Method for adjusting the lateral position of a vehicle

Abstract

A method for regulating the lateral position of a vehicle (1) on a traffic lane (10), the method comprising: a step (E21) of calculating a reference state vector (Xref) of the vehicle; a step (E22) of calculating an observed state vector (Xobs) of the vehicle; a step (E23) of calculating a steering angle setpoint (δc) of a steering wheel (2f) of the vehicle based on a difference (Xerr) between the reference state vector (Xref) and the observed state vector (Xobs); an initialization step (E1) performed when the regulation method is activated, the initialization step comprising a sub-step (E11) of calculating a component (X07) of an initial observed state vector (X0) such that the steering angle (δc) setpoint of the steering wheel is equal to the value of the steering angle (δm) of the steering wheel measured.

Description

Technical Field

[0001] The present invention relates to a method for adjusting the lateral position of a vehicle, in particular an autonomous vehicle, on a roadway. In particular, the invention relates to an initialization step performed when activating the adjustment method. The invention also relates to a motor vehicle comprising hardware and / or software means for implementing the method. Background Art

[0002] Autonomous vehicles typically include a computer-controlled steering system. The steering system controls the orientation of the vehicle's steering wheels, enabling the vehicle to autonomously follow a given trajectory. The computer implements a steering process that allows the vehicle to be positioned laterally in its lane. In particular, the computer can implement a steering process known as Lane Centering Assist (LCA), which allows the vehicle to remain centered in its lane.

[0003] When this process is activated during driving, the vehicle moves from a trajectory manually defined by the driver to an autonomously determined reference trajectory. Under certain circumstances, this trajectory change can cause significant lateral accelerations. These lateral accelerations can cause passenger anxiety or even lead to vehicle instability or loss of grip.

[0004] In particular, it should be noted that under certain specific circumstances, the lateral accelerations experienced by the vehicle during the transition from a manually controlled trajectory to an autonomously controlled trajectory can be particularly high. These lateral accelerations are particularly high when the vehicle's sensors, particularly the angle sensors on the vehicle's steering wheel, are uncalibrated or exhibit a calibration offset (i.e., a difference between the sensor's measured value and the true value). These lateral accelerations are also particularly high during this transition if the vehicle is traveling on a sloping or cambered roadway (i.e., a laterally tilted roadway).

[0005] Description of the Invention

[0006] The object of the present invention is to provide a method for regulating the position of a vehicle on a roadway which remedies the aforementioned disadvantages and improves the regulation methods known from the prior art.

[0007] More specifically, a first object of the invention is a method for regulating the lateral position of a vehicle on a roadway, which method avoids excessive lateral accelerations of the vehicle during activation.

[0008] A second object of the invention is a method for regulating the position of a vehicle on a roadway, which method allows limiting the maximum steering angle of the vehicle's steerable wheels. Summary of the Invention

[0009] The present invention relates to a method for adjusting the lateral position of a vehicle on a roadway, the adjustment method comprising:

[0010] - a step of calculating a reference state vector of the vehicle,

[0011] - a step of calculating the observed state vector of the vehicle,

[0012] - a step of calculating a steering angle setpoint for the vehicle's steered wheels as a function of the difference between the reference state vector and the observed state vector,

[0013] The regulation method includes an initialization step performed when the regulation method is activated, and the initialization step includes a sub-step, which calculates the components of the initial observation state vector so that the steering angle set points of these steering wheels when the regulation method is activated are equal to the steering angle values ​​of these steering wheels measured when the regulation method is activated.

[0014] The observed state vector may include all or some of the following components:

[0015] - the vehicle's yaw rate, and / or

[0016] - the heading angle of the vehicle, and / or

[0017] - the lateral speed of the vehicle, and / or

[0018] - the lateral distance of the vehicle relative to the reference trajectory, and / or

[0019] - the steering speed of the vehicle's steering wheels, and / or

[0020] - the steering angle of the vehicle's steering wheels, and / or

[0021] -Integral of the vehicle's lateral distance relative to the reference trajectory.

