Autonomous Vehicles Research Articles

Research on lane detection algorithms based on GTNs

Abstract. Lane detection technology plays an important role in the lane departure warning and adaptive cruise control functions of autonomous vehicle systems. Traditional lane line detection includes experimental steps such as preyreatment, color space conversion, feature extraction, and lane line tracking, but there are still some problems such as recognition accuracy in complex environments and parameter adjustment dependency. To overcome these limitations, this study uses a novel method that can directly predict the parameters of the lane shape model, thus avoiding complex post-processing steps. This research method uses the network structure based on Graph Transformer Networks (GTNS) to capture richer structural information and contextual relationships, thereby improving the accuracy of detection. In terms of feature extraction, Graph Transformer Networks is used to accurately capture the structural information of the lane through the attention mechanism. At the same time, the results in the later stage show that our method has better advantages in accuracy and speed. In short, this optimization scheme not only improves the detection speed but also ensures a certain degree of accuracy. The method will perform better in the future after further optimization such as multi-sensor fusion and real-time optimization.

A review of advances in image recognition models: From K-NN to Vision Transformer

Abstract. In the digital age, visual data are everywhere. Image recognition has become a vital branch of artificial intelligence and is used in a wide range of fields. From medical diagnostics to autonomous vehicles, its impact is profound and far-reaching. This review visits the significant advancements in the field of image recognition, summarizes and analyzes the progress from traditional methods like K-Nearest Neighbors (K-NN) to newer models like Vision Transformer. This paper analyzes image recognition models like AlexNet, GoogLeNet, ResNet, and Vision Transformer, with the principles, strengths, weaknesses, and the significance of these models to deep learning fields, along with their applications in image recognition tasks. These models not only push the boundaries of image recognition but also find extensive applications across various domains. In the future, image recognition can continue to develop with larger and better datasets, increased computing power, and more advanced network structures.

A path-planning algorithm for autonomous vehicles based on traffic stability criteria: the AS-IAPF algorithm

A path-planning algorithm for autonomous vehicles based on traffic stability criteria: the AS-IAPF algorithm

Advanced obstacle recognition and trajectory prediction for autonomous vehicles using 3D CNN and GRU

Abstract. Recent years have seen a rapid advancement in new energy vehicle technology, which has increased consumer demand for autonomous driving systems. However, current self-driving cars still face many technical barriers. Obstacle detection and obstacle trajectory prediction in dynamic environments is one of the main difficulties of the technique. To improve self-driving cars' capacity for route planning and decision-making in dynamic contexts, this thesis proposes an obstacle identification and trajectory prediction model that combines a gated recurrent unit (GRU) network with a three-dimensional convolutional neural network (3D CNN). The integrated framework utilises the feature recognition capability of 3D CNN and the prediction capability of GRU to accurately predict dynamic obstacles, thereby improving the performance of the model in various types of environments (e.g., visual SLAM). Experimental results based on publicly available datasets (e.g., KITTI and Argoverse) show that the proposed method has significant advantages in dynamic obstacle identification and prediction as well as overall model operation efficiency. The findings show that the integrated framework is useful and successful in advancing the creation of self-driving automobiles.

The effects of communication topology on mixed traffic flow: modelling and analysis

This paper aims to evaluate the combined impact of the communication topology (CT) and vehicular spatial distribution (SD) on mixed traffic of human-driven vehicles (HDVs) and connected and autonomous vehicles (CAVs). A generalized car-following (CF) model is presented to capture mixed traffic dynamics, integrating the effects of the CT, human reaction time, information delay and the optimal velocity changes with memory. Then, the linear stable condition of the model is derived using the perturbation method. Finally, extensive simulations are conducted under different CTs (i.e., multiple-predecessor leader following (MPLF) topology, predecessor following (PF) topology, and predecessor-leader following (PLF) topology) and six kinds of SDs. The results reveal that the MPLF topology demonstrates the best steady performance (i.e., stability) and dynamic performance (i.e., smoothness), and that the SD where CAVs are concentrated at the front of the traffic flow shows superior dynamic performance.

Intelligent Vehicle Trajectory Tracking Based on Horizontal and Vertical Integrated Control

To address the common issues of accuracy and stability in trajectory tracking tasks for autonomous vehicles, this study proposes an innovative composite control strategy that skillfully integrates lateral and longitudinal dynamic control. For lateral control, model predictive control (MPC) theory is introduced to compute the front wheel steering angle that ensures optimal trajectory following. On the longitudinal control level, the vehicle’s acceleration and deceleration logic are finely tuned to ensure precise adherence to the preset speed trajectory. More importantly, by deeply integrating these two control methods, the comprehensive coordination of the vehicle’s lateral and longitudinal movements is achieved. To validate the effectiveness of the proposed control strategy, simulations were conducted using the CarSim and MATLAB/Simulink platforms. The analysis of the simulation results confirms that the proposed method effectively improves speed tracking stability and significantly enhances path tracking accuracy and overall driving stability.

