, 23, as well as the radii of curvature of these turns.

[0026] The vehicle 10 is also equipped with a sensor enabling it in particular to determine the lateral acceleration experienced by the vehicle (i.e. the acceleration along the lateral axis Y). This could be an inertial sensor or any other type of sensor. Here, it will be a sensor adapted to measure the yaw rate dΨ / dt of the vehicle.

[0027] The vehicle 10 is also equipped with a longitudinal speed sensor adapted to measure the speed at which the vehicle is moving along the road 20. This speed sensor is, for example, adapted to measure the rotational speed of the vehicle's wheels to deduce therefrom the value of the longitudinal speed dx / dt. This sensor will also make it possible to calculate the value of the longitudinal acceleration d 2< x / dt 2< of the vehicle. Alternatively, an acceleration sensor could be used for this.

[0028] In order to process the information provided by these sensors and to be able to develop a longitudinal speed setpoint dx* / dt for the vehicle 10, the latter is equipped with a computer 11.

[0029] This calculator 11 includes a processor, memory and various input and output interfaces.

[0030] Thanks to its input interfaces, the ECU is adapted to receive input signals from the longitudinal speed and yaw rate sensors.

[0031] The computer's memory stores a computer application, consisting of computer programs comprising instructions whose execution by the processor allows the computer to implement the method described below.

[0032] Among these computer programs, one provides an adaptive cruise control (ACC) function.

[0033] According to the invention, another of these programs provides a curve cruise control function (CCC).

[0034] The adaptive cruise control function ACC (hereinafter referred to as the ACC function) being well known to those skilled in the art, it will not be described here. It will only be explained that it makes it possible to calculate a longitudinal speed setpoint for the vehicle 10 adapted to the road and to the speed of the motor vehicle preceding the vehicle 10 considered here.

[0035] The CCC curve cruise control function (hereinafter referred to as the CCC function) being more particularly the subject of the present invention, it will be described in detail below.

[0036] At this stage, we can only specify that this CCC function is also intended to calculate a longitudinal speed setpoint for the vehicle 10 when the latter is on a curve and certain conditions are met. This CCC function will therefore be intended to replace the ACC function under these conditions only.

[0037] Thanks to its output interfaces, the computer is adapted to transmit instructions to the various components of the vehicle, and in particular to the powertrain and the braking means. It thus makes it possible to control the acceleration, deceleration and braking of the vehicle autonomously, without intervention from the driver of the vehicle 10.

[0038] The method according to the invention is therefore intended to generate a longitudinal speed setpoint dx* / dt when the vehicle is in a bend and certain conditions are met.

[0039] As a preliminary point, we can specify that the notation d / dt will be used to designate the first derivative of a parameter (in practice this will be a speed).

[0040] The notation d 2

[0041] The notation d 3

[0042] The index “est” will be used to express the estimated nature of the parameter to which it will be affixed.

[0043] The exponent “*” will, on the other hand, be used to express that the parameter to which it will be affixed is an instruction.

[0044] We can now describe in detail how the computer implements the method of calculating the longitudinal speed setpoint dx* / dt, with reference to the figure 3 .

[0045] This process will be initiated as soon as the vehicle 10 starts. It will be implemented in a loop, at a high sampling frequency, well above the Hertz.

[0046] Here, we will consider that the vehicle is already traveling on the straight section 21 of route 20.

[0047] Then, during a first step E2, the calculator determines whether the ACC function is activated or not.

[0048] It may indeed happen that it is deactivated, in particular in the case where the driver wishes to drive his vehicle himself 10. In this event, the process is terminated.

[0049] Otherwise, the process continues in a second step E4 during which the computer 11 checks whether the navigation system detects an event such as, in our example, a turn 22.

[0050] This “detection” is carried out with a predetermined range, for example less than 100 meters (this distance may be different and may vary depending on the speed of the vehicle).

[0051] As long as no turn is detected, this step E4 is repeated in a loop. Then, as soon as a turn 22 is detected, the process continues with step E6.

[0052] During this step E6, the computer determines a longitudinal entry speed dx e / dt in turn 21.

[0053] In practice, the computer 11 seeks the speed at which the bend must be approached so that the centrifugal force applied to the vehicle is equal to a target value. In other words, the objective is for the vehicle to be able to approach the bend at a speed that guarantees it a determined lateral acceleration d 2 < y / dt 2 <.

