Continuous-scale trusted safety efficiency evaluation is crucial for the agile development and robust validation of autonomous vehicle intelligence. While the UN R157 Regulation evaluates automated lane-keeping system (ALKS) performance baselines through safe collision plots (SCPs) in various scenario clusters, quantifying the specific ALKS safety efficiency remains challenging. We propose a spectrum quantification approach to evaluate the safety efficiency of autonomous vehicles in cut-in scenarios. First, we collected speed-distance data under different cut-in scenarios and extracted essential spectral features to indicate the vehicle motion parameters during the cut-in process. Second, by utilizing Fourier analysis, a spectral analysis model was built to quantify and analyze the vehicle motion characteristics, providing insights into scenario safety. Finally, we created approximate analytical equations for the normalized disturbance frequencies in the nonlinear response scenarios of autonomous driving systems by combining the SCP with a frequency spectrum analysis model. The results showed that the normalized disturbance frequency in the cut-in scenario was approximately 0.2. When the relative longitudinal distance and speed of the vehicle are the same, if the cut-in speed of the cut-in vehicle is larger, the normalized disturbance frequency is higher, indicating that the cut-in process of the autonomous vehicle is more dangerous and may trigger a collision.
{"title":"Spectrum Quantification-Based Safety Efficiency Evaluation of Autonomous Vehicle Under Random Cut-in Scenarios","authors":"Jiang Chen;Weiwei Zhang;Miao Liu;Xiaolan Wang;Jun Gong;Jun Li;Boqi Li;Jiejie Xu","doi":"10.26599/JICV.2023.9210035","DOIUrl":"https://doi.org/10.26599/JICV.2023.9210035","url":null,"abstract":"Continuous-scale trusted safety efficiency evaluation is crucial for the agile development and robust validation of autonomous vehicle intelligence. While the UN R157 Regulation evaluates automated lane-keeping system (ALKS) performance baselines through safe collision plots (SCPs) in various scenario clusters, quantifying the specific ALKS safety efficiency remains challenging. We propose a spectrum quantification approach to evaluate the safety efficiency of autonomous vehicles in cut-in scenarios. First, we collected speed-distance data under different cut-in scenarios and extracted essential spectral features to indicate the vehicle motion parameters during the cut-in process. Second, by utilizing Fourier analysis, a spectral analysis model was built to quantify and analyze the vehicle motion characteristics, providing insights into scenario safety. Finally, we created approximate analytical equations for the normalized disturbance frequencies in the nonlinear response scenarios of autonomous driving systems by combining the SCP with a frequency spectrum analysis model. The results showed that the normalized disturbance frequency in the cut-in scenario was approximately 0.2. When the relative longitudinal distance and speed of the vehicle are the same, if the cut-in speed of the cut-in vehicle is larger, the normalized disturbance frequency is higher, indicating that the cut-in process of the autonomous vehicle is more dangerous and may trigger a collision.","PeriodicalId":100793,"journal":{"name":"Journal of Intelligent and Connected Vehicles","volume":"7 3","pages":"205-218"},"PeriodicalIF":0.0,"publicationDate":"2024-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10695163","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142323054","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This study provides a systematic analysis of the resource-consuming training of deep reinforcement-learning (DRL) agents for simulated low-speed automated driving (AD). In Unity, this study established two case studies: garage parking and navigating an obstacle-dense area. Our analysis involves training a path-planning agent with real-time-only sensor information. This study addresses research questions insufficiently covered in the literature, exploring curriculum learning (CL), agent generalization (knowledge transfer), computation distribution (CPU vs. GPU), and mapless navigation. CL proved necessary for the garage scenario and beneficial for obstacle avoidance. It involved adjustments at different stages, including terminal conditions, environment complexity, and reward function hyperparameters, guided by their evolution in multiple training attempts. Fine-tuning the simulation tick and decision period parameters was crucial for effective training. The abstraction of high-level concepts (e.g., obstacle avoidance) necessitates training the agent in sufficiently complex environments in terms of the number of obstacles. While blogs and forums discuss training machine learning models in Unity, a lack of scientific articles on DRL agents for AD persists. However, since agent development requires considerable training time and difficult procedures, there is a growing need to support such research through scientific means. In addition to our findings, we contribute to the R&D community by providing our environment with open sources.
