Autonomous Vehicles: Technology and Application

A special issue of Applied Sciences (ISSN 2076-3417). This special issue belongs to the section "Transportation and Future Mobility".

Deadline for manuscript submissions: 30 June 2024 | Viewed by 13262

Special Issue Editors


E-Mail Website
Guest Editor
School of Automobile, Chang’an University, Xi’an 710064, China
Interests: new energy vehicles; autonomous vehicles

E-Mail Website
Guest Editor
School of Intelligent Systems Engineering, Sun Yat-sen University. Room.628, Block 1 of Engineering Building, #66 Gongchang Rd., Shenzhen 518107, China
Interests: new energy vehicles; smart driving; cluster control for unmanned systems application; computer vision and its application technology
Special Issues, Collections and Topics in MDPI journals

Special Issue Information

Dear Colleagues,

We invite submissions to this Special Issue devoted to the technology and application of autonomous vehicles. Driverless driving helps improve traffic safety and efficiency, which is a hot topic in the development of artificial intelligence technology and intelligent transportation technology. Its aim is to drive more safely and reliably, free people's hands, and reduce physical labor. In the process of driving, people are the most uncertain factors, and the autonomous car works according to the computer control set by people, which can greatly reduce the occurrence of accidents, to a certain extent, while freeing hands. In general, driverless technology is a synthesis of many cutting-edge disciplines such as sensors, computers, artificial intelligence, communications, navigation and positioning, pattern recognition, machine vision, and intelligent control. According to the functional modules of driverless cars, the key technologies of driverless cars include environmental awareness, navigation and positioning, path planning, and decision control. In recent years, with the maturity of unmanned driving technology, the application and industrial system of the low-speed unmanned driving market have begun to see a larger scale. Semi-closed places with logistics transportation and personnel transfer have gradually become a typical application scenario where functional low-speed unmanned vehicles can quickly land. The application of unmanned driving will have a disruptive impact on the industry and encourage and promote the progress and technological upgrading of vehicle manufacturers.

For this Special Issue, we invite submissions exploring cutting-edge research and recent advances in the fields of autonomous vehicles. Both theoretical and experimental studies are welcome, as well as comprehensive reviews and survey papers.

Prof. Dr. Yi Han
Prof. Dr. Xiaojun Tan
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Applied Sciences is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • autonomous vehicles
  • new energy vehicles
  • cluster control for unmanned systems application
  • computer vision
  • smart driving

Published Papers (10 papers)

Order results
Result details
Select all
Export citation of selected articles as:

Research

19 pages, 2796 KiB  
Article
Efficient Multi-Source Anonymity for Aggregated Internet of Vehicles Datasets
by Xingmin Lu and Wei Song
Appl. Sci. 2024, 14(8), 3230; https://doi.org/10.3390/app14083230 - 11 Apr 2024
Viewed by 216
Abstract
The widespread use of data makes privacy protection an urgent problem that must be addressed. Anonymity is a traditional technique that is used to protect private information. In multi-source data scenarios, if attackers have background knowledge of the data from one source, they [...] Read more.
The widespread use of data makes privacy protection an urgent problem that must be addressed. Anonymity is a traditional technique that is used to protect private information. In multi-source data scenarios, if attackers have background knowledge of the data from one source, they may obtain accurate quasi-identifier (QI) values for other data sources. By analyzing the aggregated dataset, k-anonymity generalizes all or part of the QI values. Hence, some values remain unchanged. This creates new privacy disclosures for inferring other information about an individual. However, current techniques cannot address this problem. This study explores the additional privacy disclosures of aggregated datasets. We propose a new attack called a multi-source linkability attack. Subsequently, we design multi-source (k,d)-anonymity and multi-source (k,l,d)-diversity models and algorithms to protect the quasi-identifiers and sensitive attributes, respectively. We experimentally evaluate our algorithms on real datasets: that is, the Adult and Census datasets. Our work can better prevent privacy disclosures in multi-source scenarios compared to existing Incognito, Flash, Top-down, and Mondrian algorithms. The experimental results also demonstrate that our algorithms perform well regarding information loss and efficiency. Full article
(This article belongs to the Special Issue Autonomous Vehicles: Technology and Application)
Show Figures

