Improved trajectory control method for a motor vehicle

Abstract

A method for controlling the trajectory of a motor vehicle (MV) comprising a vehicle localization system including a single-antenna satellite geolocation unit (ANTE) and an inertial geolocation unit cooperating with a georeferenced high-definition map; said method, implemented by said localization system, consisting of exploiting the rear axle drift angle (RAR) of said vehicle to estimate, in real time, the yaw angle, or heading, of the vehicle throughout the determined trajectory of the vehicle, particularly during turns, using a dynamic model representing the vehicle, of the bicycle model type (BM), in an x, y reference frame, taking into account the position of the satellite geolocation unit's antenna (ANTE) in the vehicle's x, y reference frame and the antenna drift angle (ANTE). (Figure 1)

Description

Title of the invention: Method for improved trajectory control of a motor vehicle

[0001] The present invention relates generally to methods for controlling the trajectory of motor vehicles and more particularly to an improved trajectory control method exploiting a more precise estimation of the heading of a motor vehicle, especially in an automated driving context. The heading of a vehicle is also referred to as the "yaw angle".

[0002] In general, an autonomous vehicle will be defined as a vehicle having an autonomous or automated driving mode.

[0003] The autonomous driving mode is generally managed by a vehicle driving assistance system, designated by AD AS system (Anglo-Saxon acronym for "Advanced Driver Assistance Systems") allowing the automation of certain driving functions usually assigned to the driver.

[0004] The AD AS system will be deliberately confused with the AD AS “supervisor”, which is the body intended to supervise all AD AS functions within the vehicle, in particular the functions relating to autonomous driving mode.

[0005] The level of autonomy of the vehicle considered in this description typically covers levels 2 and above (the five levels of autonomy 0 to 5 are defined by the J3016 standard of the “SAE International”).

[0006] In the context of an automated driving mode relying in part on a georeferenced high-definition (HD) map and a precise absolute positioning system determining the absolute position and orientation of the vehicle, the absolute positioning and orientation information is used to control the vehicle's trajectory to a reference trajectory constructed in the HD map.

[0007] To obtain this absolute positioning and orientation information, a precise geolocation system can be used comprising a satellite geolocation unit (GNSS) associated with an inertial geolocation unit (IMU) based among other things on a satellite geolocation correction technique called real-time kinematic, RTK, (Anglo-Saxon acronym for "Real-Time Kinematic"), or PPK (Anglo-Saxon acronym for "Precise Point Kinematic") and which make it possible to reach in particular precise positions to within 5 cm; this precision being necessary to achieve sufficiently precise trajectory control with regard to the requirements for autonomous driving.

[0008] By "GNSS" (an Anglo-Saxon acronym for "Global Navigation Satellite System"), we mean a geolocation unit based on a constellation of satellites belonging to a GPS-type system (Anglo-Saxon acronym for "Global Positioning System"), Baidu, Galileo, ... and by GNSS receiver: a device on board the vehicle capable of receiving geolocation data by satellite, including a GPS receiver.

[0009] By "IMU" (an acronym for "Inertial Measurement Unit"), we mean an inertial geolocation unit comprising three accelerometers, or a tri-axis accelerometer (longitudinal, lateral, and vertical) and three gyroscopes, or a tri-axis gyroscope (roll, pitch, yaw). Thanks to these sensors (accelerometer and gyroscope), the inertial geolocation unit is capable of measuring the change in orientation but not the absolute orientation.

[0010] Most satellite geolocation units equipping current vehicles use a single antenna (single antenna geolocation), which complicates the accurate estimation of the heading and therefore the absolute orientation of the vehicle.

[0011] In order to control the trajectory of a vehicle and in particular an autonomous driving vehicle, it is common practice, for the estimation of the heading, to require an accuracy of the order of 0.1° at high speed (100km / h), with an accuracy inversely proportional to the speed of the vehicle (therefore for example 1° at 10km / h).

