scholarly journals Feasible Trajectories Generation for Autonomous Driving Vehicles

2021 ◽  
Vol 11 (23) ◽  
pp. 11143
Author(s):  
Trieu Minh Vu ◽  
Reza Moezzi ◽  
Jindrich Cyrus ◽  
Jaroslav Hlava ◽  
Michal Petru
Keyword(s):  
Tracking Error ◽  
Autonomous Driving ◽  
Angular Speed ◽  
Symmetric Polynomial ◽  
Vehicle Speed ◽  
Steering Angle ◽  
Physical Constraints ◽  
Vehicle Trajectories ◽  
Quintic Polynomial ◽  
Quartic Polynomial

This study presents smooth and fast feasible trajectory generation for autonomous driving vehicles subject to the vehicle physical constraints on the vehicle power, speed, acceleration as well as the hard limitations of the vehicle steering angle and the steering angular speed. This is due to the fact the vehicle speed and the vehicle steering angle are always in a strict relationship for safety purposes, depending on the real vehicle driving constraints, the environmental conditions, and the surrounding obstacles. Three different methods of the position quintic polynomial, speed quartic polynomial, and symmetric polynomial function for generating the vehicle trajectories are presented and illustrated with simulations. The optimal trajectory is selected according to three criteria: Smoother curve, smaller tracking error, and shorter distance. The outcomes of this paper can be used for generating online trajectories for autonomous driving vehicles and auto-parking systems.

ACDR: Autonomous-Car Drive Recorder

2020 ◽  
Vol 32 (3) ◽  
pp. 634-637
Author(s):  
Zhi Wang ◽  
Daishi Watabe ◽  
Hideyasu Sai ◽  
Yukimichi Saito ◽  
Masayoshi Wada ◽  
...  
Keyword(s):  
Autonomous Driving ◽  
Vehicle Speed ◽  
Steering Angle ◽  
Accident Investigation ◽  
Log Files ◽  
Autonomous Cars ◽  
Operational Tests ◽  
Log File

Herein, a drive recorder system for autonomous cars that can be used to analyze accidents that occur during autonomous driving is proposed. In many field operational tests of autonomous cars, vehicle speed, steering angle, and camera data are recorded in the log file. Only relying on these log files without simultaneously monitoring the operator or driver’s actions results in much information being lost that can be used for accident investigation. To solve this problem, an autonomous-car drive recorder that can respond appropriately when accidents occur and minimize the possibility of accidents is proposed. In the system, during an accident, the environment detected around the vehicle and the actions of the driver and operator are recorded, and the data are used to troubleshoot the accident occurrence.

scholarly journals An NN-Based Double Parallel Longitudinal and Lateral Driving Strategy for Self-Driving Transport Vehicles in Structured Road Scenarios

2021 ◽  
Vol 13 (8) ◽  
pp. 4531
Author(s):  
Huiyuan Xiong ◽  
Huan Liu ◽  
Jian Ma ◽  
Yuelong Pan ◽  
Ronghui Zhang
Keyword(s):  
Visual Information ◽  
Short Term Memory ◽  
Absolute Error ◽  
Autonomous Driving ◽  
Optimization Method ◽  
Steering Angle ◽  
Physical Constraints ◽  
Speed Prediction ◽  
Competing Models ◽  
Parallel Network

Studies on self-driving transport vehicles have focused on longitudinal and lateral driving strategies in automated structured road scenarios. In this study, a double parallel network (DP-Net) combined with longitudinal and lateral strategy networks is constructed for self-driving transport vehicles in structured road scenarios, which is based on a convolutional neural network (CNN) and a long short-term memory network (LSTM). First, in feature extraction and perception, a preprocessing module is introduced that can ensure the effective extraction of visual information under complex illumination. Then, a parallel CNN sub-network is designed that is based on multifeature fusion to ensure better autonomous driving strategies. Meanwhile, a parallel LSTM sub-network is designed, which uses vehicle kinematic features as physical constraints to improve the prediction accuracy for steering angle and speed. The Udacity Challenge II dataset is used as the training set with the proposed DP-Net input requirements. Finally, for the proposed DP-Net, the root mean square error (RMSE) is used as the loss function, the mean absolute error (MAE) is used as the metric, and Adam is used as the optimization method. Compared with competing models such as PilotNet, CgNet, and E2E multimodal multitask network, the proposed DP-Net is more robust in handling complex illumination. The RMSE and MAE values for predicting the steering angle of the E2E multimodal multitask network are 0.0584 and 0.0163 rad, respectively; for the proposed DP-Net, those values are 0.0107 and 0.0054 rad, i.e., 81.7% and 66.9% lower, respectively. In addition, the proposed DP-Net also has higher accuracy in speed prediction. Upon testing the collected SYSU Campus dataset, good predictions are also obtained. These results should provide significant guidance for using a DP-Net to deploy multi-axle transport vehicles.

scholarly journals Research on the fuel consumption conservation potential of ADAS on passenger cars