[0054] At this stage, the computer can only rely on the information stored in the navigation system (alternatively or in addition, it could rely on information acquired by means of a camera) and on a map stored in its memory, which is illustrated in the figure 2 .

[0055] This figure illustrates the variation, as a function of the radius of curvature Rc of road 20, of the lateral acceleration d 2< y / dt 2< which would be experienced by the vehicle at different longitudinal speeds V (or dx / dt).

[0056] A reference curve C1 has also been shown which illustrates the variation of the lateral acceleration setpoint d 2< y / dt 2< at which we want the vehicle to approach a bend (so that it experiences a determined centrifugal force).

[0057] The longitudinal input velocity dx e / dt is then read from this map.

[0058] In practice, the longitudinal input speed dx e / dt is more precisely obtained by linear interpolation, as a function of the radius of curvature Rc of the input on turn 22 stored in the navigation system, so as to be on the reference curve C1.

[0059] In the following step E8, the ACC function of the computer 11 then establishes a speed reference profile making it possible to control the powertrain and the braking means in such a way that the vehicle arrives at the entrance to bend 22 at the longitudinal entry speed dx e / dt.

[0060] This strategy allows the computer 11 to anticipate turn 22 by decelerating the vehicle well before the turn. Here, this deceleration phase is achieved by first using the engine brake, then the engine brake coupled with the braking means if necessary. Of course, any other suitable strategy would be usable.

[0061] During the next step E10, the computer seeks to determine whether vehicle 10 is still located upstream of turn 22, or whether it is located at the entrance to or in turn 22.

[0062] In fact, this condition is necessary to activate the CCC function and substitute it for the ACC function.

[0063] In this step, the calculator uses a Boolean indicator β, which is initially set to zero.

[0064] Here, three conditions are specifically provided for activating the CCC function for cornering the vehicle.

[0065] The first condition, which is already met at this stage, is that the event detected by the navigation system is indeed of the curve type (here it is a turn 22).

[0066] The second condition is that the distance separating vehicle 10 from turn 22 is less than a predefined distance threshold S d, here chosen to be equal to 10 meters.

[0067] This distance threshold S d is fixed here but it would be possible to use a threshold which varies according to the speed of the vehicle and / or the reaction time of the system and / or the characteristics of the bend.

[0068] The third condition is that the vehicle's steering wheel is turned by an angle of inclination α (measured relative to the steering wheel's mid-position) greater than a predetermined angular threshold S a1, here equal to 3 degrees. Here again, it would be possible for this threshold to vary according to different parameters.

[0069] As long as all three of these conditions are not met, step E10 is repeated in a loop and the indicator β is left equal to zero.

[0070] As soon as they are combined, which means that the vehicle has entered turn 22, the computer 11 assigns the value 1 to the indicator β.

[0071] Then, the process continues in steps E12 then E14, allowing the longitudinal speed setpoint dx* / dt to be calculated.

[0072] These steps and the following ones are implemented in a loop so that the longitudinal speed setpoint dx* / dt is adjusted dynamically throughout turn 22, at a high frequency, according to the actual conditions encountered by the vehicle 10 (possibly independently of the data from the navigation system).

[0073] The objective is to ensure that the lateral acceleration d 2< y / dt 2< experienced by the vehicle is controlled in real time.

[0074] The vehicle speed could be controlled so that the lateral acceleration d 2< y / dt 2< remains constant through the bend.

[0075] However, here, we will ensure that this lateral acceleration d 2< y / dt 2< varies according to the radius of curvature of the part of turn 22 taken, so that it follows the reference curve C1 illustrated on the figure 2 .

[0076] The objective is also for the computer 11 to take into account the longitudinal input speed dx e / dt, then to adjust it according to the lateral acceleration d 2< y / dt 2< experienced by the vehicle, avoiding any excessively sudden braking.

[0077] In practice, the calculation of the longitudinal speed setpoint dx* / dt is carried out in a closed loop, using a control law such as that illustrated in the figure 4 This regulation is therefore not only a limitation but authorizes or even controls the acceleration of the vehicle if necessary.

[0078] This control law uses two regulators K 1 , K 2 .

[0079] The first regulator is used so that the lateral acceleration d 2< y / dt 2< of the vehicle (or rather its estimate noted d 2< y est / dt 2< ) is equal to a lateral acceleration setpoint d 2< y* / dt 2< .

