Improving autonomous vehicle in‐traffic safety using learning‐based action governor
نویسندگان
چکیده
Abstract The Action Governor (AG) is a supervisory scheme augmenting nominal control system in order to enhance the system's safety and performance. It acts as an action filter, monitoring commands generated by policy adjusting ones that might lead undesirable behavior. In this article, we present approach based on learning developing AG for autonomous vehicle (AV) decision policies improve their operating mixed‐autonomy traffic (i.e., involving both AVs human‐operated vehicles (HVs)). To achieve this, demonstrate it possible train simulator capable of representing in‐traffic interactions among HVs. We illustrate effectiveness learning‐based improving AV through simulation case studies.
منابع مشابه
Obstacle avoidance for an autonomous vehicle using force field method
This paper presents a force field concept for guiding a vehicle at a high speed maneuver. This method is similar to potential field method. In this paper, motion constrains like vehicles velocity, distance to obstacle and tire conditions and such lane change conditions as zero slop condition and zero lateral acceleration are discussed. After that, possible equations as vehicles path ar...
متن کاملReference Governor Strategies for Vehicle Rollover Avoidance
The paper addresses the problem of vehicle rollover avoidance using reference governors applied to modify the driver steering input in vehicles with an active steering system. Several reference governor designs are presented and tested with a detailed nonlinear simulation model. The vehicle dynamics are highly nonlinear for large steering angles, including the conditions where the vehicle appro...
متن کاملEquity, Safety, and Privacy in the Autonomous Vehicle Era
In 2015, an estimated 38,300 people were killed and 4.4 million injured in motor vehicle crashes in the US alone, with damage costs exceeding $400 billion.1 In contrast, there were no commercial aviation fatalities and 136 civil aviation deaths in the US in 2015.2 More than 90 percent of auto accidents result from human impairment such as drunk driving or road rage, errant pedestrians, or just ...
متن کاملAutonomous Vehicle Control using Image Processing
(ABSTRACT) This thesis describes the design of an inexpensive autonomous vehicle system using a small scaled model vehicle. The system is capable of operating in two different modes: telerobotic manual mode and automated driving mode. In telerobotic manual mode, the model vehicle is controlled by a human driver at a stationary remote control station with full-scale steering wheel and gas pedal....
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Advanced control for applications
سال: 2022
ISSN: ['2578-0727']
DOI: https://doi.org/10.1002/adc2.101