{"title":"Development of Deep-Learning-Based Autonomous Agents for Low-Speed Maneuvering in Unity","authors":"Riccardo Berta;Luca Lazzaroni;Alessio Capello;Marianna Cossu;Luca Forneris;Alessandro Pighetti;Francesco Bellotti","doi":"10.26599/JICV.2023.9210039","DOIUrl":"https://doi.org/10.26599/JICV.2023.9210039","url":null,"abstract":"This study provides a systematic analysis of the resource-consuming training of deep reinforcement-learning (DRL) agents for simulated low-speed automated driving (AD). In Unity, this study established two case studies: garage parking and navigating an obstacle-dense area. Our analysis involves training a path-planning agent with real-time-only sensor information. This study addresses research questions insufficiently covered in the literature, exploring curriculum learning (CL), agent generalization (knowledge transfer), computation distribution (CPU vs. GPU), and mapless navigation. CL proved necessary for the garage scenario and beneficial for obstacle avoidance. It involved adjustments at different stages, including terminal conditions, environment complexity, and reward function hyperparameters, guided by their evolution in multiple training attempts. Fine-tuning the simulation tick and decision period parameters was crucial for effective training. The abstraction of high-level concepts (e.g., obstacle avoidance) necessitates training the agent in sufficiently complex environments in terms of the number of obstacles. While blogs and forums discuss training machine learning models in Unity, a lack of scientific articles on DRL agents for AD persists. However, since agent development requires considerable training time and difficult procedures, there is a growing need to support such research through scientific means. In addition to our findings, we contribute to the R&D community by providing our environment with open sources.","PeriodicalId":100793,"journal":{"name":"Journal of Intelligent and Connected Vehicles","volume":"7 3","pages":"229-244"},"PeriodicalIF":0.0,"publicationDate":"2024-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10695161","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142324272","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The mobility landscape is experiencing major changes due to two emerging transportation trends, autonomous vehicles (AVs) and on-demand transportation, and the convergence of these smart mobility innovations as shared autonomous vehicles (SAVs) can considerably alter travel behavior and consequently the ecological and societal aspects of the transportation sector. On-demand autonomous mobility is a promising transportation mode, but further research is necessary to evaluate its various aspects and implications prior to widespread adoption. Thus, this study investigates the effects of integrating automation and on-demand mobility by analyzing the effects on the environment, public transportation, land use, vehicle ownership, and public acceptance. A comprehensive literature review was performed, and through a detailed review of 210 articles, the impacts of each of these categories were determined and classified according to their causes, and the number of publications with which they were cited in the literature was determined. The review showed that SAVs can either positively or negatively impact categories and have the potential to minimize mobility obstacles and transportation inequity if legislators use technology to develop a better transportation system by initiating effective policies that govern the four impacted areas. A list of 22 policy recommendations designed to avoid the negative consequences of SAVs by maximizing the benefits of the technology while limiting the associated risks was also identified. The findings of this review will be beneficial to AV manufacturers, transportation professionals, and especially policymakers, who play an integral role in shaping how society benefits from SAV technology.
