Page 446 - Automotive Engineering Powertrain Chassis System and Vehicle Body

Decisional architecture C HAPTER 14.2

4. Obtain the vehicle’s location relative to environ- parameters are evaluated using a linear camera (equipped
mental objects at the parking place. If the ‘parked’ with appropriate optical lenses) located in the automatic
location is reached, stop; otherwise, go to step 1. vehicle, and an infrared target located at the rear side of
the leading vehicle (see Section 14.2.4). The position/
When the vehicle A1 moves backwards into the parking
place from the start location shown in Fig. 14.2-15, the orientation parameters are represented by the longitu-
corner A12 (front right corner of the vehicle) must dinal and lateral distances DX and DY between the two
not collide with the corner B24 (front left corner of vehicles, and by the angle Dj between the main axes of
the place). The start location must ensure that the sub- the two vehicles; the velocity parameter is obtained by
sequent motions will be collision-free with objects lim- derivating the position parameters.
iting the parking place. To obtain a convenient start
location, the vehicle has to stop at a distance D3 that will 14.2.3.6.2 Generating the required controls
ensure a desired minimal safety distance D5 between the
vehicle and the nearest corner of the parking place during Following the leading vehicle is performed by controlling,
the subsequent backward motion. The relation between at the servo-loop frequency, the acceleration/decelera-
the distance D1, D2, D3, D4 and D5 is described by tion of the automated vehicle along with the angular
a function F(D1, D2, D3, D4, D5) ¼ 0. velocity of its steering wheel.
This function can not be expressed in closed form, but As for the longitudinal control, the basic idea is to set
it can be estimated for a given type of vehicle by using the a linear relation between the distance and the speed of
model (Equation 14.2.1) when the commands (Equa- the two vehicles:
tions 14.2.9 and 14.2.10) are applied. The computations
are carried out off-line and the results are stored in X X ¼ d min þ hV f (14.2.17)
f
l
a look-up table which is used on-line, to obtain an esti-
mate of D3 corresponding to a desired minimal safety where X l ,X f , and V f are respectively the position of the
distance D5 for given D1, D2 and D4 (Paromtchik and leading vehicle, the position of the following vehicle, and
Laugier, 1996a). When the necessary parking ’depth’ has the velocity of the following vehicle, d min is the mini-
been reached, clearance between the vehicle and the mum distance between the two vehicles, and h is a time
parked ones is provided, i.e. the vehicle moves forwards constant ðd min ¼ 1 m and h ¼ 0:35 s in the reported
or backwards so as to be in the middle of the parking experiments). This approach has led us to make use of
place between the two parked vehicles. the following controller (see Daviet and Parent (1996))
for more details:

14.2.3.6 Platooning A ¼ C v DV þ C p ðDX hV d min Þ (14.2.18)
f
f
The platooning SBM allows the controlled vehicle to where A f is the acceleration of the following vehicle,
automatically follow an other vehicle (this leading vehicle DV ¼ V V ; and DX ¼ X X : the control gains
l
l
f
f
can either have been moved autonomously, or driven by C p and C v . have been chosen as follows: C v ¼ 1=h and
a human driver). This SBM takes as input the current C p ¼ minð1=h; A max =V Þ: The fact that the position gain
f
(velocity, position, orientation) parameters of the vehicle factor is variable allows the controller to take into account
3
to control, and it generates in real-time the required the acceleration saturation and to deal with large initial
lateral and longitudinal controls. The platooning SBM errors (since C p decreases when the speed increases).
operates in two phases (Daviet and Parent, 1996): (1) As for the lateral control we have applied a simple
determining the relative velocity and position/orienta- approach based onto the classical ‘tractor model’. This
tion parameters, and (2) generating the required longi- approach leads the controller to always set the orienta-
tudinal and lateral controls. tion of the steering wheel in a direction parallel to the
orientation of the leading vehicle. This approach generate
stable behaviours, but it leads the following vehicle to
14.2.3.6.1 Determining the state parameters
weakly cut the turns (this might be a problem for con-
The assessment of the velocity and of the position/ori- trolling a platoon of several vehicles in a constrained
entation parameters of the leading vehicle has to be area).
performed at a rate consistent with the servo-loop fre- In a more recent work, we have slightly modified the
quency (50 Hz in practice). In our implementation, these longitudinal and lateral controls in order to avoid the

3 The (velocity, position, orientation) parameters of the following vehicle are computed in real-time from the sensory data; they are expressed
relatively to the leading vehicle reference frame.

453