[0012] In [Fig. 1], a VHL vehicle according to the invention is represented by a simple dynamic model, called the bicycle model MDB, in a horizontal plane, assuming that the VHL vehicle is symmetrical about its longitudinal axis XX', that it has a constant longitudinal speed, and that there is no left-to-right or front-to-rear load transfer. Pitch or roll has no effect on this MDB model, which comprises a single front wheel and a single rear wheel, identified respectively by Rav and Rar, with only a front steering system (no rear-wheel steering system).

[0013] The chassis of the VHL vehicle is represented in the bicycle model MDB by the segment AB. Point A is the midpoint of the front axis of the chassis (the front axle and the chassis are here coincident with point A which corresponds to the center of the front wheel Rav), point B is the midpoint of the rear axis of the chassis (the rear axle and the chassis are here coincident with point B which corresponds to the center of the rear wheel Rar), and G is the center of gravity of the VHL vehicle.

[0014] The VHL vehicle is represented in a two-dimensional reference frame x, y, having as its origin the center of gravity G of the VHL vehicle. The longitudinal axis is the "x" axis and the lateral axis is the "y" axis.

[0015] North is also indicated on [Fig. 1] to the right of the MDB model of the VHL vehicle, using the x, y reference frame of the VHL vehicle.

[0016]

[0017]

[0018]

[0019]

[0020]

[0021]

[0022]

[0023]

[0024]

[0025]

[0026]

[0027]

[0028]

[0029]

[0030]

[0031] The position of the ANT antenna of the satellite geolocation unit, relative to the chassis (segment AB) to which it is linked, was represented by the point R with coordinates Xr,Yr in the x, y reference frame of the VHL vehicle. V is the absolute velocity vector of point B, y is the absolute velocity vector of point A and is the velocity vector of point R of the antenna ANT. is the heading angle, or yaw angle, between the "x" axis and North. 7' is the yaw rate of the VHL vehicle. droue is the steering angle of the front wheel Rav. 5ak is the angle that the velocity vector y makes with the longitudinal axis XX' of the VHL vehicle, also designated as the drift angle or rear axle drift (rear wheel Rar). ôG is the angle that the velocity vector y$ of point G makes with the longitudinal axis XX' of the VHL vehicle, also designated as the drift angle or drift of the VHL vehicle at the center of gravity G. 5r is the angle that the velocity vector of point R makes with the longitudinal axis XX' of the VHL vehicle: it is the drift angle at the level of the ANT antenna. Ôav is the angle that the velocity vector y makes with the plane of the front wheels (here, the front wheel Rav), also referred to as the slip angle or front axle slip (front wheel Rav). This angle will not be used in the present invention. If we call Vx and Vy the components of the vector along the x and y axes respectively, and stating that the distance between the center of gravity G and the midpoint of the rear axle (point B) is equal to lr, then we can express the angles ÔG, d^ct ÔR by the following formulas (assuming that the values ​​of these angles are small): yv ^G = ïÿ y^ *ar - yt y^v,. °R - Moreover, classically, the inertial geolocation unit UNI is capable of accurately estimating the "East" and "North" components of the absolute velocity vector y^ of point R, which means that it can accurately estimate the quantity W = 5Æ + ip. The quantity W corresponds to the sum of the drift angle ôR of point R (antenna ANT) with the yaw angle (heading) between the axis "x" and North. If we now make the classic assumption that the rear axle drift angle 5^ is zero, that is to say that the rear axle does not slip laterally "at all" (drift null), then we can deduce that Vy = and by substituting into the expression for ÔR, we obtain:

[0033] We can therefore ultimately extract and estimate the heading according to the following formula:

[0034]

[0035] IL is a quantity directly measured by the receiver of the single-antenna geolocation unit. The quantity IV is typically measured every second.

[0036] V is the yaw rate directly measured by the inertial geolocation unit.