2021 ◽  
Vol 268 ◽  
pp. 01035
Author(s):  
Guogang Qian ◽  
Tieqiang Fu ◽  
Long Sun
Keyword(s):  
Traffic Management ◽  
Green Light ◽  
Autonomous Driving ◽  
Energy Utilization ◽  
Active Management ◽  
Vehicle Speed ◽  
Fuel Saving ◽  
Passenger Cars ◽  
The Road ◽  
Network Connection

Under the trend of automobile electrification, network connection, and intelligence, EU and USA have carried out fuel-saving research and initiatives on ADAS and CAV. The eCoMove project has aimed at economically optimal driving control and traffic management; MAVEN discusses the technical path of GLOSA (Green Light Optimal Speed Advisory) and ecological auto-driving EAD (Eco-Autonomous Driving) by smoothing the vehicle speed. The American NEXTCAR project contains multiple projects. When supplemented with DSF (Dynamic Skip Fire) and 48V technology, the road test led by Ohio State University resulted in a 15% fuel saving rate. Platoon and optimizing intersection signal lights can offer vehicles a more fuel-efficient condition; slope energy utilization, HEV SOC active management, cold storage evaporator, coasting, 48V and mDSF (miller cycle Dynamic Skip Fire) fuel-saving potential has been fully utilized.

A Model Predictive Control Strategy for Lateral and Longitudinal Dynamics in Autonomous Driving

2020 ◽  
Author(s):  
Irfan Khan ◽  
Stefano Feraco ◽  
Angelo Bonfitto ◽  
Nicola Amati
Keyword(s):  
Model Predictive Control ◽  
Predictive Control ◽  
Control Strategy ◽  
Autonomous Driving ◽  
Maximum Speed ◽  
Reference Trajectory ◽  
Vehicle Speed ◽  
The Road ◽  
Speed Regulation ◽  
Road Geometry

Abstract This paper presents a controller dedicated to the lateral and longitudinal vehicle dynamics control for autonomous driving. The proposed strategy exploits a Model Predictive Control strategy to perform lateral guidance and speed regulation. To this end, the algorithm controls the steering angle and the throttle and brake pedals for minimizing the vehicle’s lateral deviation and relative yaw angle with respect to the reference trajectory, while the vehicle speed is controlled to drive at the maximum acceptable longitudinal speed considering the adherence and legal speed limits. The technique exploits data computed by a simulated camera mounted on the top of the vehicle while moving in different driving scenarios. The longitudinal control strategy is based on a reference speed generator, which computes the maximum speed considering the road geometry and lateral motion of the vehicle at the same time. The proposed controller is tested in highway, interurban and urban driving scenarios to check the performance of the proposed method in different driving environments.

Guaranteeing Lanekeeping Performance With Tire Saturation Using Computed Polynomial Lyapunov Functions

2005 ◽  
Author(s):  
Joshua P. Switkes ◽  
J. Christian Gerdes
Keyword(s):  
Lyapunov Function ◽  
Lyapunov Functions ◽  
Vehicle Speed ◽  
The Road ◽  
Vehicle Trajectories ◽  
Lane Departure ◽  
Polynomial Fit ◽  
Tire Forces ◽  
Polynomial Dynamics ◽  
Bounded Disturbance

Lanekeeping assistance systems hold the promise to save thousands of lives every year by preventing unintended lane departure. The potential field lanekeeping assistance system assists the driver in the lanekeeping task by effectively placing the vehicle in an artificial potential well with minimum at lane center. Previous work mathematically guarantees the performance of the system in the linear region of tire forces, but no guarantees of performance or even stability exist for saturating tires. These guarantees are crucial to ensure safety when the vehicle speed is too high for a given turn or the friction coefficient of the road is low due to surface conditions. Here we explore ways to numerically find Lyapunov functions for a vehicle with lanekeeping assistance and realistic tires. First the nonlinearity is modeled as a sector bounded disturbance, and a Lyapunov function is found for all vehicle trajectories that fit this sector bounded disturbance. Next, a polynomial fit is performed on the HSRI tire model, and a Lyapunov function is found for these polynomial dynamics. Each of these approaches provide Lyapunov functions valid well into the nonlinear region.

The Effects of Tire Nonlinearity on Vehicle Steering Stability at High Speed

2014 ◽  
Vol 505-506 ◽  
pp. 301-309
Author(s):  
Hua Dong Xu
Keyword(s):  
High Speed ◽  
Necessary Conditions ◽  
Front Wheel ◽  
Vehicle Safety ◽  
Slip Angle ◽  
Vehicle Speed ◽  
Tire Model ◽  
Dynamics Model ◽  
Steering Angle ◽  
Yaw Rate

The steering stability of a vehicle at high speed is the urgent key problem to be solved of automobile independent development. And it is also the premise and one of the necessary conditions of vehicle safety. Considering of the effects of tire nonlinearity, a 4-DOF dynamics model for a vehicle is established. The yaw rate responses, side slip angle, carriage roll angle and front wheel steering angle with different vehicle speeds are calculated. The calculated values are then compared with the values without considering of the effects of tire nonlinearity. The simulations results show that the vehicle responses can be reflected accurately by using nonlinear tire model. With the bigger vehicle speed, the effects of tire nonlinearity on vehicle high-speed steering stability become more obvious.