[0022] During the sub-step of calculating the components of the initial observed state vector, said components can be calculated as a function of the steering angle values ​​measured when the regulation method is activated.

[0023] The observed state vector may comprise a component equal to the integral of the lateral distance of the vehicle relative to the reference trajectory, the initial value of the integral of the lateral distance defined during the initialization step being calculated so that the steering angle setpoints of the steered wheels when the regulation method is activated are equal to the steering angle values ​​of the steered wheels measured when the regulation method is activated.

[0024] The step of calculating the steering angle setpoint may comprise multiplying the difference between the reference state vector and the observed state vector by a regulation vector, the regulation vector being dependent on the speed of the vehicle.

[0025] An initial value for the integral of the vehicle's lateral distance relative to the reference trajectory can be calculated from the components of the difference between the reference state vector and the observed state vector, from the components of the adjustment vector, and from the steering angle value measured when the adjustment method is activated.

[0026] The calculation of the observed state vector can be based on quantities measured by on-board sensors of the vehicle, constants characterizing the vehicle, the speed of the vehicle and the setpoint value of the steering angle of the steered wheels of the vehicle calculated during a previous iteration of the regulation method.

[0027] The invention also relates to a computer program product comprising program code instructions stored on a computer-readable medium for implementing the steps of the regulation method as defined above when said program is run on a computer.

[0028] The invention also relates to a computer-readable data storage medium storing a computer program comprising program code instructions for implementing the regulation method as defined above.

[0029] The invention also relates to a motor vehicle comprising hardware and / or software means for implementing the regulation method as defined above. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] These objects, features and advantages of the present invention will be explained in detail in the following description of specific embodiments given in a purely non-limiting manner with reference to the accompanying drawings, in which:

[0031] [ Figure 1 ] Figure 1 is a schematic top view of a motor vehicle according to an embodiment of the present invention.

[0032] [ Figure 2 ] Figure 2 It is a top-down schematic diagram of vehicles on the lane.

[0033] [ Figure 3 ] Figure 3 It is a front view schematic diagram of vehicles on the driving lane.

[0034] [ Figure 4 ] Figure 4 is an overview of an adjustment method according to an embodiment of the present invention.

[0035] [ Figure 5 ] Figure 5 is a schematic diagram of a regulator used in the regulation method.

[0036] [ Figure 6 ] Figure 6is a graph showing the performance of a regulation method according to an embodiment of the present invention compared with a regulation method according to the prior art. DETAILED DESCRIPTION

[0037] Figure 1 A motor vehicle 1 according to an embodiment of the present invention is schematically illustrated. Vehicle 1 can be of any type. Specifically, it can be, for example, a car, a utility vehicle, a truck, or a bus. Vehicle 1 has two steered front wheels 2f and two rear wheels 2r. The orientation of the steered wheels 2f can be controlled by a steering system 3. Steering system 3 includes a steering device 4 mechanically connected to the two front wheels 2f and a steering wheel 5 mechanically connected to the steering device 4. The steering system may also include a power-assisted steering module, which is, for example, integrated into the steering device 4. Steering system 3 also includes an electronic control unit 6 and a steering wheel angle sensor 7. Steering wheel angle sensor 7 is capable of measuring the orientation of steering wheel 7. Specifically, the steering wheel angle sensor can measure the angular position of a steering column connected to steering wheel 7. The orientation of steering wheel 7 is proportional to the orientation of steered wheels 2f (i.e., the steering angle of steered wheels 2f). Therefore, controlling the steering angle of the steered wheels is equivalent to controlling the angle at the steering wheel of vehicle 1. Vehicle 1 may also be equipped with other sensors, such as a yaw sensor and / or at least one wheel speed sensor that allows the vehicle speed to be determined.