[0037] To estimate the heading (yaw angle), an estimator, generally a computer in the vehicle localization system VHL, calculates the integral of the yaw rate between two "direct" heading measurements, typically every second.

[0038] The heading can therefore be estimated each time the receiver of the single-antenna geolocation unit estimates the quantity IV (i.e., every second). However, taking into account the possible delay of this measurement, and also to estimate the heading between two direct measurements of the quantity W, the integration of the yaw rate is used (an integral that does not have time to drift because it is performed over periods less than one second).

[0039] However, the assumption of "zero drift" at the rear axle is not correct in a turn. Thus, for a typical road vehicle, the drift at the rear axle can quite well be on the order of one degree, even for a turn with relatively low lateral acceleration.

[0040] The heading estimate based on this assumption is therefore not satisfactory for a drift on the order of one degree or even 0.5°.

[0041] To obtain precise control of the trajectory of an autonomously controlled vehicle, it would be necessary to be able to take into account a drift of the order of 0.1°.

[0042] This is why in practice, the heading is estimated according to the previous formula, recalled below, when the lateral acceleration is low, along a straight line, and when the vehicle takes a turn, the heading is estimated only by "integrating" the yaw rate V at the time of taking the turn.

[0044] But this method has its limits, depending on the duration of the turn(s) and the "quality" of the measurement carried out by the inertial geolocation unit in terms of accuracy and noise, because the integration of the yaw rate will also integrate noise, and create an error with exponential evolution.

[0045] To this end, the invention has as its first object a method for controlling the trajectory of a motor vehicle comprising a vehicle localization system

[0046]

[0047] comprising a single-antenna satellite geolocation unit and an inertial geolocation unit cooperating with a georeferenced high-definition map; said location system being capable of providing absolute positioning and orientation information to a trajectory control system enabling the vehicle's trajectory to be controlled in real time to a determined trajectory constructed from the map;said method implemented by said localization system, consisting of exploiting the drift angle 5^ of the rear axle of said vehicle to estimate in real time, the yaw angle, or heading, of the vehicle during the entire determined trajectory of the vehicle, in particular during turns, using a dynamic model representative of the vehicle, of the bicycle model type, in an x, y reference frame, taking into account the position of the antenna of the satellite geolocation unit in the x, y reference frame of the vehicle and the drift angle ÔR of the antenna in said x, y reference frame. ; According to one characteristic, the process consists of estimating the drift angle 5^ of the rear axle, in particular during a turn, at a determined frequency, in particular every second, then estimating the lateral speed Vy of the vehicle considered at the center of gravity of the vehicle in the bicycle-type dynamic model, to deduce the drift angle ÔR at the antenna level and to estimate the yaw angle, or heading, 'P' at the frequency determined from the drift angle ôR of the antenna and a quantity IV measured by the satellite geolocation unit, and the yaw rate ¢, measured at the center of gravity of the vehicle by the inertial geolocation unit. According to another characteristic, the estimated rear axle drift angle 5 is expressed by the following formula: 5^ = .a^ with f ( V ) a known function of the vehicle speed y^ and a^t, the lateral acceleration measured by the unit of inertial geolocation.

[0048] According to one variant, the rear axle drift angle is measured by a sensor on the vehicle.

[0049] According to another characteristic, the Vy component of the velocity vector is estimated. of the vehicle along the y-axis according to the following formula: Vy = + with: the components of the vehicle's velocity vector along the x-axis, lr: the distance between the center of gravity of the vehicle and the middle of the rear axle of the vehicle and 7'; the yaw rate of the vehicle, then, we deduce the drift angle ÔR of the antenna from the following formula: s, and we estimate the yaw angle v at the frequency °R = determined, from the following formula: w , with W: a w quantity corresponding to the sum of the drift angle 5r at the antenna level and the yaw angle V between the x-axis and North.

