US 12,735,053 B2 · Maneuver prediction based speed profile adaptation for behavior based automated cruise control

Abstract

A system for maneuver prediction based speed profile adaptation for behavior based automated cruise control in a host vehicle includes sensors and actuators in communication with a controller. The controller executing a maneuver prediction speed profile adaptation (MPSPA) application that captures static and dynamic information about the host vehicle and about the host vehicle's environment. The MPSA application processes sensor information and performs: feature extraction, synchronization, normalization, generates a probabilistic path prediction, and detects an intent of a vehicle operator. The MPSA Application generates a control command to the actuators to selectively alter at least a longitudinal velocity of the host vehicle to approximate vehicle operator historical preferences for lateral and longitudinal velocities, yaw and vehicle operator observed forces while the host vehicle is being driven in a fully autonomous and/or a semi-autonomous mode.

A system for maneuver prediction based speed profile adaptation for behavior based automated cruise control in a host vehicle comprises:

a host vehicle;
one or more sensors, the one or more sensors detecting static and dynamic state information about the host vehicle;
one or more actuators, the one or more actuators altering static and dynamic performance of the host vehicle;
a controller having a processor, a non-transitory memory, and input/output (I/O) ports, the I/O ports in communication with the one or more sensors and the one or more actuators, the processor executing program code portions stored in the non-transitory memory, the program code portions including a maneuver prediction speed profile adaptation (MPSPA) application comprising:
a first control logic for capturing static and dynamic information about the host vehicle and about an environment surrounding the host vehicle;
a second control logic for processing the static and dynamic information about the host vehicle and about the environment surrounding the host vehicle, wherein the processing includes at least:
feature extraction, synchronization, normalization, and generating a probabilistic path prediction;
wherein the second control logic further comprises:
control logic for filtering and performing feature extraction on data from the one or more sensors;
wherein filtering causes the system and MPSPA application to reduce computational resource utilization from a first level to a second level less than the first by selecting only certain portions of the data from the one or more sensors that may indicate a host vehicle state change, and wherein the feature extraction further determines that the host vehicle state change is of a magnitude calling for additional processing;
control logic for synchronizing the data from the one or more sensors, wherein synchronizing the data causes the data from the one or more sensors to be accurately aligned and associated in time, so that data from each of the one or more sensors is correlated in time with data from each of the other sensors of the one or more sensors;
control logic for generating a probabilistic path prediction utilizing a probability function for each state variable, following a sigmoid function where:
host vehicle state variables include:
host vehicle speed, vx, host vehicle steering angle, δ, torsion bar torque, τ, and torsion bar torque rate, {dot over (τ)}:
P
M
,
x
(
k
)
=
1
1
+
e
-
β
M
,
x
(
❘
"\[LeftBracketingBar]"
x
❘
"\[RightBracketingBar]"
-
∖
α
M
,
x
)
Eq
.
1
where x is a variable, vx, δ, τ, {dot over (τ)}, and M is maneuver type including turn, lane change, u-turn, evasive steering, and/or lane following, and wherein a maneuver probability is obtained by multiplication of state probability functions as follows:
P
M
(
k
)
=
g
d
⁢
∏
x
=
1
n
〚
P
M
,
x
(
k
〛
)
Eq
.
2
where PM,x:
is a probability function of variable x for maneuver M;
k is a sampling time, βM, x is a tuning parameter, sigmoid steepness of variable x for maneuver M;
αM, x is a tuning parameter, sigmoid midpoint of variable x for maneuver M;
n is a number or quantity of variables;
and gd is a direction of the maneuver M such that for a right side maneuver is gd=−1, and for a left side maneuver gd=1;
a third control logic for detecting an intent of a vehicle operator;
and
a fourth control logic for generating a control command to the one or more actuators, wherein the control command selectively alters at least a longitudinal velocity of the host vehicle to approximate vehicle operator historical preferences for lateral and longitudinal velocities, yaw and vehicle operator observed forces while the host vehicle is being driven in a fully autonomous and/or a semi-autonomous mode.

80 more elements

Maneuver prediction based speed profile adaptation for behavior based automated cruise control

Life of this patent

Maneuver prediction based speed profile adaptation for behavior based automated cruise control

Life of this patent

Abstract

Claims14