由于自动驾驶汽车（AV）和按需运输这两种新兴的交通趋势，交通格局正在经历重大变化，而共享自动驾驶汽车（SAV）这种智能交通创新的融合可以大大改变人们的出行行为，进而改变交通领域的生态和社会方面。按需自主交通是一种前景广阔的交通模式，但在广泛采用之前，有必要开展进一步研究，以评估其各个方面和影响。因此，本研究通过分析对环境、公共交通、土地使用、车辆所有权和公众接受度的影响，调查了自动化与按需移动相结合的影响。本研究进行了全面的文献综述，通过对 210 篇文章的详细审查，确定了上述各类影响，并根据其成因进行了分类，还确定了这些影响在文献中的引用数量。审查结果表明，如果立法者利用技术开发出更好的交通系统，启动有效的政策来管理这四个受影响的领域，那么小型自动变速器就有可能最大限度地减少流动障碍和交通不公平。此外，还确定了一份 22 条政策建议清单，旨在通过最大限度地发挥技术的益处，同时限制相关风险，避免 SAVs 带来的负面影响。本次审查的结果将有益于自动驾驶汽车制造商、交通专业人士，特别是政策制定者，他们在塑造社会如何从自动驾驶汽车技术中受益方面发挥着不可或缺的作用。
{"title":"Convergence of Emerging Transportation Trends: A Comprehensive Review of Shared Autonomous Vehicles","authors":"Deema Almaskati;Sharareh Kermanshachi;Apurva Pamidimukkala","doi":"10.26599/JICV.2023.9210043","DOIUrl":"https://doi.org/10.26599/JICV.2023.9210043","url":null,"abstract":"The mobility landscape is experiencing major changes due to two emerging transportation trends, autonomous vehicles (AVs) and on-demand transportation, and the convergence of these smart mobility innovations as shared autonomous vehicles (SAVs) can considerably alter travel behavior and consequently the ecological and societal aspects of the transportation sector. On-demand autonomous mobility is a promising transportation mode, but further research is necessary to evaluate its various aspects and implications prior to widespread adoption. Thus, this study investigates the effects of integrating automation and on-demand mobility by analyzing the effects on the environment, public transportation, land use, vehicle ownership, and public acceptance. A comprehensive literature review was performed, and through a detailed review of 210 articles, the impacts of each of these categories were determined and classified according to their causes, and the number of publications with which they were cited in the literature was determined. The review showed that SAVs can either positively or negatively impact categories and have the potential to minimize mobility obstacles and transportation inequity if legislators use technology to develop a better transportation system by initiating effective policies that govern the four impacted areas. A list of 22 policy recommendations designed to avoid the negative consequences of SAVs by maximizing the benefits of the technology while limiting the associated risks was also identified. The findings of this review will be beneficial to AV manufacturers, transportation professionals, and especially policymakers, who play an integral role in shaping how society benefits from SAV technology.","PeriodicalId":100793,"journal":{"name":"Journal of Intelligent and Connected Vehicles","volume":"7 3","pages":"177-189"},"PeriodicalIF":0.0,"publicationDate":"2024-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10695164","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142324349","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With the rapid growth of urbanization and the increasing demand for transportation, urban traffic congestion has become a hindrance to individuals' travel experience. Urban intersections are one of the primary sources of traffic congestion, and these bottlenecks have a negative impact not only on traffic efficacy but also on the surrounding road traffic in the region. To alleviate urban traffic congestion, cyber-physical systems have been widely implemented in the transportation industry, allowing for the perception, analysis, calculation, and dispatching of urban traffic flow, as well as making urban transportation safe, efficient, and quick. As the system scale and functions increase, system design has become increasingly complex, necessitating a deeper comprehension of the system's structure and interaction relationships to construct a stable and reliable system. Therefore, this study proposes a method for designing cyber-physical systems for urban traffic intersections based on Model-Based Systems Engineering (MBSE). This method models and analyses exhaustively the system's requirements, functions, and logical architecture using System Modeling Language (SysML). After the architecture design has been completed, an architecture verification and optimization method based on Failure Mode and Effect Analysis (FMEA) for urban road intersection cyber-physical systems is utilized to analyze the architecture's reliability by analyzing the failure modes of activities and to optimize the system architecture to improve the design's efficiency and reliability.