[0050] The present invention has as its second object a computer program product comprising instructions which, when the program is executed by a computer, lead the latter to implement the steps of the process as described above, for the estimation of the yaw angle, or heading.

[0051] The present invention has as its third object a motor vehicle comprising a vehicle localization system including a single-antenna satellite geolocation unit and an inertial geolocation unit cooperating with a georeferenced high-definition map and a trajectory control system for said vehicle; said localization system further comprising a heading estimator capable of estimating the heading according to the steps of the process as described above.

[0052] According to one feature, the vehicle includes an automated driving mode.

[0053] According to another feature of the vehicle, the trajectory control system belongs to an automated driving mode supervision system based on an AD AS driving assistance system of the vehicle.

[0054] The main advantages of the present invention are to improve the accuracy of heading estimation in a vehicle trajectory control method, in particular an automated driving vehicle, without adding any additional sensor or antenna and therefore without additional cost.

[0055] Other advantages and features of the present invention will become clearer from the following description, given solely by way of non-limiting example and with reference to the drawings in which:

[0056] [Fig-1] already described, illustrates a model of a motor vehicle according to the invention by a bicycle model;

[0057] [Fig.2] illustrates a block diagram of a motor vehicle according to the invention, implementing a method for controlling the vehicle's trajectory, according to the invention; and

[0058] [Fig.3] illustrates in the form of a flowchart, the main steps implemented by the trajectory control method according to the invention for estimating the yaw angle.

[0059] Fig. 2 illustrates, in the form of a block diagram, a VHL motor vehicle implementing the method according to the invention.

[0060] The VHL vehicle includes a vehicle trajectory control system (VTC) belonging, for example, to an autonomous driving supervisor, coupled to, or integrated into, a driver assistance system or AD AS system.

[0061] The vehicle trajectory control system SCT acts on vehicle trajectory control elements VHL, controlling the dynamic behavior of the vehicle VHL, including at least one longitudinal control element OLG and one lateral control element OLT.

[0062] The VHL vehicle further comprises, an SLC vehicle location system providing in real time, absolute positioning and orientation information of the VHL vehicle.

[0063] The SLC location system includes a single-antenna ANT satellite geolocation unit UNS, for example a GPS receiver, comprising a single ANT antenna, and an inertial geolocation unit UNI comprising three accelerometers, or a tri-axis accelerometer (longitudinal, lateral, and vertical) and three gyroscopes, or a tri-axis gyroscope (roll, pitch, yaw), not shown.

[0064] The satellite geolocation unit UNS and inertial geolocation unit UNI cooperate with a georeferenced high-definition map CHD to construct a reference trajectory to which the trajectory of the vehicle VHL must be controlled.

[0065] The SLC localization system further includes a CAP estimator, capable of estimating the yaw angle, or heading, which is provided to the SCT trajectory control system, from the information provided by the UND satellite geolocation units and UNI inertial units.

[0066] Fig. 3 illustrates, in the form of a flowchart, the main steps implemented by the trajectory control method according to the invention for estimating the yaw angle.

[0067] The method consists, in a first step 100, of estimating in real time, the drift angle of the rear axle 5^ in particular during a turn, at a determined frequency, in particular every second, then in a second step 200 of estimating the lateral speed Vy of the vehicle VHL, considered at the center of gravity G of the vehicle VHL in the dynamic bicycle type model MDB, then deducing, in a third step 300, the drift angle 5R at the level of the antenna ANT and in a fourth step 400, of estimating the yaw angle, or heading, , at the frequency determined from the drift angle 5R of the antenna ANT and a quantity W measured by the satellite geolocation unit UNS, and the yaw rate ¢, measured at the center of gravity G of the vehicle VHL by the inertial navigation unit UNI.

[0068] The present invention also implements a computer program product comprising instructions which, when the program is executed by a computer, lead the latter to implement the steps of the process described above for estimating the yaw angle, or heading.