[0080] The lateral acceleration setpoint d 2< y* / dt 2< is read from the map illustrated on the figure 2 , depending on the measured speed dx / dt of the vehicle and the radius of curvature Rc of the bend (which can be measured in various known ways using sensors, or acquired using the navigation system).

[0081] The estimation of the lateral acceleration d 2< y est / dt 2< experienced by the vehicle 10 is carried out here as a function of the measured values of the yaw rate dΨ / dt and the longitudinal speed dx / dt of the vehicle. Alternatively, one and / or the other of these values could be calculated, for example by means of a Kalman filter. The computer 11 then estimates the lateral acceleration d 2< y est / dt 2< by means of the following mathematical equation: d 2 y est / dt 2 = dΨ / dt . dx / dt

[0082] The K 1 regulator then makes it possible to obtain a first acceleration setpoint d 2< x 1 / dt 2< , taking into account the difference between the lateral acceleration setpoint d 2< y* / dt 2< and the estimate of the lateral acceleration d 2< y est / dt 2< .

[0083] The second regulator K 2 is designed to limit oscillations of longitudinal acceleration in turn 22 (better known under the English name of “Jerk”).

[0084] To do this, the calculator estimates the derivative of the longitudinal acceleration d 3< x est / dt 3< of the vehicle 10. This estimation is carried out here by means of a filtered derivative, as a function of the longitudinal acceleration d 2< x est / dt 2< .

[0085] The acceleration derivative instruction, noted d 3< x* / dt 3< , is chosen to be zero.

[0086] The K 2 regulator then makes it possible to obtain a second acceleration setpoint d 2< x 2 / dt 2< , taking into account the estimate of the longitudinal acceleration d 2< x est / dt 2< .

[0087] On this figure 4 , the parameters T der and T lp are time constants.

[0088] The time constant T der is used to perform the "pseudo-derivative" or "filtered derivative" of the vehicle's longitudinal acceleration to obtain the value of d 3< x est / dt 3< .

[0089] The time constant T lp is used for the “low-pass” filter and allows to reduce the measurement noise on the estimation of the lateral acceleration d 2< y / dt 2< .

[0090] Both regulators K 1 , K 2 have gains that must be calibrated during vehicle design and / or during vehicle use.

[0091] Conventionally, the gain of the K 1 regulator is used to adjust the dynamics of the speed setpoint. As with any control law, significant dynamics can lead to overshoots and oscillations in the vehicle speed. Here, to obtain the best possible performance, while avoiding excessive overshoots and too many speed oscillations, the gain of the K 1 regulator is chosen to be equal to 1.5.

[0092] The gain of the K2 regulator limits oscillations in the vehicle's longitudinal acceleration. Increasing this gain limits overshoot and oscillations but results in a slowdown in control dynamics. The compromise chosen for this gain is 0.85.

[0093] Of course, these fixed values may vary depending on the type of vehicle.

[0094] Alternatively, these values may vary depending on the conditions in which the vehicle is driven, and in particular depending on the mass of the vehicle (which will depend on the number of occupants and the load on board). For this purpose, it will be possible to use maps developed during the design of the vehicle and stored in the memory of the computer 11.

[0095] As shown in the figure 2 , the two acceleration instructions d 2< x 2 / dt 2< , d 2< x 2 / dt 2< are summed and this sum is then multiplied by a variable Gw.

[0096] This variable Gw is used as a weighting factor so as to progressively give weight to the two acceleration instructions when vehicle 10 enters turn 22.

[0097] The determination of this variable Gw is carried out in two stages, by means of two maps illustrated on the figures 5 et 6 .

[0098] Firstly, the calculator determines, based on the longitudinal speed dx / dt of the vehicle 10, a maximum steering wheel angle value α max (see figure 5 ). This maximum value corresponds to a value beyond which the safety and comfort of the vehicle's passengers would no longer be ensured satisfactorily.

[0099] In a second step, the calculator determines the variable Gw as a function of the ratio between the measured steering wheel angle α and the maximum steering wheel angle value α max.

[0100] The calculator thus obtains a third acceleration setpoint d 2< x 3 / dt 2< , which is then saturated so as to limit the acceleration setpoint depending in particular on the acceleration capabilities of the vehicle 1, then integrated so as to obtain the longitudinal speed setpoint dx* / dt.