scholarly journals Vehicle trajectory prediction and generation using LSTM models and GANs

PLoS ONE ◽  
2021 ◽  
Vol 16 (7) ◽  
pp. e0253868
Author(s):  
Luca Rossi ◽  
Andrea Ajmar ◽  
Marina Paolanti ◽  
Roberto Pierdicca
Keyword(s):  
Traffic Congestion ◽  
Short Term Memory ◽  
Autonomous Driving ◽  
Generative Models ◽  
Trajectory Prediction ◽  
Generative Adversarial Network ◽  
Displacement Error ◽  
Adversarial Network ◽  
Vehicle Trajectories ◽  
Vehicle Trajectory

Vehicles’ trajectory prediction is a topic with growing interest in recent years, as there are applications in several domains ranging from autonomous driving to traffic congestion prediction and urban planning. Predicting trajectories starting from Floating Car Data (FCD) is a complex task that comes with different challenges, namely Vehicle to Infrastructure (V2I) interaction, Vehicle to Vehicle (V2V) interaction, multimodality, and generalizability. These challenges, especially, have not been completely explored by state-of-the-art works. In particular, multimodality and generalizability have been neglected the most, and this work attempts to fill this gap by proposing and defining new datasets, metrics, and methods to help understand and predict vehicle trajectories. We propose and compare Deep Learning models based on Long Short-Term Memory and Generative Adversarial Network architectures; in particular, our GAN-3 model can be used to generate multiple predictions in multimodal scenarios. These approaches are evaluated with our newly proposed error metrics N-ADE and N-FDE, which normalize some biases in the standard Average Displacement Error (ADE) and Final Displacement Error (FDE) metrics. Experiments have been conducted using newly collected datasets in four large Italian cities (Rome, Milan, Naples, and Turin), considering different trajectory lengths to analyze error growth over a larger number of time-steps. The results prove that, although LSTM-based models are superior in unimodal scenarios, generative models perform best in those where the effects of multimodality are higher. Space-time and geographical analysis are performed, to prove the suitability of the proposed methodology for real cases and management services.

scholarly journals Optimal Geno-Fuzzy Lateral Control of Powered Parachute Flying Vehicles

Aerospace ◽  
2021 ◽  
Vol 8 (12) ◽  
pp. 400
Author(s):  
Hanafy M. Omar
Keyword(s):  
Vehicle Dynamics ◽  
Degrees Of Freedom ◽  
Fuzzy Logic Controller ◽  
Optimization Technique ◽  
Maximum Error ◽  
Steering Angle ◽  
Physical Constraints ◽  
Six Degrees Of Freedom ◽  
Fuzzy Membership Functions ◽  
Yaw Angle

In this work, we propose a systematic procedure to design a fuzzy logic controller (FLC) to control the lateral motion of powered parachute (PPC) flying vehicles. The design process does not require knowing the details of vehicle dynamics. Moreover, the physical constraints of the system, such as the maximum error of the yaw angle and the maximum allowed steering angle, are naturally included in the designed controller. The effectiveness of the proposed controller was assessed using the nonlinear six degrees of freedom (6DOF) mathematical model of the PPC. The genetic algorithm (GA) optimization technique was used to optimize the distribution of the fuzzy membership functions in order to improve the performance of the suggested controller. The robustness of the proposed controller was evaluated by changing the values of the parafoil aerodynamic coefficients and the initial flight conditions.

End-to-End Multi-Task Machine Learning of Vehicle Dynamics for Steering Angle Prediction for Autonomous Driving

2019 ◽  
Author(s):  
Nicholas Merrill ◽  
Azim Eskandarian
Keyword(s):  
Network Architecture ◽  
Autonomous Driving ◽  
Sensor Data ◽  
Dynamic Information ◽  
Steering Angle ◽  
Robust Algorithms ◽  
Error Accumulation ◽  
End To End ◽  
Traditional Approaches ◽  
Quality Maps

Abstract The traditional approaches to autonomous, vision-based vehicle systems are limited by their dependency on robust algorithms, sensor fusion, detailed scene construction, and high-quality maps. End-to-end models offer a means of circumventing these limitations by directly mapping an image input to a steering angle output for lateral control. Existing end-to-end models, however, either fail to capture temporally dynamic information or rely on computationally expensive Recurrent Neural Networks (RNN), which are prone to error accumulation via feedback. This paper proposes a Multi-Task Learning (MTL) network architecture that leverages available dynamic sensor data as a target for auxiliary tasks. This method improves steering angle prediction by facilitating the extraction of temporal dependencies from sequentially stacked image inputs. Evaluations performed on the publicly available Comma.ai dataset show a 28.6% improvement in steering angle prediction over existing end-to-end methods.

Sign in / Sign up

Export Citation Format

Share Document