A simplified dynamic model with driver’s NMS characteristic for human-vehicle shared control of autonomous vehicle

2021 ◽  
pp. 095440702110189
Author(s):  
Huateng Wu ◽  
Hanbing Wei ◽  
Zheng Liu ◽  
Jin Xu
Keyword(s):  
Dynamic Model ◽  
Autonomous Vehicles ◽  
Large Scale ◽  
Autonomous Vehicle ◽  
Lyapunov Stability Theory ◽  
Shared Control ◽  
Steering Wheel ◽  
Adaptive Genetic Algorithm ◽  
Model Parameters ◽  
Control Authority

Since the large-scale application of fully autonomous vehicles is difficult to be commercialized in the short term, human-vehicle shared control (HVSC) is a promising technique. To implement the control authority allocation and observe the driver characteristic, it is essential to develop an efficient HVSC dynamic model with the driver’s neuromuscular characteristic (NMS). To further our previous research, a simplified HVSC dynamic model is proposed in this paper. This model simplifies the non-critical NMS parameters such as muscle spindle feedback, which has no significant feedback effect while retaining essential NMS characteristics such as stretch reflection and intrinsic properties. The model consists of a model predictive controller (MPC) coupled with a driver NMS model and a 2 DOF vehicle model. The stability is proved by Lyapunov stability theory. Moreover, a field experiment was conducted for validation of the model. The V-Box is utilized to measure the vehicle’s state signals, such as steering wheel angle and pedal stroke. Subsequently, the adaptive genetic algorithm (AGA) is employed to identify the model parameters based on the experimental results. The comparison between the experiment and the model output shows that the proposed model can accurately represent the driver’s NMS characteristics and vehicle dynamic parameters. This paper will serve as a theoretical basis for the control authority allocation for L3 class autonomous vehicles.

scholarly journals A Steering-Following Dynamic Model with Driver’s NMS Characteristic for Human-Vehicle Shared Control

2020 ◽  
Vol 10 (7) ◽  
pp. 2626 ◽  
Author(s):  
Hanbing Wei ◽  
Yanhong Wu ◽  
Xing Chen ◽  
Jin Xu
Keyword(s):  
Dynamic Model ◽  
Unscented Kalman Filter ◽  
Autonomous Vehicle ◽  
Least Square ◽  
Experimental Results ◽  
Shared Control ◽  
Dynamic State ◽  
Dynamic Mode ◽  
Identification Algorithm ◽  
Control Authority

For investigating driver characteristic as well as control authority allocation during the process of human–vehicle shared control (HVSC) for an autonomous vehicle (AV), a HVSC dynamic mode with a driver’s neuromuscular (NMS) state parameters was proposed in this paper. It takes into account the driver’s NMS characteristics such as stretch reflection and reflex stiffness. By designing a model predictive control (MPC) controller, the vehicle’s state feedback and driver’s state are incorporated to construct the HVSC dynamic model. For the validation of the model, a field experiment was conducted. The vehicle state signals are collected by V-BOX, and the driver’s state signals are obtained with the electromyography instrument. Subsequently, the hierarchical least square (HLS) parameter identification algorithm was implemented to identify the parameters of the model based on the experimental results. Moreover, the Unscented Kalman Filter (UKF) was utilized to estimate the important NMS parameters which cannot be measured directly. The experimental results showed that the model we proposed has excellent accuracy in characterizing the vehicle’s dynamic state and estimating the driver’s NMS parameter. This paper will serve as a theoretical basis for the new control strategy allocation between human and vehicle for L3 class AVs.

