Hybrid deterministic override of probabilistic advanced driving assistance systems (ADAS)

Patent metadata

Field	Value
Publication number	US-11926322-B2
Application number	US-202117398666-A
Country	US
Kind code	B2
Filing date	Aug 10, 2021
Priority date	Aug 10, 2021
Publication date	Mar 12, 2024
Grant date	Mar 12, 2024

Abstract

A hybrid deterministic override to cloud based probabilistic advanced driver assistance systems. Under default driving conditions, an ego vehicle is controlled by a probabilistic controller in a cloud. An overall gap between the ego vehicle and a leading vehicle is divided into an emergency collision gap and a driver specified gap. The vehicle sensors monitor the overall gap. When the gap between the ego vehicle and the leading vehicle is less than or equal to the emergency collision gap, a deterministic controller of the ego vehicle overrides the cloud based probabilistic controller to control the braking and acceleration of the ego vehicle.

First claim

The invention claimed is:

1. A method for controlling an ego vehicle equipped with advanced driver assistance, comprising:
receiving an input of a personalized gap of the ego vehicle;
determining an emergency collision gap between the ego vehicle and a leading vehicle;
when the personalized gap is greater than the emergency collision gap, controlling the ego vehicle by a cloud based probabilistic controller; and
when the emergency collision gap is greater than or equal to the personalized gap, overriding the cloud based probabilistic controller and controlling the ego vehicle by a deterministic controller of the ego vehicle.

2. The method of claim 1 , further comprising:
determining an overall gap from a series of radar and LIDAR measurements made by the ego vehicle;
determining a position of the ego vehicle;
calculating, using at least two successive measurements of the series of radar and LIDAR measurements, a velocity of the leading vehicle and determining, by a deterministic controller of the ego vehicle, the emergency collision gap from the overall gap, a velocity of the ego vehicle, and the velocity of the leading vehicle; and
transmitting the personalized gap, the overall gap, the emergency collision gap, the velocity of the leading vehicle, the velocity of the ego vehicle, and the position of the ego vehicle to the cloud based probabilistic controller.

3. The method of claim 2 , further comprising:
correlating by the probabilistic controller, the personalized gap, the overall gap, the emergency collision gap, the velocity of the leading vehicle, the velocity of the ego vehicle, the position of the ego vehicle, and a position of the leading vehicle with a driver profile and historic driving habits of a driver of the ego vehicle;
generating, by the cloud based probabilistic controller, a desired acceleration based on the correlating;
transmitting, by a cloud-vehicle management communication device of the cloud based probabilistic controller, the desired acceleration to the deterministic controller;
calculating, by an upper level controller of the deterministic controller, a first set of braking control signals and a second set of acceleration control signals based on the desired acceleration;
when the personalized gap is greater than the emergency collision gap, the deterministic controller actuating one or more of:
a hydraulic braking system using the first set of braking control signals, and
an acceleration system using the second set of acceleration control signals; and

when the emergency collision gap is greater than or equal to the personalized gap
overriding the cloud based probabilistic controller by invoking an override function.

4. The method of claim 3 , further comprising:
generating, by the upper level controller, an acceleration command based on the desired acceleration;
transmitting, by the upper level controller, the acceleration command to a longitudinal dynamics controller including a mode switch, an acceleration processor, and a deceleration processor;
receiving, by the longitudinal dynamics controller, a set of feedback signals from the hydraulic braking system and the acceleration system;
calculating an updated emergency collision gap from a new series of radar and LIDAR measurements made by the ego vehicle;
determining, by the longitudinal dynamics controller, whether to continue to invoke, by the mode switch, the override function based on the updated emergency collision gap;
continuing to override the cloud based probabilistic controller when the override function is invoked: and
returning control to the cloud based probabilistic controller when the override function is not invoked.

5. The method of claim 4 , further comprising:
when overriding the cloud based probabilistic controller,
determining, by the upper level controller, a desired longitudinal force needed to control the ego vehicle;
transmitting the desired longitudinal force to a lower level controller with the first set of braking control signals, the second set of acceleration control signals, the velocity of the ego vehicle and the overall gap;
outputting, by the lower level controller, the velocity of the ego vehicle, the overall gap and the first set of braking control signals to the hydraulic braking system; and
outputting, by the lower level controller, the velocity of the ego vehicle, the overall gap and the second set of acceleration control signals to the acceleration system.

6. The method of claim 2 , further comprising:
receiving, by the cloud based probabilistic controller, the personalized gap and the overall gap from the ego vehicle;
retrieving, from a memory of the cloud based probabilistic controller, a stored historical driving data of a driver of the ego vehicle;
applying the personalized gap, the overall gap, the emergency collision gap, the velocity of the leading vehicle and the velocity of the ego vehicle to a machine learning algorithm and generating a personalized acceleration profile based on the personalized gap, the historical driving data, the overall gap, the emergency collision gap, the velocity of the leading vehicle, the position of the ego vehicle, and the velocity of the ego vehicle;
determining, by a personalized adaptive cruise control algorithm a desired acceleration for the ego vehicle based on the personalized acceleration profile and the overall gap, the emergency collision gap, the velocity of the leading vehicle, and the velocity of the ego vehicle;
generating the desired acceleration; and
transmitting the desired acceleration to the deterministic controller.

7. The method of claim 6 , wherein the machine learning algorithm is a Gaussian process algorithm.

8. The method of claim 6 , wherein the machine learning algorithm is a Bayesian inference algorithm.

9. The method of claim 1 , further comprising:
overriding, by the deterministic controller, the cloud based probabilistic controller when one or more conditions exist, the one or more conditions including
the overall gap is less than or equal to the emergency collision gap;
the personalized gap is greater than or equal to the overall gap; and
the emergency collision gap is greater than the personalized gap.

Assignees

• Toyota Eng & Mfg North America

Inventors

• Gupta Rohit
• Wang Yanbing
• Wang Ziran
• Han Kyungtae
• Tiwari Prashant

Classifications

• B60W30/162 Primary

  Speed limiting therefor · CPC title

• B60W50/0098

  Details of control systems ensuring comfort, safety or stability not otherwise provided for · CPC title

• B60W2050/0095

  Automatic control mode change · CPC title

• B60W2420/52

  Operations & Transport · mapped topic

• B60W2520/10

  Longitudinal speed · CPC title