Claims

Demands

1. A method for controlling the trajectory of a motor vehicle (MV) comprising a vehicle (MV) location system (LSS) comprising a single-antenna satellite geolocation unit (SNU) and an inertial geolocation unit (ISU) cooperating with a georeferenced high-definition map (HDM); said location system (LSS) being capable of providing absolute positioning and orientation information to a trajectory control system (TCS) enabling the real-time control of the vehicle's (MV) trajectory to a determined trajectory constructed from the map (HDM);said method implemented by said localization system (SLC), consisting of exploiting the drift angle 5AR of the rear axle (Rar) of said vehicle (VHL) to estimate in real time, the yaw angle, or heading, of the vehicle (VHL) during the entire determined trajectory of the vehicle (VHL), in particular during turns, using a dynamic model representative of the vehicle (VHL), of bicycle model type (MDB), in an x, y reference frame, taking into account the position of the antenna (ANT) of the satellite geolocation unit in the x, y reference frame of the vehicle (VHL) and the drift angle Ôr of the antenna (ANT) in said x, y reference frame.;

2. A method according to the preceding claim, consisting of estimating (100) the rear axle drift angle (Rar), in particular during a turn, at a determined frequency, in particular every second, and then estimating (200) the lateral velocity Vy of the vehicle (VHL) considered at the center of gravity (G) of the vehicle (VHL) in the bicycle-type dynamic model (MDB), to deduce (300) the drift angle Ôr at the antenna (ANT) and estimating (400) the yaw angle, or heading, 4L at the frequency determined from the drift angle ÔR of the antenna (ANT) and a quantity IV measured by the satellite geolocation unit (SNU), and the yaw rate ¢, measured at the center of gravity (G) of the vehicle (VHL) by the inertial geolocation unit (INU).

3. A method according to the preceding claim, wherein the estimated rear axle drift angle θ (Rar) (100) is expressed by the following formula: 5AR = f(V)alat, with f(V) a known function of the vehicle speed y^ and ai«t, the lateral acceleration measured by the inertial geolocation unit (UNI).

4. Method according to claim 2, wherein the drift angle <5^ of the rear axle (Rar) is measured (100) by a sensor of the vehicle (VHL).

5. A method according to claims 2 and 3 or 2 and 4, wherein (200) the component Vy of the velocity vector y\ of the vehicle (VHL) along the y-axis is estimated according to the following formula: Vj = + with Vx ; the component of the velocity vector y^ of the vehicle along the x-axis, lr : the distance between the center of gravity (G) of the vehicle (VHL) and the middle of the rear axle (Rar) of the vehicle (VHL) and V' : the yaw rate of the vehicle (VHL), then (300) the drift angle ôR of the antenna (ANT) is deduced from the following formula: o ~ . and we estimate (400) the yaw angle °R = V^-Y^p at the determined frequency, from the following formula: _ w KvGAit+ilr+Xrï-w, with W: a quantity corresponding to the v- — ” Vx-Yrïp sum of the drift angle 8r at the level of the antenna (ANT) and the yaw angle 7' between the x axis and North.

6. Product computer program comprising instructions which, when the program is executed by a computer, cause the computer to carry out the steps of the process according to any one of the preceding claims for estimating the yaw angle, or heading.

7. Motor vehicle (MV) comprising a vehicle (MV) location system (LSS) comprising a single-antenna satellite geolocation unit (SNU) and an inertial geolocation unit (ISU) cooperating with a georeferenced high-definition map (HDM) and a trajectory control system (TCS) of said vehicle (MV); said location system (LSS) further comprising a heading estimator (CA) capable of estimating the heading according to the steps of the method according to any one of claims 1 to 5.

8. Motor vehicle (MV) according to the preceding claim, comprising an automated driving mode.

9. Motor vehicle (MV) according to the preceding claim, wherein the trajectory control system (TCS) belongs to a automated driving mode supervision system based on a vehicle AD AS (VHL) driver assistance system.

Images