Vision-Based Navigation of Autonomous Vehicles in Roadway Environments with Unexpected Hazards

2019 ◽  
Vol 2673 (12) ◽  
pp. 494-507 ◽  
Author(s):  
Mhafuzul Islam ◽  
Mashrur Chowdhury ◽  
Hongda Li ◽  
Hongxin Hu
Keyword(s):  
Object Detection ◽  
Autonomous Vehicles ◽  
Autonomous Vehicle ◽  
Semantic Segmentation ◽  
Steering Wheel ◽  
Potential Hazard ◽  
Driving System ◽  
Hazardous Object ◽  
Vision Based Navigation ◽  
Navigational System

Vision-based navigation of autonomous vehicles primarily depends on the deep neural network (DNN) based systems in which the controller obtains input from sensors/detectors, such as cameras, and produces a vehicle control output, such as a steering wheel angle to navigate the vehicle safely in a roadway traffic environment. Typically, these DNN-based systems in the autonomous vehicle are trained through supervised learning; however, recent studies show that a trained DNN-based system can be compromised by perturbation or adverse inputs. Similarly, this perturbation can be introduced into the DNN-based systems of autonomous vehicles by unexpected roadway hazards, such as debris or roadblocks. In this study, we first introduce a hazardous roadway environment that can compromise the DNN-based navigational system of an autonomous vehicle, and produce an incorrect steering wheel angle, which could cause crashes resulting in fatality or injury. Then, we develop a DNN-based autonomous vehicle driving system using object detection and semantic segmentation to mitigate the adverse effect of this type of hazard, which helps the autonomous vehicle to navigate safely around such hazards. We find that our developed DNN-based autonomous vehicle driving system, including hazardous object detection and semantic segmentation, improves the navigational ability of an autonomous vehicle to avoid a potential hazard by 21% compared with the traditional DNN-based autonomous vehicle driving system.

YouTube as a Source of Information in Understanding Autonomous Vehicle Consumers: Natural Language Processing Study

2019 ◽  
Vol 2673 (8) ◽  
pp. 242-253 ◽  
Author(s):  
Subasish Das ◽  
Anandi Dutta ◽  
Tomas Lindheimer ◽  
Mohammad Jalayer ◽  
Zachary Elgart
Keyword(s):  
Natural Language Processing ◽  
Natural Language ◽  
Language Processing ◽  
Automotive Industry ◽  
Autonomous Vehicles ◽  
Large Scale ◽  
Keyword Search ◽  
Autonomous Vehicle ◽  
Perception Of Safety ◽  
Automation Level

The automotive industry is currently experiencing a revolution with the advent and deployment of autonomous vehicles. Several countries are conducting large-scale testing of autonomous vehicles on private and even public roads. It is important to examine the attitudes and potential concerns of end users towards autonomous cars before mass deployment. To facilitate the transition to autonomous vehicles, the automotive industry produces many videos on its products and technologies. The largest video sharing website, YouTube.com, hosts many videos on autonomous vehicle technology. Content analysis and text mining of the comments related to the videos with large numbers of views can provide insight about potential end-user feedback. This study examines two questions: first, how do people view autonomous vehicles? Second, what polarities exist regarding (a) content and (b) automation level? The researchers found 107 videos on YouTube using a related keyword search and examined comments on the 15 most-viewed videos, which had a total of 60.9 million views and around 25,000 comments. The videos were manually clustered based on their content and automation level. This study used two natural language processing (NLP) tools to perform knowledge discovery from a bag of approximately seven million words. The key issues in the comment threads were mostly associated with efficiency, performance, trust, comfort, and safety. The perception of safety and risk increased in the textual contents when videos presented full automation level. Sentiment analysis shows mixed sentiments towards autonomous vehicle technologies, however, the positive sentiments were higher than the negative.

Stability of Controlled Road Vehicles: A Preliminary Fundamental Study

2015 ◽  
Author(s):  
Fabio della Rossa ◽  
Massimiliano Gobbi ◽  
Giampiero Mastinu ◽  
Carlo Piccardi ◽  
Giorgio Previati
Keyword(s):  
Autonomous Vehicles ◽  
Autonomous Vehicle ◽  
Preliminary Investigation ◽  
Basin Of Attraction ◽  
Lateral Acceleration ◽  
Steering Wheel ◽  
Vehicle Speed ◽  
Vehicle Model ◽  
The Stability ◽  
Positive Effect