{"title":"CPS Architecture Design for Urban Roadway Intersections Based on MBSE","authors":"Chen Wang;Xiaoping Ma;Limin Jia;Zheng Lai;Zhexuan Yang;Han Yan;Jing Zhao","doi":"10.26599/JICV.2023.9210030","DOIUrl":"https://doi.org/10.26599/JICV.2023.9210030","url":null,"abstract":"With the rapid growth of urbanization and the increasing demand for transportation, urban traffic congestion has become a hindrance to individuals' travel experience. Urban intersections are one of the primary sources of traffic congestion, and these bottlenecks have a negative impact not only on traffic efficacy but also on the surrounding road traffic in the region. To alleviate urban traffic congestion, cyber-physical systems have been widely implemented in the transportation industry, allowing for the perception, analysis, calculation, and dispatching of urban traffic flow, as well as making urban transportation safe, efficient, and quick. As the system scale and functions increase, system design has become increasingly complex, necessitating a deeper comprehension of the system's structure and interaction relationships to construct a stable and reliable system. Therefore, this study proposes a method for designing cyber-physical systems for urban traffic intersections based on Model-Based Systems Engineering (MBSE). This method models and analyses exhaustively the system's requirements, functions, and logical architecture using System Modeling Language (SysML). After the architecture design has been completed, an architecture verification and optimization method based on Failure Mode and Effect Analysis (FMEA) for urban road intersection cyber-physical systems is utilized to analyze the architecture's reliability by analyzing the failure modes of activities and to optimize the system architecture to improve the design's efficiency and reliability.","PeriodicalId":100793,"journal":{"name":"Journal of Intelligent and Connected Vehicles","volume":"7 3","pages":"190-204"},"PeriodicalIF":0.0,"publicationDate":"2024-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10695165","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142323055","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-16DOI: 10.26599/JICV.2023.9210049
Yongsheng Zhang;Chen Tu;Kun Gao;Liang Wang
As urban transportation increasingly impacts daily life, efficiently utilizing traffic resources and developing public transportation have become crucial for addressing issues such as congestion, frequent accidents, and noise pollution. The rapid advancement of intelligent autonomous driving technologies, particularly environmental perception technologies, offers new directions for solving these problems. This review discusses the application of multisensor information fusion technology in environmental perception for intelligent vehicles, analyzing the components and performance of various sensors and their specific applications in autonomous driving. Through multisensor information fusion, the accuracy of environmental perception is enhanced, optimizing decision support for autonomous driving systems and thereby improving vehicle safety and driving efficiency. This study also discusses the challenges faced by information fusion technology and future development trends, providing references for further research and application in intelligent transportation systems.
{"title":"Multisensor Information Fusion: Future of Environmental Perception in Intelligent Vehicles","authors":"Yongsheng Zhang;Chen Tu;Kun Gao;Liang Wang","doi":"10.26599/JICV.2023.9210049","DOIUrl":"https://doi.org/10.26599/JICV.2023.9210049","url":null,"abstract":"As urban transportation increasingly impacts daily life, efficiently utilizing traffic resources and developing public transportation have become crucial for addressing issues such as congestion, frequent accidents, and noise pollution. The rapid advancement of intelligent autonomous driving technologies, particularly environmental perception technologies, offers new directions for solving these problems. This review discusses the application of multisensor information fusion technology in environmental perception for intelligent vehicles, analyzing the components and performance of various sensors and their specific applications in autonomous driving. Through multisensor information fusion, the accuracy of environmental perception is enhanced, optimizing decision support for autonomous driving systems and thereby improving vehicle safety and driving efficiency. This study also discusses the challenges faced by information fusion technology and future development trends, providing references for further research and application in intelligent transportation systems.","PeriodicalId":100793,"journal":{"name":"Journal of Intelligent and Connected Vehicles","volume":"7 3","pages":"163-176"},"PeriodicalIF":0.0,"publicationDate":"2024-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10600093","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142324356","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Positioning and mapping technology is a difficult and hot topic in autonomous driving environment sensing systems. In a complex traffic environment, the signal of the Global Navigation Satellite System (GNSS) will be blocked, leading to inaccurate vehicle positioning. To ensure the security of automatic electric campus vehicles, this study is based on the Lightweight and Ground-Optimized Lidar Odometry and Mapping on Variable Terrain (LEGO-LOAM) algorithm with a monocular vision system added. An algorithm framework based on Lidar-IMU-Camera (Lidar means light detection and ranging) fusion was proposed. A lightweight monocular vision odometer model was used, and the LEGO-LOAM system was employed to initialize monocular vision. The visual odometer information was taken as the initial value of the laser odometer. At the back-end opti9mization phase error state, the Kalman filtering fusion algorithm was employed to fuse the visual odometer and LEGO-LOAM system for positioning. The visual word bag model was applied to perform loopback detection. Taking the test results into account, the laser radar loopback detection was further optimized, reducing the accumulated positioning error. The real car experiment results showed that our algorithm could improve the mapping quality and positioning accuracy in the campus environment. The Lidar-IMU-Camera algorithm framework was verified on the Hong Kong city dataset UrbanNav. Compared with the LEGO-LOAM algorithm, the results show that the proposed algorithm can effectively reduce map drift, improve map resolution, and output more accurate driving trajectory information.