Figure 1

15 pages, 4731 KiB  
Article
Enhanced YOLO Network for Improving the Efficiency of Traffic Sign Detection
by Yang Cui, Dong Guo, Hao Yuan, Hengzhi Gu and Hongbo Tang
Appl. Sci. 2024, 14(2), 555; https://doi.org/10.3390/app14020555 - 08 Jan 2024
Cited by 2 | Viewed by 826
Abstract
One important task for autonomous driving is the precise detection and recognition of road traffic signs. This research focuses on a comprehensive set of 72 distinct traffic signs that are prevalent on urban roads in China, with the goal of developing an enhanced [...] Read more.
One important task for autonomous driving is the precise detection and recognition of road traffic signs. This research focuses on a comprehensive set of 72 distinct traffic signs that are prevalent on urban roads in China, with the goal of developing an enhanced You Only Look Once (YOLO) network model tailored for this specific task. The modifications include the omission of the terminal convolution module and Conv3 (C3) module within the backbone network. Additionally, the 32-fold downsampling is replaced with a 16-fold downsampling, and a feature fusion module with dimensions of 152 × 152 is introduced in the feature layer. To capture a more encompassing context, a novel hybrid space pyramid pooling module, referred to as Hybrid Spatial Pyramid Pooling Fast (H-SPPF), is introduced. Furthermore, a channel attention mechanism is integrated into the framework, combined with three other improved methodologies. Upon evaluation, the enhanced algorithm demonstrates impressive results, achieving a precision rate of 91.72%, a recall rate of 91.77%, and a mean average precision (mAP) of 93.88% at an intersection over union (IoU) threshold of 0.5. Additionally, the method also achieves an mAP of 75.81% for a variety of IoU criteria between 0.5 and 0.95. These achievements are validated on an augmented dataset established for this study. Full article
(This article belongs to the Special Issue Autonomous Vehicles: Technology and Application)
Show Figures

Figure 1

22 pages, 5810 KiB  
Article
Research on Driving Style Recognition of Autonomous Vehicles Based on ACO-BP
by Feng Cheng, Wei Gao and Shuchun Jia
Appl. Sci. 2023, 13(22), 12367; https://doi.org/10.3390/app132212367 - 15 Nov 2023
Cited by 1 | Viewed by 872
Abstract
To enhance the lane-changing safety of autonomous vehicles, it is crucial to accurately identify the driving styles of human drivers in scenarios involving the coexistence of autonomous and human-driven vehicles, aiming to avoid encountering vehicles exhibiting hazardous driving patterns. In this study, based [...] Read more.
To enhance the lane-changing safety of autonomous vehicles, it is crucial to accurately identify the driving styles of human drivers in scenarios involving the coexistence of autonomous and human-driven vehicles, aiming to avoid encountering vehicles exhibiting hazardous driving patterns. In this study, based on the real traffic flow data from the Next Generation Simulation (NGSIM) dataset in the United States, 301 lane-changing vehicles that meet the criteria are selected. Six evaluation parameters are chosen, and principal component analysis (PCA) is employed for dimensionality reduction in the data. The K-means algorithm is then utilized to cluster the driving styles, classifying them into three categories. Finally, ant colony optimization (ACO) of a backpropagation (BP) neural network model was constructed, utilizing the dimensionality reduction results as inputs and the clustering results as outputs for the purpose of driving style recognition. Simulation experiments are conducted using MATLAB Version 9.10 (R2021a) for comparative analysis. The results indicate that the constructed ACO-BP model achieved an overall recognition accuracy of 96.7%, significantly higher than the recognition accuracies of the BP, artificial neural network (ANN), and gradient boosting machine (GBM) models. The ACO-BP model also exhibited the fastest recognition speed among the four models. Moreover, the ACO-BP model shows varied improvements in recognition accuracy for each of the three driving styles, with an increase of 13.7%, 4.4%, and 4.3%, respectively, compared to the BP model. The simulation results validate the high accuracy, real-time capability, and classification effectiveness of this model in driving style recognition, providing new insights for this field. Full article
(This article belongs to the Special Issue Autonomous Vehicles: Technology and Application)
Show Figures