[0101] On the figure 3 , step E12 corresponds to the calculation of the variable Gw and step E14 corresponds to the calculation of the longitudinal speed setpoint dx* / dt and to the sending of this setpoint to the powertrain and / or the braking means.

[0102] The method then continues in steps E16 and E18 which make it possible to check whether one or other of two conditions for deactivating the CCC function are met (the process being interrupted as soon as one of these conditions is met).

[0103] The first condition is to determine whether or not the driver returns the steering wheel to its middle position, which means that vehicle 10 is exiting the bend.

[0104] For this, the maximum inclination angle α up reached by the steering wheel since entering turn 22 is detected, then as soon as the inclination angle α drops beyond a predetermined percentage of this maximum inclination angle α up , the process stops.

[0105] This percentage is predefined and is for example between 30 and 90%. Here it is equal to 50%. Thus, as soon as the steering wheel returns to its middle position by turning beyond a second angular threshold S a2 equal to half of its maximum inclination angle α up , the process is interrupted. Then, the indicator β is reset to zero.

[0106] Before describing the second condition, we can observe on the figures 7 à 11 an example of implementation of the piloting process.

[0107] In this example, we are considering the case where route 20 only has one turn 22.

[0108] On the figure 7 , we observe that the indicator β, which we recall indicates whether or not the vehicle has entered turn 22, is equal to zero as long as the vehicle is traveling on the straight section 21, then it is set to 1 at time t 2 at which the three aforementioned conditions are met.

[0109] We observe on the figure 11 that the longitudinal speed dx / dt of the vehicle decreases progressively from the instant t 0 when turn 22 was detected. It decreases slowly at first, by means of the engine brake alone, then more significantly from the instant t 1 , by means of this engine brake and the braking means. We observe on the figure 10 the variations of longitudinal acceleration d 2< x / dt 2< .

[0110] On the figure 8 , we observe that at time t 2 , the angle of inclination α of the steering wheel has started to increase (and exceeds 3°). It continues its progression then stabilizes at approximately 12 degrees until time t 3 where it falls sharply.

[0111] This fall causes, a short time later (at time t 4 ), the indicator β to be reset to zero.

[0112] On the figure 9 , we have represented, by curve C2, the variation of the lateral acceleration setpoint d 2< y* / dt 2< . We have also represented, by curve C3, the variation of the estimate of the lateral acceleration d 2< y est / dt 2< . We observe that this estimate is continuous and continuously derivable, which avoids any suddenness for the passengers. We also observe that it quickly follows the shape of the setpoint, which makes it possible to avoid the lateral acceleration undergone by the vehicle being too great.

[0113] This method of calculating the longitudinal speed setpoint dx* / dt has the advantage of being able to slow down the vehicle if conditions require it, or to accelerate the vehicle, for example, because the bend opens gradually. The vehicle is therefore not blocked at its longitudinal entry speed dx e / dt. In fact, this calculation method is not only active upstream of the bend but throughout it.

[0114] We can now describe the second condition.

[0115] This second condition consists of determining whether an event located in the bend or further down the road requires the vehicle to brake, and returning to the ACC function in this case.

[0116] To do this, the computer seeks to detect any event located downstream of vehicle 10 along route 20, within a given range, which is for example a range of three kilometers using information from the navigation system.

[0117] For each of these events, the computer determines whether it will be necessary to reduce the longitudinal speed setpoint dx* / dt of the vehicle 10 by decelerating with a longitudinal acceleration Decel, in absolute value, greater than a predetermined deceleration threshold.

[0118] This longitudinal acceleration is estimated here by considering the instantaneous speed V 0 of the vehicle, the speed V 1 at which the vehicle will have to approach this event and the distance D separating the vehicle 10 from this event, and by carrying out the following calculation: Decel = V 0 2 − V 1 2 / 2 . D

[0119] If the longitudinal acceleration Decel is greater than the predetermined deceleration threshold, the process stops. Then, the indicator β is reset to zero.

[0120] We can observe on the figures 12 à 16 another example of implementation of the piloting process, illustrating this second condition well.

[0121] In this example, we are considering the case where route 20 has two separate and successive turns 22, 23.