[0038] The electronic control unit 6 is electrically connected to the steering wheel angle sensor 7, the steering device 4, and in some cases other sensors of the vehicle. In particular, the electronic control unit includes a memory, a microprocessor and an input / output interface, which are used to receive data provided by other devices of the vehicle 1 or to transmit data to other devices of the vehicle 1. The memory of the electronic control unit is a data storage medium for storing a computer program, which includes program code instructions for implementing a method according to an embodiment of the present invention. The microprocessor can execute the method. In particular, the electronic control unit 6 can send a control command to the steering device 4 via its input / output interface so as to steer the steering wheel (i.e., orient) according to the calculated angle. The longitudinal axis X1 of the vehicle can be defined as the axis parallel to the direction of the vehicle's straight-line travel.

[0039] The vehicle 1 further comprises a device 8 for detecting the environment of the vehicle 1, for example a radar, a lidar or a camera. The detection device 8 is also connected to the electronic control unit 6. The vehicle 1 is an autonomous vehicle, i.e. the steering system can control the orientation of the steered wheels 2f without the driver having to operate the steering wheel 5. The vehicle 1 can thus be driven along a steering track and remain on the steering track without driver intervention. The vehicle 1 can also be conventionally controlled by the driver actuating the steering wheel. The vehicle 1 can thus be used in two separate functional modes: in a first mode, referred to as manual mode, the orientation of the steered wheels is controlled by the driver. In a second mode, referred to as autonomous mode, the orientation of the steered wheels is controlled by the electronic control unit, in particular by means of the detection device 8 connected thereto.

[0040] Figure 2 A vehicle 1 is shown traveling on a lane 10. The left and right sides of the lane 10 are delimited by two dividing lines 11, which are marked on the lane, for example, in the form of continuous or dashed white or yellow lines. The detection device 8 is capable of identifying the dividing lines 11. The electronic control unit includes software means allowing the calculation of a reference trajectory 12 or setpoint trajectory. The reference trajectory 12 is Figure 2 denoted by a dashed line. Reference trajectory 12 is a theoretical line not visible on lane 10. Reference trajectory 12 may be, for example, a line located equidistant from both dividing lines 11. Alternatively, reference trajectory 12 may be defined in different ways. It may be offset more toward one or the other of the two dividing lines 11. It may also be calculated based on the detection of obstacles or other vehicles on lane 10 or on lanes adjacent to lane 10.

[0041] In this document, the longitudinal axis X is defined as the lane axis parallel to the reference track at the height of vehicle 1. The transverse axis Y is the lane axis perpendicular to the reference track at the height of vehicle 1. The axis Z is the axis perpendicular to the plane formed by the lane. Axes X, Y, and Z form an orthogonal reference system.

[0042] Figure 2 The lanes shown in FIG. 1 are straight lines. However, the present invention can also be used when the lanes describe curves or bends. Figure 3As shown, the lane 10 may be inclined. Therefore, the Z axis perpendicular to the surface of the lane 10 is inclined by an angle A1 that is not equal to zero relative to the vertical axis Z'. The angle A1 is oriented transversely. This means that the lane is inclined along the transverse axis Y. When a vehicle travels on such a lane, the steered wheels must be turned slightly (usually towards the top of the slope) so that the vehicle maintains a straight trajectory. Slip may then occur at the junction between the steered wheels 2f and the road surface. In this case, the steered wheels are not perfectly aligned with the direction followed by the vehicle. Therefore, the steering wheel angle value provided by the steering wheel angle measured by the steering wheel angle sensor is offset relative to the steering wheel angle value that would be applied if the vehicle followed the same heading on a flat road surface. Compared to the methods of adjusting the vehicle position known from the prior art, the described invention can be advantageously used for such inclined lanes and achieve smooth adjustment of the vehicle position.

[0043] The present invention can also be used to advantage on roadways with a lateral curvature. On such roads, the road surface portion contacting the left front wheel and the road surface portion contacting the right front wheel are not parallel. In countries where vehicles drive on the right, the curved roadway can be constructed so that vehicles traveling on it drift slightly to the right. This prevents the driver from colliding with vehicles traveling from the opposite direction and therefore to their left if they are not paying attention. More generally, the present invention can be used to advantage in all situations where the steering wheel angle obtained by the steering wheel angle sensor does not correspond to the direction the vehicle is actually traveling.