Figure 1

18 pages, 4158 KiB  
Article
Multi-Lidar System Localization and Mapping with Online Calibration
by Fang Wang, Xilong Zhao, Hengzhi Gu, Lida Wang, Siyu Wang and Yi Han
Appl. Sci. 2023, 13(18), 10193; https://doi.org/10.3390/app131810193 - 11 Sep 2023
Viewed by 933
Abstract
Currently, the demand for automobiles is increasing, and daily travel is increasingly reliant on cars. However, accompanying this trend are escalating traffic safety issues. Surveys indicate that most traffic accidents stem from driver errors, both intentional and unintentional. Consequently, within the framework of [...] Read more.
Currently, the demand for automobiles is increasing, and daily travel is increasingly reliant on cars. However, accompanying this trend are escalating traffic safety issues. Surveys indicate that most traffic accidents stem from driver errors, both intentional and unintentional. Consequently, within the framework of vehicular intelligence, intelligent driving uses computer software to assist drivers, thereby reducing the likelihood of road safety incidents and traffic accidents. Lidar, an essential facet of perception technology, plays an important role in vehicle intelligent driving. In real-world driving scenarios, the detection range of a single laser radar is limited. Multiple laser radars can improve the detection range and point density, effectively mitigating state estimation degradation in unstructured environments. This, in turn, enhances the precision and accuracy of synchronous positioning and mapping. Nonetheless, the relationship governing pose transformation between multiple lidars is intricate. Over extended periods, perturbations arising from vibrations, temperature fluctuations, or collisions can compromise the initially converged external parameters. In view of these concerns, this paper introduces a system capable of concurrent multi-lidar positioning and mapping, as well as real-time online external parameter calibration. The method first preprocesses the original measurement data, extracts linear and planar features, and rectifies motion distortion. Subsequently, leveraging degradation factors, the convergence of the multi-lidar external parameters is detected in real time. When deterioration in external parameters is identified, the local map of the main laser radar and the feature point cloud of the auxiliary laser radar are associated to realize online calibration. This is succeeded by frame-to-frame matching according to the converged external parameters, culminating in laser odometer computation. Introducing ground constraints and loop closure detection constraints in the back-end optimization effectuates global estimated pose rectification. Concurrently, the feature point cloud is aligned with the global map, and map update is completed. Finally, experimental validation is conducted on data acquired from Chang’an University to substantiate the system’s online calibration and positioning mapping accuracy, robustness, and real-time performance. Full article
(This article belongs to the Special Issue Autonomous Vehicles: Technology and Application)
Show Figures

Figure 1

21 pages, 3755 KiB  
Article
Path Planning for Autonomous Vehicles in Unknown Dynamic Environment Based on Deep Reinforcement Learning
by Hui Hu, Yuge Wang, Wenjie Tong, Jiao Zhao and Yulei Gu
Appl. Sci. 2023, 13(18), 10056; https://doi.org/10.3390/app131810056 - 06 Sep 2023
Cited by 2 | Viewed by 1466
Abstract
Autonomous vehicles can reduce labor power during cargo transportation, and then improve transportation efficiency, for example, the automated guided vehicle (AGV) in the warehouse can improve the operation efficiency. To overcome the limitations of traditional path planning algorithms in unknown environments, such as [...] Read more.
Autonomous vehicles can reduce labor power during cargo transportation, and then improve transportation efficiency, for example, the automated guided vehicle (AGV) in the warehouse can improve the operation efficiency. To overcome the limitations of traditional path planning algorithms in unknown environments, such as reliance on high-precision maps, lack of generalization ability, and obstacle avoidance capability, this study focuses on investigating the Deep Q-Network and its derivative algorithm to enhance network and algorithm structures. A new algorithm named APF-D3QNPER is proposed, which combines the action output method of artificial potential field (APF) with the Dueling Double Deep Q Network algorithm, and experience sample rewards are considered in the experience playback portion of the traditional Deep Reinforcement Learning (DRL) algorithm, which enhances the convergence ability of the traditional DRL algorithm. A long short-term memory (LSTM) network is added to the state feature extraction network part to improve its adaptability in unknown environments and enhance its spatiotemporal sensitivity to the environment. The APF-D3QNPER algorithm is compared with mainstream deep reinforcement learning algorithms and traditional path planning algorithms using a robot operating system and the Gazebo simulation platform by conducting experiments. The results demonstrate that the APF-D3QNPER algorithm exhibits excellent generalization abilities in the simulation environment, and the convergence speed, the loss value, the path planning time, and the path planning length of the APF-D3QNPER algorithm are all less than for other algorithms in diverse scenarios. Full article
(This article belongs to the Special Issue Autonomous Vehicles: Technology and Application)
Show Figures