[0122] On the figure 12 , we observe that the indicator β is equal to zero as long as the vehicle is traveling on the straight section 21, then it is set to 1 when the vehicle enters the bend. Indeed, at this stage, the second bend 23 is so far away that the longitudinal acceleration Decel is lower than the predetermined threshold.

[0123] The β indicator is however reset to zero when the computer notes that the longitudinal acceleration Decel required to enter the second turn 23 with an adequate speed exceeds the predetermined deceleration threshold.

[0124] Therefore, the vehicle's longitudinal speed setpoint is developed using the CCC function only in the first part of the first bend 22. The end of this bend and the straight section between the two bends is approached at a speed defined by the ACC function. Then, the β indicator is reset to 1 when the vehicle enters the second bend. This β indicator is then maintained at this value until the exit of this second bend 23.

[0125] We observe on the figure 16 that the longitudinal speed dx / dt of the vehicle decreases before entering the first bend and then before entering the second bend.

[0126] We observe on the figure 15 the variations of longitudinal acceleration d 2< x / dt 2< .

[0127] On the figure 13 , we observe that the angle of inclination α of the steering wheel varies in one direction in the first bend, and in the other in the second. It is this which causes the indicator β to be reset to zero at the exit of the second bend.

[0128] On the figure 14 , we have represented, by curve C4, the variation of the lateral acceleration setpoint d 2< y* / dt 2< . We have also represented, by curve C5, the variation of the estimate of the lateral acceleration d 2< y est / dt 2< . We observe that this estimate is continuous and continuously derivable and that it quickly follows the form of the setpoint.

[0129] The present invention is in no way limited to the embodiment described but those skilled in the art will be able to provide any variation in accordance with the invention.

Claims

1. Method for controlling a motor vehicle (10), comprising: - a checking step during which a computer (11) on board the motor vehicle (10) checks whether the motor vehicle (10) is entering a curve (21; 22) or in a curve (21; 22), and then, if such is the case, - an acquisition step of acquiring a lateral acceleration (d2yest / dt2) experienced by the motor vehicle (10), - a determination step during which the computer (11) determines a longitudinal-speed setpoint (dx* / dt) as a function of the lateral acceleration (d2yest / dt2) acquired, and - a control step during which the vehicle is controlled according to the longitudinal-speed setpoint (dx* / dt), and wherein the derivative of the longitudinal-speed setpoint (dx* / dt) has a value that is dependent upon an angle of inclination (α) of the steering wheel of the motor vehicle (10) and is determined so as to have: - a value of zero if the angle of inclination (α) is zero, - a maximum value if the angle of inclination (α) is equal to a maximum determined according to the radius of curvature of the curve (21; 22), and otherwise - a value that is dependent on the angle of inclination (α) and that is between zero and said maximum value.

2. Control method according to the preceding claim, wherein the longitudinal-speed setpoint (dx* / dt) is determined such that the lateral acceleration (d2yest / dt2) experienced by the motor vehicle (10) is less than or equal to a lateral-acceleration setpoint (d2y* / dt2), and preferably equal to said lateral-acceleration setpoint (d2y* / dt2).

3. Control method according to one of the preceding claims, wherein the longitudinal-speed setpoint (dx* / dt) is determined such that the derivative of the longitudinal acceleration (d3xest / dt3) experienced by the motor vehicle (10) is as close as possible to zero.

4. Control method according to one of the preceding claims, wherein the derivative of the longitudinal-speed setpoint (dx* / dt) has a saturated value between two limits.

5. Control method according to one of the preceding claims, wherein the checking, determination and control steps are iterated loopwise, at a frequency higher than one Hertz.

6. Control method according to one of the preceding claims, wherein, during the checking step, the computer (11) checks whether the distance between the vehicle (10) and said curve is less than a distance threshold (Sd), and whether the angle of inclination (α) of the steering wheel is greater than an angular threshold (Sa1).

7. Control method according to one of the preceding claims, wherein the method is terminated when the angle of inclination (α) of the steering wheel has reached a maximum and has then decreased by an angle greater than a threshold (Sa2).

8. Control method according to one of the preceding claims, wherein the method is terminated when an event beyond said curve (21; 22) and requiring braking of the motor vehicle (10) is detected.

9. Motor vehicle (10) including a power unit, braking means and a computer (11), characterized in that the computer (11) is programmed to implement a control method according to one of the preceding claims.