A comparison of the lateral stability behaviour between an autonomous vehicle, a vehicle with driver and a vehicle without driver (fixed steering wheel) is made by introducing a simple mathematical model of a vehicle running on even road. The mechanical model of the vehicle has two degrees of freedom and the related equations of motion contain the nonlinear tyre characteristics. The driver is described by a well-known model proposed in the literature. The autonomous vehicle has a virtual driver (robot) that behaves substantially like a human, but with its proper reaction time and gain. The road vehicle model has been validated. The study of vehicle stability has to be based on bifurcation analysis and a preliminary investigation is proposed here. The accurate computation of steady-state equilibria is crucial to study the stability of the three kinds of vehicles here compared. The stability of the bare vehicle without driver (fixed steering wheel) is studied in a rather complete way referring to a number of combinations of tyre characteristics. The (known) conclusion is that the understeering vehicle is stable at each lateral acceleration level and at each vehicle speed. The additional (partially unknown) conclusion is that the vehicle (model) with degradated tyres may exhibit a huge number of different bifurcations. The driver has many effects on the stability of the vehicle. One positive effect is to eliminate the many possible different equilibria of the bare vehicle and keep active one single equilibrium only. Another positive effect is to broaden the basin of attraction of stable equilibria (at least at relatively low speed). A negative effect is that, even for straight running, the driver seem introducing a subcritical Hopf bifurcation which limits the maximum forward speed of some understeering vehicles (that could run faster with fixed steering wheel). Both the mentioned positive and negative effects appear to be applicable to autonomous vehicles as well. Further studies could be useful to overcome the limitations on the stability of current autonomous vehicles that have been identified in the present research.

Canyon Building Ventilation System Dynamic Model Optimization Study

1993 ◽  
Author(s):  
Franklin F. K. Chen ◽  
B. Ronald Moncrief
Keyword(s):  
Dynamic Model ◽  
Parameter Space ◽  
Large Scale ◽  
Age Factors ◽  
Ventilation System ◽  
Operating Conditions ◽  
Data Availability ◽  
Model Parameters ◽  
Desktop Computer ◽  
Plant Data

Abstract A canyon building houses special nuclear material processing facilities in two canyon like structures, each with approximately a million cubic feet of air space and a hundred thousand hydraulic equivalent feet of ductwork of various cross sections. The canyon ventilation system is a “once through” design with separate supply and exhaust fans, utilizes two large sand filters to remove radionuclide particulate matter, and exhausts through a tall stack. The ventilation equipment is similar to most industrial ventilation systems. However, in a canyon building, nuclear contamination prohibits access to a large portion of the system and therefore limits the kind of plant data possible. The facility investigated is 40 years old and is operating with original or replacement equipment of comparable antiquity. These factors, access and aged equipment, present a challenge in gauging the performance of canyon ventilation, particularly under uncommon operating conditions. The ability to assess canyon ventilation system performance became critical with time, as the system took on additional exhaust loads and aging equipment approached design maximum. Many “What if?” questions, needed to address modernization/safety issues, are difficult to answer without a dynamic model. This paper describes the development, the validation and the utilization of a dynamic model to analyze the capacity of this ventilation system, under many unusual but likely conditions. The development of a ventilation model with volume and hydraulics of this scale is unique. The resultant model resolutions of better than 0.05″wg under normal plant conditions and approximately 0.2″wg under all plant conditions achievable with a desktop computer is a benchmark of the power of micro-computers. The detail planning and the persistent execution of large scale plant experiments under very restrictive conditions not only produced data to validate the model but lent credence to subsequent applications of the model to mission oriented analysis. Modelling methodology adopted a two parameter space approach, rational parameters and irrational parameters. Rational parameters, such as fan age-factors, idle parameters, infiltration areas and tunnel hydraulic parameters are deduced from plant data based on certain hydraulic models. Due to limited accessibility and therefore partial data availability, the identification of irrational model parameters, such as register positions and unidentifiable infiltrations, required unique treatment of the parameter space. These unique parameters were identified by a numerical search strategy to minimize a set of performance indices. With the large number of parameters, this further attests to our strategy in utilizing the computing power of modern micros. Nine irrational parameters at five levels and 12 sets of plant data, counting up to 540 runs, were completely searched over the time span of a long weekend. Some key results, in assessing emergency operation, in evaluating modernization options, are presented to illustrate the functions of the dynamic model.