[0044] The vehicle state (ie, the position of the vehicle on the lane 10 and the trajectory of the vehicle) can be represented as a set of physical quantities or state variables, Figure 2 The above section shows these physical quantities or state variables. In particular, the vehicle state can be characterized as:

[0045] - the vehicle's yaw rate dψ, and / or

[0046] - the vehicle's heading angle ψ, and / or

[0047] - the lateral velocity dy of the vehicle, and / or

[0048] - the lateral distance y of the vehicle relative to the reference trajectory, and / or

[0049] - the steering speed dδ of the vehicle's steering wheels, and / or

[0050] - the steering angle δ of the vehicle's steering wheels, and / or

[0051] - The integral ly of the vehicle's lateral distance relative to the reference trajectory.

[0052] The yaw rate dψ is the rotational speed of the vehicle about the axis Z. The yaw rate can be measured, for example, by means of a yaw sensor. The heading angle ψ can be defined as the angle formed between the longitudinal axis X of the roadway 10 (i.e. the axis tangent to the reference trajectory 12 at the height of the vehicle) and the longitudinal axis X1 of the vehicle. The lateral distance y can be defined as the distance between a point on the vehicle (in particular the center of gravity C of the vehicle) and the reference trajectory 12. As a variant, the lateral distance y can also measure the distance of a point on the vehicle from the dividing line 11. The heading angle ψ and the lateral distance y can be calculated, for example, by the detection device 8 of the vehicle. The lateral rate dy is the derivative of the lateral distance y with respect to time. The integral ly of the lateral distance can be calculated with respect to time from an initial moment, which corresponds to the moment when the control method is activated.

[0053] The steering angle δ of the steered wheel can be defined as the angle formed between the axis X2 parallel to the direction of travel of the steered wheel and the longitudinal axis X1 of the vehicle. The steering angle δ is proportional to the steering wheel angle measured by the steering wheel angle sensor 7. It should be noted that when the vehicle turns, the two steered wheels of the vehicle usually turn at slightly different angles in order to take into account the different radii of curvature followed by each steered wheel. Advantageously, the vehicle state can be simplified using a so-called "bicycle" model (i.e., by considering a single steered wheel whose steering angle causes the vehicle to follow the same curve as two steered wheels with different steering angles). When using the "bicycle" model, the steering angle of the steered wheel can be, for example, the average angle between the steering angle of the left steered wheel and the steering angle of the right steered wheel. The steering speed dδ of the steered wheel is the derivative of the steering angle δ with respect to time.

[0054] Using the state variables just described, the state vector X can be defined as follows:

[0055] [Math 1]

[0056]

[0057] Thus, the state vector X comprises seven components. This state vector allows characterizing the position and trajectory of vehicle 1 at a given moment. Alternatively, the components of the state vector may be presented in a different order, or even include different variables instead of or in addition to the variables presented. The following description will then be adjusted accordingly. In general, the state vector can characterize the position, velocity, and acceleration of the vehicle at a given moment. Thus, the state vector is a vector that can change over time.

[0058] refer to Figure 4We will now describe a method for adjusting the lateral position of vehicle 1 on lane 10. This method can be broken down into two main steps E1 and E2. The first step E1 is an initialization step performed upon activation of the adjustment method. In particular, the first step E1 can be performed during the transition of vehicle 1 from manual to autonomous driving mode. The second step E2 is a step for adjusting the position of vehicle 1, performed after initialization step E1. This second step is used to calculate the steering angle setpoint δc for the vehicle's steered wheels. This second step can then be repeated at a given frequency as long as vehicle 1 is autonomously controlled.

[0059] For a clear understanding of the present invention, we will first describe the second step E2 and then the first step E1. Figure 5 Schematically shown in FIG, which shows a closed-loop control system.