{"title":"Localization and Mapping Algorithm Based on Lidar-IMU-Camera Fusion","authors":"Yibing Zhao;Yuhe Liang;Zhenqiang Ma;Lie Guo;Hexin Zhang","doi":"10.26599/JICV.2023.9210027","DOIUrl":"https://doi.org/10.26599/JICV.2023.9210027","url":null,"abstract":"Positioning and mapping technology is a difficult and hot topic in autonomous driving environment sensing systems. In a complex traffic environment, the signal of the Global Navigation Satellite System (GNSS) will be blocked, leading to inaccurate vehicle positioning. To ensure the security of automatic electric campus vehicles, this study is based on the Lightweight and Ground-Optimized Lidar Odometry and Mapping on Variable Terrain (LEGO-LOAM) algorithm with a monocular vision system added. An algorithm framework based on Lidar-IMU-Camera (Lidar means light detection and ranging) fusion was proposed. A lightweight monocular vision odometer model was used, and the LEGO-LOAM system was employed to initialize monocular vision. The visual odometer information was taken as the initial value of the laser odometer. At the back-end opti9mization phase error state, the Kalman filtering fusion algorithm was employed to fuse the visual odometer and LEGO-LOAM system for positioning. The visual word bag model was applied to perform loopback detection. Taking the test results into account, the laser radar loopback detection was further optimized, reducing the accumulated positioning error. The real car experiment results showed that our algorithm could improve the mapping quality and positioning accuracy in the campus environment. The Lidar-IMU-Camera algorithm framework was verified on the Hong Kong city dataset UrbanNav. Compared with the LEGO-LOAM algorithm, the results show that the proposed algorithm can effectively reduce map drift, improve map resolution, and output more accurate driving trajectory information.","PeriodicalId":100793,"journal":{"name":"Journal of Intelligent and Connected Vehicles","volume":"7 2","pages":"97-107"},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10586905","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141543845","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-06-01DOI: 10.26599/JICV.2023.9210037
Miklós Lukovics;Szabolcs Prónay;Barbara Nagy
Previous studies have identified trust as one of the key factors in the technology acceptance of autonomous vehicles. As these studies mostly investigated the population in general, little is known about segment-specific differences. Furthermore, the widely used survey methods are less able to capture the deeper forms of trust—which neuroscientific methods are much better suited to capture. The main objective of our research is to study trust as one of the key factors of technology acceptance related to autonomous vehicles by using neuroscientific methods for specific consumer segments. Real-time eye-tracking tests were applied to a sample of 113 participants, combined with a posttest self-report. The tests were carried out under laboratory conditions during which our subjects watched videos recorded with the internal cameras of autonomous vehicles. Based on the fixation count, total fixation duration, and pupil dilation, we empirically verified that the trust level of all five identified segments is relatively low, while the trust level of the “traditional rejecting” segment is the lowest. An increase in trust level can be shown if the subjects receive extra information about the journey. Another important finding is that the self-reported trust level is not always congruent with the eye-tracking analysis results; therefore, combined approaches can lead to greater measurement validity.