Figure 1

15 pages, 7504 KiB  
Article
Estimation of Intelligent Commercial Vehicle Sideslip Angle Based on Steering Torque
by Yafei Li, Yiyong Yang, Xiangyu Wang, Yongtao Zhao and Chengbiao Wang
Appl. Sci. 2023, 13(13), 7974; https://doi.org/10.3390/app13137974 - 07 Jul 2023
Cited by 4 | Viewed by 1127
Abstract
The sideslip angle is crucial for the lateral stability state and stability control of intelligent commercial vehicles. However, sensors that can be used for direct measurements are often complex, expensive, and difficult to install in commercial vehicles. To estimate the vehicle sideslip angle, [...] Read more.
The sideslip angle is crucial for the lateral stability state and stability control of intelligent commercial vehicles. However, sensors that can be used for direct measurements are often complex, expensive, and difficult to install in commercial vehicles. To estimate the vehicle sideslip angle, a state observer derived from the extended Kalman filter (EKF) method is proposed, and the state observer is estimated based on steering torque rather than steering angle. The transfer functions between the sideslip angle–steering torque and sideslip angle–steering angle are established, respectively, and the analysis shows that the steering torque signal has a more rapid and more direct reaction due to the hydraulic pressure in the steering system. Finally, the proposed method is validated using Simulink/TruckSim simulation hardware-in-the-loop bench test, and the results show that the proposed method can accurately reflect the actual state of the sideslip angle with good reliability and effectiveness. Full article
(This article belongs to the Special Issue Autonomous Vehicles: Technology and Application)
Show Figures

Figure 1

24 pages, 2796 KiB  
Article
Wheel Odometry with Deep Learning-Based Error Prediction Model for Vehicle Localization
by Ke He, Haitao Ding, Nan Xu and Konghui Guo
Appl. Sci. 2023, 13(9), 5588; https://doi.org/10.3390/app13095588 - 30 Apr 2023
Cited by 1 | Viewed by 1595
Abstract
Wheel odometry is a simple and low-cost localization technique that can be used for localization in GNSS-deprived environments; however, its measurement accuracy is affected by many factors, such as wheel slip, wear, and tire pressure changes, resulting in unpredictable and variable errors, which [...] Read more.
Wheel odometry is a simple and low-cost localization technique that can be used for localization in GNSS-deprived environments; however, its measurement accuracy is affected by many factors, such as wheel slip, wear, and tire pressure changes, resulting in unpredictable and variable errors, which in turn affect positioning performance. To improve the localization performance of wheel odometry, this study developed a wheel odometry error prediction model based on a transformer neural network to learn the measurement uncertainty of wheel odometry and accurately predict the odometry error. Driving condition characteristics including features describing road types, road conditions, and vehicle driving operations were considered, and models both with and without driving condition characteristics were compared and analyzed. Tests were performed on a public dataset and an experimental vehicle. The experimental results demonstrate that the proposed model can predict the odometry error with higher accuracy, stability, and reliability than the LSTM and WhONet models under multiple challenging and longer GNSS outage driving conditions. At the same time, the transformer model’s overall performance can be improved in longer GNSS outage driving conditions by considering the driving condition characteristics. Tests on the experimental vehicle demonstrate the model’s generalization capability and the improved positioning performance of dead reckoning when using the proposed model. This study explored the possibility of applying a transformer model to wheel odometry and provides a new solution for using deep learning in localization. Full article
(This article belongs to the Special Issue Autonomous Vehicles: Technology and Application)
Show Figures

Figure 1

14 pages, 428 KiB  
Article
Multi-View Joint Learning and BEV Feature-Fusion Network for 3D Object Detection
by Qunming Liu, Xiaodong Li, Xiaofei Zhang, Xiaojun Tan and Bodong Shi
Appl. Sci. 2023, 13(9), 5274; https://doi.org/10.3390/app13095274 - 23 Apr 2023
Viewed by 1834
Abstract
Traditional 3D object detectors use BEV (bird’s eye view) feature maps to generate 3D object proposals, but in a single BEV feature map, there are inevitably the problems of high compression and information loss. To solve this problem, we propose a multi-view joint [...] Read more.
Traditional 3D object detectors use BEV (bird’s eye view) feature maps to generate 3D object proposals, but in a single BEV feature map, there are inevitably the problems of high compression and information loss. To solve this problem, we propose a multi-view joint learning and BEV feature-fusion network. In this network, we mainly propose two fusion modules: the multi-view feature-fusion module and the multi-BEV feature-fusion module. The multi-view feature fusion module performs joint learning from multiple views, fusing features learned from multiple views, and supplementing missing information in the BEV feature map. The multi-BEV feature-fusion module fuses BEV feature map outputs from different feature extractors to further enrich the feature information in the BEV feature map, in order to generate better quality 3D object proposals. We conducted experiments on a widely used KITTI dataset. The results show that our method has significantly improved the detection accuracy of the cyclist category.For cyclist detection tasks at the easy, moderate, and hard levels of the KITTI test dataset, our method improves by 1.57%, 2.03%, and 0.67%, respectively, compared to PV-RCNN. Full article
(This article belongs to the Special Issue Autonomous Vehicles: Technology and Application)
Show Figures