[0060] The second step E2 comprises: a step E21 of calculating a reference state vector Xref of the vehicle; a step E22 of calculating an observed state vector Xobs of the vehicle; and a step E23 of calculating a steering angle setpoint δc for the vehicle's steered wheels based on the difference Xerr between the reference state vector Xref and the observed state vector Xobs. The steering angle setpoint δc is then used to calculate the observed state vector Xobs during subsequent iterations of the second step E2.

[0061] During calculation step E21, a reference state vector Xref is calculated. This state vector describes the vehicle's position and the desired vehicle trajectory. This state vector can be calculated in particular based on reference trajectory 12. Since vehicle 1 is subject to various disturbances caused, for example, by factors external to vehicle 1 or factors internal to vehicle 1, at a given moment, the reference state vector Xref may differ from the observed state vector Xobs of the vehicle.

[0062] During the calculation step E22, based on the vehicle's onboard sensors, the vehicle's kinematic model (in Figure 5 21) and the observer (in Figure 5 The kinematic model 21 can be described by the following equations:

[0063] [Math 2]

[0064] dX=A.X+B1.δc+B2.ρ

[0065] In this equation:

[0066] -dX is the time derivative of the state vector X,

[0067] δc is the steering angle setpoint of the steered wheels calculated during the previous iteration of the second step E2,

[0068] -ρ is the curvature of the lane. This value is zero when the lane is straight. To simplify the remainder of this description, it is assumed that the lane is actually straight. However, the control method according to the present invention can be applied to situations where the vehicle is traveling around a curve.

[0069] -A is a 7×7 matrix, which will be defined below,

[0070] - B1 and B2 are vectors with seven components defined below.

[0071] The matrix A can be defined by the following formula:

[0072] [Math 3]

[0073]

[0074] in:

[0075] -cf represents the drift stiffness of the vehicle's front driveline,

[0076] -cr represents the drift stiffness of the vehicle's rear driveline,

[0077] -lf represents the distance between the vehicle's center of gravity C and the front driveline,

[0078] -lr represents the distance between the vehicle's center of gravity C and the rear driveline,

[0079] -lz represents the inertia of the vehicle,

[0080] -m represents the mass of the vehicle,

[0081] -v indicates the speed of the vehicle,

[0082] -ζ represents the damping coefficient of the steering system,

[0083] -ω represents the natural frequency of the steering system filter.

[0084] Therefore, the parameters cf, cr, lf, lr, lz, m, ζ, and ω are constants that characterize the vehicle. These parameters can be defined once during vehicle development and stored in the memory of the electronic control unit.

[0085] Vector B1 can be defined by the following formula:

[0086] [Math 4]

[0087]

[0088] Where ω represents the natural frequency of the steering system filter.

[0089] Vector B2 can be defined by the following formula:

[0090] [Math 5]

[0091]

[0092] Where v represents the speed of the vehicle.

[0093] Furthermore, the measurement vector Y can be defined by the following formula:

[0094] [Math 6]

[0095] Y=CX

[0096] Therein, C denotes a diagonal matrix allowing the isolation of the state variables constituting the vector X, these state variables being directly accessible by means of the sensors on board the vehicle 1. The matrix C is therefore a matrix dependent on the measurements accessible on board the vehicle.

[0097] Based on the kinematic model 21 presented above, the observed state vector Xobs can be calculated with the aid of an observer 22. The observer 22 allows the estimation of the unmeasured components of the observed state vector. In the case of zero curvature of the road, the observed state vector can thus be calculated with the aid of the following equation:

[0098] [Math 7]

[0099] dXobs=(A-Lp.C).Xobs+B1.δ+Lp.Y

[0100] in,

[0101] -Xobs represents the observation state vector,

[0102] -dXobs represents the derivative of the observed state vector with respect to time,

[0103] -Lp is a positive linear matrix that depends on the vehicle speed,

[0104] - A, B1, C and δ are the quantities defined above.