scholarly journals Weakly Supervised Reinforcement Learning for Autonomous Highway Driving via Virtual Safety Cages

Sensors ◽  
2021 ◽  
Vol 21 (6) ◽  
pp. 2032
Author(s):  
Sampo Kuutti ◽  
Richard Bowden ◽  
Saber Fallah
Keyword(s):  
Reinforcement Learning ◽  
Autonomous Vehicles ◽  
Autonomous Vehicle ◽  
Model Performance ◽  
Model Parameters ◽  
Safe Driving ◽  
Weak Supervision ◽  
Rule Based ◽  
Learning Agent ◽  
Weakly Supervised

The use of neural networks and reinforcement learning has become increasingly popular in autonomous vehicle control. However, the opaqueness of the resulting control policies presents a significant barrier to deploying neural network-based control in autonomous vehicles. In this paper, we present a reinforcement learning based approach to autonomous vehicle longitudinal control, where the rule-based safety cages provide enhanced safety for the vehicle as well as weak supervision to the reinforcement learning agent. By guiding the agent to meaningful states and actions, this weak supervision improves the convergence during training and enhances the safety of the final trained policy. This rule-based supervisory controller has the further advantage of being fully interpretable, thereby enabling traditional validation and verification approaches to ensure the safety of the vehicle. We compare models with and without safety cages, as well as models with optimal and constrained model parameters, and show that the weak supervision consistently improves the safety of exploration, speed of convergence, and model performance. Additionally, we show that when the model parameters are constrained or sub-optimal, the safety cages can enable a model to learn a safe driving policy even when the model could not be trained to drive through reinforcement learning alone.

scholarly journals A Novel Data-Driven Modeling and Control Design Method for Autonomous Vehicles

Energies ◽  
2021 ◽  
Vol 14 (2) ◽  
pp. 517
Author(s):  
Dániel Fényes ◽  
Balázs Németh ◽  
Péter Gáspár
Keyword(s):  
Autonomous Vehicles ◽  
Design Method ◽  
Nonlinear Modeling ◽  
Autonomous Vehicle ◽  
Control Design ◽  
Path Following ◽  
Simulation Software ◽  
Model Parameters ◽  
Modeling And Control ◽  
And Control

This paper presents a novel modeling method for the control design of autonomous vehicle systems. The goal of the method is to provide a control-oriented model in a predefined Linear Parameter Varying (LPV) structure. The scheduling variables of the LPV model through machine-learning-based methods using a big dataset are selected. Moreover, the LPV model parameters through an optimization algorithm are computed, with which accurate fitting on the dataset is achieved. The proposed method is illustrated on the nonlinear modeling of the lateral vehicle dynamics. The resulting LPV-based vehicle model is used for the control design of path following functionality of autonomous vehicles. The effectiveness of the modeling and control design methods through comprehensive simulation examples based on a high-fidelity simulation software are illustrated.

The Kinematic Analysis and Simulation of the Large-Scale Linear Vibration Screen

2012 ◽  
Vol 215-216 ◽  
pp. 726-729 ◽  
Author(s):  
Xin Du
Keyword(s):  
Dynamic Model ◽  
Kinematic Analysis ◽  
Large Scale ◽  
Vertical Direction ◽  
Model Parameters ◽  
Linear Vibration ◽  
Theory Analysis ◽  
Kinematic Theory ◽  
Reasonable Range ◽  
Displacement Equation

The kinematic analysis had carried on large-scale linear vibration screen, which was based on the theory of multi-rigid-body system and the theory of self-synchronous vibration. According to the dynamic analysis, the displacement equation was established. The vibration screen of motion was decomposed into two vertical directions. It included the vertical direction and the horizontal direction. The model of vibration screen is analyzed in the vertical direction. The dynamic model is established and analyzed by COSMOSMotion software. The paper is consisted of three parts.1) Kinematic theory analysis of double motors on the linear vibrating machine. 2) Dynamic model is built. Fixed frame is replaced with simple parts in order that the model is observed later. Through the motion analysis of the vertical direction, it is concluded that the different stiffness of the spring group was corresponding to the moving distance. 3) It is judged that reasonable range of spring parameters and the system damping, and model parameters will be modified

scholarly journals Establishment and tracking control of trapezoidal steering wheel angle model for autonomous vehicles