{"title":"Segmented Trust Assessment in Autonomous Vehicles via Eye-Tracking","authors":"Miklós Lukovics;Szabolcs Prónay;Barbara Nagy","doi":"10.26599/JICV.2023.9210037","DOIUrl":"https://doi.org/10.26599/JICV.2023.9210037","url":null,"abstract":"Previous studies have identified trust as one of the key factors in the technology acceptance of autonomous vehicles. As these studies mostly investigated the population in general, little is known about segment-specific differences. Furthermore, the widely used survey methods are less able to capture the deeper forms of trust—which neuroscientific methods are much better suited to capture. The main objective of our research is to study trust as one of the key factors of technology acceptance related to autonomous vehicles by using neuroscientific methods for specific consumer segments. Real-time eye-tracking tests were applied to a sample of 113 participants, combined with a posttest self-report. The tests were carried out under laboratory conditions during which our subjects watched videos recorded with the internal cameras of autonomous vehicles. Based on the fixation count, total fixation duration, and pupil dilation, we empirically verified that the trust level of all five identified segments is relatively low, while the trust level of the “traditional rejecting” segment is the lowest. An increase in trust level can be shown if the subjects receive extra information about the journey. Another important finding is that the self-reported trust level is not always congruent with the eye-tracking analysis results; therefore, combined approaches can lead to greater measurement validity.","PeriodicalId":100793,"journal":{"name":"Journal of Intelligent and Connected Vehicles","volume":"7 2","pages":"151-161"},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10586910","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141543846","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-06-01DOI: 10.26599/JICV.2023.9210036
Zihao Sheng;Zilin Huang;Sikai Chen
Trajectory prediction for heterogeneous traffic agents plays a crucial role in ensuring the safety and efficiency of automated driving in highly interactive traffic environments. Numerous studies in this area have focused on physics-based approaches because they can clearly interpret the dynamic evolution of trajectories. However, physics-based methods often suffer from limited accuracy. Recent learning-based methods have demonstrated better performance, but they cannot be fully trusted due to the insufficient incorporation of physical constraints. To mitigate the limitations of purely physics-based and learning-based approaches, this study proposes a kinematics-aware multigraph attention network (KA-MGAT) that incorporates physics models into a deep learning framework to improve the learning process of neural networks. Besides, we propose a residual prediction module to further refine the trajectory predictions and address the limitations arising from simplified assumptions in kinematic models. We evaluate our proposed model through experiments on two challenging trajectory datasets, namely, ApolloScape and NGSIM. Our findings from the experiments demonstrate that our model outperforms various kinematics-agnostic models with respect to prediction accuracy and learning efficiency.
{"title":"Kinematics-Aware Multigraph Attention Network with Residual Learning for Heterogeneous Trajectory Prediction","authors":"Zihao Sheng;Zilin Huang;Sikai Chen","doi":"10.26599/JICV.2023.9210036","DOIUrl":"https://doi.org/10.26599/JICV.2023.9210036","url":null,"abstract":"Trajectory prediction for heterogeneous traffic agents plays a crucial role in ensuring the safety and efficiency of automated driving in highly interactive traffic environments. Numerous studies in this area have focused on physics-based approaches because they can clearly interpret the dynamic evolution of trajectories. However, physics-based methods often suffer from limited accuracy. Recent learning-based methods have demonstrated better performance, but they cannot be fully trusted due to the insufficient incorporation of physical constraints. To mitigate the limitations of purely physics-based and learning-based approaches, this study proposes a kinematics-aware multigraph attention network (KA-MGAT) that incorporates physics models into a deep learning framework to improve the learning process of neural networks. Besides, we propose a residual prediction module to further refine the trajectory predictions and address the limitations arising from simplified assumptions in kinematic models. We evaluate our proposed model through experiments on two challenging trajectory datasets, namely, ApolloScape and NGSIM. Our findings from the experiments demonstrate that our model outperforms various kinematics-agnostic models with respect to prediction accuracy and learning efficiency.","PeriodicalId":100793,"journal":{"name":"Journal of Intelligent and Connected Vehicles","volume":"7 2","pages":"138-150"},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10586904","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141543876","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-06-01DOI: 10.26599/JICV.2023.9210040
Yang Fei;Peng Shi;Yang Liu;Liang Wang
In recent years, advancements in onboard computing hardware and wireless communication technology have remarkably stimulated the development of intelligent and connected vehicles (ICVs). Specifically, some researchers have investigated the issue of employing various advanced control techniques to optimize the performance of autonomous vehicles in practice (Sun et al., 2023; Zhang et al., 2023a, 2023b). Therefore, this article aims to discuss why and how control engineering plays an essential role in the development of ICVs.