[0105] During the third calculation step E23 , the difference between the reference state vector Xref and the observed state vector Xobs is first calculated. This gives a vector Xerr representing the error between the theoretical state of the vehicle and its actual state. The vector Xerr can thus be defined by the following formula:

[0106] [Math 8]

[0107] Xerr=Xref-Xobs

[0108] The steering angle setpoint of the steered wheel is then deduced from the vector Xerr by multiplying Xerr by the control vector Ks, which is a vector that depends on the vehicle speed. The steering angle setpoint δc of the steered wheel can therefore be calculated using the following formula:

[0109] [Math 9]

[0110] δc=Ks.Xerr+δvir

[0111] where δvir represents the steering angle of the steered wheels necessary to follow the curvature of a possible curve around which the vehicle is traveling. Assuming that the method is performed on a straight lane, the term δvir will therefore be zero.

[0112] It should be noted that the second step E2 uses the vector Ks and the matrix Lp, both of which are defined as functions of vehicle speed. Therefore, the memory of the electronic control unit may contain different values ​​for the vector Ks and the matrix Lp defined for predefined speed values. When the vehicle is actually traveling at an intermediate speed between two predefined speed values, the vector Ks and the matrix Lp can be extrapolated by weighting the stored vector Ks and matrix Lp for speeds higher and lower than the vehicle's actual speed, respectively.

[0113] A first step E1 will now be described during which the regulation method is initialized.

[0114] Initialization step E1 includes a sub-step E11, which calculates the components of the initial observed state vector X0 so that the steering angle setpoint δc of the steered wheel when the control method is activated is equal to the steering angle δm of the steered wheel measured when the control method is activated. The initial observed state vector X0 corresponds to the observed state vector at the moment of transition between manual and autonomous driving modes (i.e., when the control method is initialized). Therefore, specific initial conditions are imposed on the observed state vector in order to achieve a smooth transition between manual and autonomous driving modes. The steering angle δm of the steered wheel is measured using a steering wheel angle sensor by multiplying the steering wheel angle by a scaling factor that depends on the steering device. Even if the measured value returned by the steering wheel angle sensor 7 differs from the actual steering wheel angle (e.g., has an offset relative to the actual value), the steering angle setpoint is equal to this measured value. This prevents abrupt changes in the steering angle of the steered wheel during initialization.

[0115] The initial observation state vector X0 (hereinafter simply referred to as vector X0) corresponds to the observation state vector at the time of initialization step E1. The initial observation state vector can be defined by the following formula:

[0116] [Math 10]

[0117]

[0118] in:

[0119] -X01, X02, X03, X04, X05, X06, X07 represent the first, second, third, fourth, fifth, sixth and seventh components of vector X0 respectively,

[0120] - dψ0 is the yaw rate of the vehicle measured or calculated at the current moment (i.e. when the regulation method is initialized),

[0121] -ψ0 is the heading angle of the vehicle measured or calculated during the initialization step E1,

[0122] - dy0 is the lateral velocity dy of the vehicle measured or calculated at the initialization step E1,

[0123] - y0 is the lateral distance of the vehicle relative to the reference trajectory measured or calculated during the initialization step E1,

[0124] - dδ0 is the steering speed of the vehicle's steering wheels measured or calculated during the initialization step E1,

[0125] - δ0 is the steering angle of the vehicle's steering wheels measured or calculated during the initialization step E1,

[0126] -X07 is the seventh component of vector X0. X07 is calculated according to the following formula:

[0127] [Math 11]

[0128]

[0129] in:

[0130] -Ks,i represents the component i of the adjustment vector Ks,

[0131] -Xerr,i represents the component i of the difference between the reference vector and the vector X0 at the initialization step E1,

[0132] δm is the steering angle of the vehicle's steering wheels measured at the initialization step E1 and obtained using a steering wheel angle sensor,

[0133] - δvir represents the steering angle of the steered wheels necessary to follow the curvature of a possible curve around which the vehicle is traveling.