2020 ◽  
Vol 17 (6) ◽  
pp. 172988142098278
Author(s):  
Haobin Jiang ◽  
Aoxue Li ◽  
Xinchen Zhou ◽  
Yue Yu
Keyword(s):  
Autonomous Vehicles ◽  
Autonomous Vehicle ◽  
Steering System ◽  
Free Play ◽  
Driver Model ◽  
Steering Wheel ◽  
Feature Points ◽  
Steering Control ◽  
Power Steering ◽  
Nonlinear Vehicle Model

Human drivers have rich and diverse driving characteristics on curved roads. Finding the characteristic quantities of the experienced drivers during curve driving and applying them to the steering control of autonomous vehicles is the research goal of this article. We first recruited 10 taxi drivers, 5 bus drivers, and 5 driving instructors as the representatives of experienced drivers and conducted a real car field experiment on six curves with different lengths and curvatures. After processing the collected driving data in the Frenet frame and considering the free play of a real car’s steering system, it was interesting to observe that the shape enclosed by steering wheel angles and the coordinate axis was a trapezoid. Then, we defined four feature points, four feature distances, and one feature steering wheel angle, and the trapezoidal steering wheel angle (TSWA) model was developed by backpropagation neural network with the inputs were vehicle speeds at four feature points, and road curvature and the outputs were feature distances and feature steering wheel angle. The comparisons between TSWA model and experienced drivers, model predictive control, and preview-based driver model showed that the proposed TSWA model can best reflect the steering features of experienced drivers. What is more, the concise expression and human-like characteristic of TSWA model make it easy to realize human-like steering control for autonomous vehicles. Lastly, an autonomous vehicle composed of a nonlinear vehicle model and electric power steering (EPS) system was established in Simulink, the steering wheel angles generated by TSWA model were tracked by EPS motor directly, and the results showed that the EPS system can track the steering angles with high accuracy at different vehicle speeds.

scholarly journals Learning-Based Stochastic Driving Model for Autonomous Vehicle Testing

2021 ◽  
pp. 036119812110357
Author(s):  
Lin Liu ◽  
Shuo Feng ◽  
Yiheng Feng ◽  
Xichan Zhu ◽  
Henry X. Liu
Keyword(s):  
Autonomous Vehicles ◽  
Large Scale ◽  
Short Term Memory ◽  
Autonomous Vehicle ◽  
Deterministic Models ◽  
Pilot Model ◽  
Proposed Model ◽  
Driving Model ◽  
Naturalistic Driving ◽  
Testing And Evaluation

In the simulation-based testing and evaluation of autonomous vehicles (AVs), how background vehicles (BVs) drive directly influences the AV’s driving behavior and further affects the test results. Most existing simulation platforms use either predetermined trajectories or deterministic driving models to model BV behaviors. However, predetermined BV trajectories cannot react to AV maneuvers, and deterministic models are different from real human drivers because of the lack of stochastic components and errors. Both methods lead to unrealistic traffic scenarios. This paper presents a learning-based stochastic driving model that meets the unique needs of AV testing (i.e., interactive and human-like stochasticity). The model is built based on the long short-term memory architecture. By incorporating the concept of quantile regression into the loss function of the model, the stochastic behaviors are reproduced without prior assumption of human drivers. The model is trained with the large-scale naturalistic driving data (NDD) from the Safety Pilot Model Deployment project and compared with a stochastic intelligent driving model (IDM). Analysis of individual trajectories shows that the proposed model can reproduce more similar trajectories of human drivers than IDM. To validate the ability of the proposed model in generating a naturalistic driving environment, traffic simulation experiments are implemented. The results show that traffic flow parameters such as speed, range, and time headway distribution match closely with the NDD, which is of significant importance for AV testing and evaluation.

Sign in / Sign up

Export Citation Format

Share Document