Figure 1

19 pages, 8728 KiB  
Article
Cooperative Decision-Making for Mixed Traffic at an Unsignalized Intersection Based on Multi-Agent Reinforcement Learning
by Huanbiao Zhuang, Chaofan Lei, Yuanhang Chen and Xiaojun Tan
Appl. Sci. 2023, 13(8), 5018; https://doi.org/10.3390/app13085018 - 17 Apr 2023
Cited by 3 | Viewed by 1857
Abstract
Despite rapid advances in vehicle intelligence and connectivity, there is still a significant period in mixed traffic where connected, automated vehicles and human-driven vehicles coexist. The behavioral uncertainty of human-driven vehicles makes decision-making a challenging task in an unsignalized intersection scenario. In this [...] Read more.
Despite rapid advances in vehicle intelligence and connectivity, there is still a significant period in mixed traffic where connected, automated vehicles and human-driven vehicles coexist. The behavioral uncertainty of human-driven vehicles makes decision-making a challenging task in an unsignalized intersection scenario. In this paper, a decentralized multi-agent proximal policy optimization (MAPPO) based on an attention representations algorithm (Attn-MAPPO) was developed to make joint decisions at an intersection to avoid collisions and cross the intersection effectively. To implement this framework, by exploiting the shared information, the system was modeled as a model-free, fully cooperative, multi-agent system. The vehicle employed an attention module to extract the most valuable information from its neighbors. Based on the observation and traffic rules, a joint policy was identified to work more cooperatively based on the trajectory prediction of all the vehicles. To facilitate the collaboration between the vehicles, a weighted reward assignment scheme was proposed to focus more on the vehicles approaching intersections. The results presented the advantages of the Attn-MAPPO framework and validated the effectiveness of the designed reward function. Ultimately, the comparative experiments were conducted to demonstrate that the proposed approach was more adaptive and generalized than the heuristic rule-based model, which revealed its great potential for reinforcement learning in the decision-making of autonomous driving. Full article
(This article belongs to the Special Issue Autonomous Vehicles: Technology and Application)
Show Figures

Figure 1

19 pages, 1834 KiB  
Article
Feature Identification, Solution Disassembly and Cost Comparison of Intelligent Driving under Different Technical Routes
by Zongwei Liu, Wang Zhang, Hong Tan and Fuquan Zhao
Appl. Sci. 2023, 13(7), 4361; https://doi.org/10.3390/app13074361 - 29 Mar 2023
Cited by 2 | Viewed by 1136
Abstract
Technical route decision making of intelligent driving has always been the focus of attention of automotive enterprises and even the industry. Firstly, this study combs the main technical routes of intelligent driving at different levels from three dimensions: development strategy, intelligence allocation and [...] Read more.
Technical route decision making of intelligent driving has always been the focus of attention of automotive enterprises and even the industry. Firstly, this study combs the main technical routes of intelligent driving at different levels from three dimensions: development strategy, intelligence allocation and sensor combination. Then, the methodology of technical component combination is designed to disassemble different technical routes into corresponding technical component combinations. Finally, an improved evaluation model of total cost of ownership of intelligent driving is developed and the total cost of ownership of intelligent driving system under different technical routes is compared. For the development strategy, even if the function superposition can follow some research and development achievements of low-level intelligent driving, scenario-driven is still the option with lower cost and better sustainability. For intelligence allocation, collaborative intelligence can effectively reduce the cost of the vehicle compared with single-vehicle intelligence by up to 46%, but the cost reduction depends on the original on-board hardware. For sensor combination, the multi-source fusion always has the cost advantage compared with vision-only, but the advantage is more obvious in the medium-level and high-level stage of single-vehicle intelligence. Full article
(This article belongs to the Special Issue Autonomous Vehicles: Technology and Application)
Show Figures

Figure 1

Back to TopTop