{"title":"Critical Roles of Control Engineering in the Development of Intelligent and Connected Vehicles","authors":"Yang Fei;Peng Shi;Yang Liu;Liang Wang","doi":"10.26599/JICV.2023.9210040","DOIUrl":"https://doi.org/10.26599/JICV.2023.9210040","url":null,"abstract":"In recent years, advancements in onboard computing hardware and wireless communication technology have remarkably stimulated the development of intelligent and connected vehicles (ICVs). Specifically, some researchers have investigated the issue of employing various advanced control techniques to optimize the performance of autonomous vehicles in practice (Sun et al., 2023; Zhang et al., 2023a, 2023b). Therefore, this article aims to discuss why and how control engineering plays an essential role in the development of ICVs.","PeriodicalId":100793,"journal":{"name":"Journal of Intelligent and Connected Vehicles","volume":"7 2","pages":"79-85"},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10586906","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141543874","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-06-01DOI: 10.26599/JICV.2023.9210026
Qianwen Li;Peng Zhang;Handong Yao;Zhiwei Chen;Xiaopeng Li
Motivated by the promising benefits of connected and autonomous vehicles (CAVs) in improving fuel efficiency, mitigating congestion, and enhancing safety, numerous theoretical models have been proposed to plan CAV multiple-step trajectories (time-specific speed/location trajectories) to accomplish various operations. However, limited efforts have been made to develop proper trajectory control techniques to regulate vehicle movements to follow multiple-step trajectories and test the performance of theoretical trajectory planning models with field experiments. Without an effective control method, the benefits of theoretical models for CAV trajectory planning can be difficult to harvest. This study proposes an online learning-based model predictive vehicle trajectory control structure to follow time-specific speed and location profiles. Unlike single-step controllers that are dominantly used in the literature, a multiple-step model predictive controller is adopted to control the vehicle's longitudinal movements for higher accuracy. The model predictive controller output (speed) cannot be interpreted by vehicles. A reinforcement learning agent is used to convert the speed value to the vehicle's direct control variable (i.e., throttle/brake). The reinforcement learning agent captures real-time changes in the operating environment. This is valuable in saving parameter calibration resources and improving trajectory control accuracy. A line tracking controller keeps vehicles on track. The proposed control structure is tested using reduced-scale robot cars. The adaptivity of the proposed control structure is demonstrated by changing the vehicle load. Then, experiments on two fundamental CAV platoon operations (i.e., platooning and split) show the effectiveness of the proposed trajectory control structure in regulating robot movements to follow time-specific reference trajectories.
{"title":"Online Learning-Based Model Predictive Trajectory Control for Connected and Autonomous Vehicles: Modeling and Physical Tests","authors":"Qianwen Li;Peng Zhang;Handong Yao;Zhiwei Chen;Xiaopeng Li","doi":"10.26599/JICV.2023.9210026","DOIUrl":"https://doi.org/10.26599/JICV.2023.9210026","url":null,"abstract":"Motivated by the promising benefits of connected and autonomous vehicles (CAVs) in improving fuel efficiency, mitigating congestion, and enhancing safety, numerous theoretical models have been proposed to plan CAV multiple-step trajectories (time-specific speed/location trajectories) to accomplish various operations. However, limited efforts have been made to develop proper trajectory control techniques to regulate vehicle movements to follow multiple-step trajectories and test the performance of theoretical trajectory planning models with field experiments. Without an effective control method, the benefits of theoretical models for CAV trajectory planning can be difficult to harvest. This study proposes an online learning-based model predictive vehicle trajectory control structure to follow time-specific speed and location profiles. Unlike single-step controllers that are dominantly used in the literature, a multiple-step model predictive controller is adopted to control the vehicle's longitudinal movements for higher accuracy. The model predictive controller output (speed) cannot be interpreted by vehicles. A reinforcement learning agent is used to convert the speed value to the vehicle's direct control variable (i.e., throttle/brake). The reinforcement learning agent captures real-time changes in the operating environment. This is valuable in saving parameter calibration resources and improving trajectory control accuracy. A line tracking controller keeps vehicles on track. The proposed control structure is tested using reduced-scale robot cars. The adaptivity of the proposed control structure is demonstrated by changing the vehicle load. Then, experiments on two fundamental CAV platoon operations (i.e., platooning and split) show the effectiveness of the proposed trajectory control structure in regulating robot movements to follow time-specific reference trajectories.","PeriodicalId":100793,"journal":{"name":"Journal of Intelligent and Connected Vehicles","volume":"7 2","pages":"86-96"},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10586903","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141543844","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：