[0134] The seventh component X07 is therefore calculated as a function of the value of the steering angle δm measured when the control method is activated and also as a function of the control vector Ks.

[0135] During the first iteration of calculation step E2, a steering angle setpoint δc of the steered wheels is obtained by applying an initial value to the seventh component of vector X0, which is equal to the steering angle δm of the steered wheels of the vehicle measured when the control method is activated. Indeed, during the first iteration of calculation step E2, the seventh component of reference vector Xref is zero, and the solution of equation Math 9 defined above therefore yields δc=δm.

[0136] Therefore, the seventh component X07 of the vector X0 (representing the initial value of the integral of the vehicle's lateral distance ly relative to the reference trajectory) is defined so that the setpoint value of the steering angle of the steered wheels upon activation of the control method is equal to the steering angle of the steered wheels measured upon activation of the control method. Consequently, the initial value of the integral of the vehicle's lateral distance relative to the reference trajectory is not defined as a constant value, and in particular, not as a zero value. Instead, this initial value is calculated based on the measured value of the steering angle of the steered wheels upon activation of the control method. Then, during subsequent iterations of calculation step E2, the integral of the lateral distance ly (i.e., the seventh component of the observed state vector Xobs) can deviate from this calculated initial value and be incremented or decremented (depending on whether the vehicle is drifting laterally to the left or right).

[0137] Figure 6 is a graph that allows comparison of trajectory T1 followed by vehicle 1 implementing a control method according to an embodiment of the present invention with trajectory T2 followed by vehicle 1 implementing a control method according to the prior art. To achieve this comparison, a 15° offset is applied to the steering wheel angle values ​​emitted by the steering wheel angle sensor. Consequently, there is a 15° difference between the actual and measured steering wheel angles. This scenario illustrates an extreme case where vehicle 1 is traveling on a particularly inclined or cambered roadway, or where there is a steering wheel angle sensor calibration fault. It should be noted that such a calibration fault can occur, for example, when the vehicle is leaving a factory or garage and the steering wheel angle sensor calibration algorithm has not yet been executed. Assume that the vehicle switches between manual and autonomous driving modes while driving in a straight line, and its actual trajectory already substantially corresponds to reference trajectory 12.

[0138] The graph can be broken down into three sections, P1, P2, and P3, with the abscissa being common to all three sections. This axis represents time in seconds. Assume that the vehicle switches from manual to autonomous driving mode at abscissa 0. Trajectory T2 is interrupted approximately 1.7 seconds after activating the control method because the driver has to interrupt the control method and resume manual control of the vehicle.

[0139] The upper portion P1 of the graph shows the steering wheel angles (expressed in degrees) for trajectory T1 and trajectory T2. It should be noted that trajectory T1 results in a steering wheel angle variation of up to plus or minus five degrees. However, trajectory T2 results in a larger variation in the steering wheel angle. Over a duration of approximately one second, these variations are even greater than 10°. In fact, because the vehicle is already following the reference trajectory I2, the steering angle setpoint calculated by the vehicle according to the prior art will be essentially equal to 0°. Consequently, the abrupt switch in steering wheel angle from a value equal to the offset (in this case, 15°) to a value equal to 0° will have the effect of a sudden, large steering input.

[0140] The middle portion P2 of the graph shows the lateral distance (expressed in meters) of the vehicle 1 relative to the reference trajectory 12 for the trajectories T1 and T2 .

[0141] The lower portion P3 of the graph shows the lateral acceleration of the vehicle 1 for the trajectories T1 and T2 (expressed in m / s2). The lateral acceleration of the vehicle following the trajectory T1 remains less than approximately 0.7 m / s2. The lateral acceleration of the vehicle following the trajectory T2 is approximately 1.5 m / s2.

[0142] When the vehicle follows trajectory T2, passengers experience abrupt steering inputs, which can make them anxious and disrupt their comfort. When the vehicle follows trajectory T1, changes in steering wheel angle and lateral acceleration are gentler. Passenger comfort is improved, and the driver does not feel the need to regain control of vehicle 1.

[0143] The present invention provides a method for adjusting the lateral position that, while active, does not result in any abrupt steering inputs, even when the vehicle 1 is traveling on a sloped or cambered roadway, or when the value provided by the steering wheel angle sensor deviates by a certain offset from the actual value of the steering wheel angle. In all cases, the transition between manual and autonomous driving modes is comfortable and reassuring.

Claims

1. A method for adjusting the lateral position of a vehicle (1) on a roadway (10), characterized in that The adjustment method includes: - a step (E21) of calculating a reference state vector (Xref) of the vehicle (1); - a step (E22) of calculating the observed state vector (Xobs) of the vehicle (1); - a step (E23) of calculating a steering angle setpoint (δc) of the steered wheels (2f) of the vehicle (1) as a function of the difference (Xerr) between the reference state vector (Xref) and the observed state vector (Xobs); The regulation method comprises an initialization step (E1) performed when the regulation method is activated, the initialization step (E1) comprising a sub-step (E11) of calculating the components (X07) of the initial observation state vector (X0) so that the steering angle setpoint (δc) of the steered wheels (2f) when the regulation method is activated is equal to the steering angle value (δm) of the steered wheels (2f) measured when the regulation method is activated, The observed state vector (Xobs) includes all or some of the following components: - the yaw velocity (dψ) of the vehicle (1), and - the heading angle (ψ) of the vehicle (1), and - the lateral velocity (dy) of the vehicle (1), and - the lateral distance (y) of the vehicle (1) relative to the reference trajectory (12), and - the steering speed (dδ) of the steering wheel (2f) of the vehicle (1), and - the steering angle (δ) of the steering wheel (2f) of the vehicle (1), and - the integral (ly) of the lateral distance of the vehicle (1) relative to the reference trajectory (12), wherein the observed state vector (Xobs) comprises a component equal to the integral (ly) of the lateral distance of the vehicle (1) relative to the reference trajectory (12), the initial value of the integral (ly) of the lateral distance defined during the initialization step (E1) being calculated so that the steering angle setpoint (δc) of the steered wheels (2f) when the regulation method is activated is equal to the steering angle value (δm) of the steered wheels measured when the regulation method is activated.

2. The adjustment method according to claim 1, wherein: During the sub-step (E11) of calculating the component (X07) of the initial observed state vector (X0), said component (X07) is calculated as a function of the steering angle value (δm) measured when the regulation method is activated.

3. The adjustment method according to claim 1 or 2, characterized in that: The step (E23) of calculating the steering angle setpoint (δc) consists in multiplying the difference (Xerr) between the reference state vector (Xref) and the observed state vector (Xobs) by a regulation vector (Ks) that depends on the speed (v) of the vehicle (1).

4. The adjustment method according to claim 3, wherein: An initial value of the integral (ly) of the lateral distance of the vehicle (1) relative to the reference trajectory (12) is calculated based on the components of the difference (Xerr) between the reference state vector (Xref) and the observed state vector (Xobs), based on the components of the control vector (Ks), and based on the steering angle value (δm) measured when the control method is activated.

5. The adjustment method according to claim 1 or 2, characterized in that: The calculation of the observed state vector (Xobs) is based on the quantities measured by the on-board sensors (7) of the vehicle (1), the constants characterizing the vehicle (cf, cr, lf, lz, m, ζ, ω), the speed (v) of the vehicle (1) and the setpoint value of the steering angle (δr) of the steering wheel (2f) of the vehicle (1) calculated during the previous iteration of the regulation method. 6 . A computer program product comprising program code instructions stored on a computer-readable medium, the program code instructions being configured to implement the steps of the regulation method according to claim 1 when the program is run on a computer. 7 . A computer-readable data storage medium storing a computer program comprising program code instructions for implementing the regulation method according to claim 1 .

8. A motor vehicle (1), characterized in that The motor vehicle comprises hardware and / or software means for implementing the regulation method as claimed in any one of claims 1 to 5 .

Images