Evaluation of Autonomous Driving Safety by Operational Design Domains (ODD) in Mixed Traffic

This study derived effective driving behavior indicators to assess the driving safety of autonomous vehicles (AV). A variety of operation design domains (ODD) in urban road networks, which include intersections, illegal parking, bus stop, bicycle lanes, and pedestrian crossings, were taken into consideration in traffic simulation analyses. Both longitudinal and interaction driving indicators were investigated to identify the driving performance of AVs in terms of traffic safety in mixed traffic stream based on simulation experiments. As a result of identifying the appropriate evaluation indicator, time-varying stochastic volatility (VF) headway time was selected as a representative evaluation indicator for left turn and straight through signalized intersections among ODDs related to intersection types. VF headway time is suitable for evaluating driving ability by measuring the variation in driving safety in terms of interaction with the leading vehicle. In addition to ODDs associated with intersection type, U-turns, additional lane segments, illegal parking, bus stops, and merging lane have common characteristics that increase the likelihood of interactions with neighboring vehicles. The VF headway time for these ODDs was derived as driving safety in terms of interaction between vehicles. The results of this study would be valuable in establishing a guideline for driving performance evaluation of AVs. The study found that unsignalized left turns, signalized right turns, and roundabouts had the highest risk scores of 0.554, 0.525, and 0.501, respectively, indicating these as the most vulnerable ODDs for AVs. Additionally, intersection and mid-block crosswalks, as well as bicycle lanes, showed high risk scores due to frequent interactions with pedestrians and cyclists. These areas are particularly risky because they involve unpredictable movements from non-vehicular road users, which require AVs to make rapid adjustments in speed and trajectory. These findings provide a foundation for improving AV algorithms to enhance safety and establishing objective criteria for AV policy-making.

Causality-Sensitive Scheduling to Reduce Latency in Vehicle-to-Vehicle Interactions.

This paper shows through real-life measurement that bi-directional vehicle-to-vehicle (V2V) communication latency can be dominated by sidelink scheduling delay when causality is not taken into account. Moreover, the large delay persists for a few seconds at a time once it occurs. In applications like maneuver coordination between autonomous vehicles or in platoon, such delay can be highly detrimental to safety and efficiency. We investigate the source of the problem and propose a solution that factors in causality in interactive communication. Specifically, we develop a constraint under which the resource positions are automatically aligned between the communicating vehicles, and the delay spikes are provably eliminated. Through the measurements on commercial V2X devices, we confirm that enforcing the constraint can remove latency spikes so that 5G sidelink can be more easily applied to time-sensitive interactions between vehicles.

Non-stationary spatio-temporal modeling using the stochastic advection–diffusion equation

We construct flexible spatio-temporal models through stochastic partial differential equations (SPDEs) where both diffusion and advection can be spatially varying. Computations are done through a Gaussian Markov random field approximation of the solution of the SPDE, which is constructed through a finite volume method. The new flexible non-separable model is compared to a flexible separable model both for reconstruction and forecasting, and evaluated in terms of root mean square errors and continuous rank probability scores. A simulation study demonstrates that the non-separable model performs better when the data is simulated from a non-separable model with diffusion and advection. Further, we estimate surrogate models for emulating the output of a ocean model in Trondheimsfjorden, Norway, and simulate observations of autonomous underwater vehicles. The results show that the flexible non-separable model outperforms the flexible separable model for real-time prediction of unobserved locations.

Hybrid offline-online reinforcement learning for obstacle avoidance in autonomous underwater vehicles

ABSTRACT This study presents a novel control framework for autonomous underwater vehicles (AUVs) that integrates offline and online reinforcement learning to enhance navigation accuracy and obstacle avoidance. Recognizing the limitations of online reinforcement learning due to high interaction demands and the challenges of offline learning from suboptimal data, we construct kinematic and dynamic models of AUVs as a foundation for our control strategies. Our controller employs a fusion strategy, utilizing offline datasets generated via the Model Predictive Path Integral (MPPI) method and exploration rewards based on kernel density estimation (KDE) to improve exploration of low-confidence areas. Extensive simulations demonstrate the effectiveness of our approach in complex scenarios, including navigation to multiple targets, power insufficiencies, water flow interference, and dynamic obstacles. The trained agents exhibited superior navigation accuracy and obstacle avoidance, underscoring the practicality of our combined learning methods for robust AUV performance in intricate environments.

Global point cloud registration network for large transformations

Three-dimensional registration is an established yet challenging problem that is key in many different applications, such as mapping the environment for autonomous vehicles, or modeling people for avatar creation, among others. Registration refers to the process of mapping multiple data into the same coordinate system by means of matching correspondences and transformation estimation. Novel proposals exploit the benefits of deep learning architectures for this purpose, as they learn the best features for the data, providing better matches and hence results. However, the state of the art is usually focused on cases of relatively small transformations, although in certain applications and in a real and practical environment, large transformations are very common. In this paper, we present ReLaTo (Registration for Large Transformations), an architecture that addresses the cases where large transformations happen while maintaining good performance for local transformations. This proposal uses a novel Softmax pooling layer to find correspondences in a bilateral consensus manner between two point sets, sampling the most confident matches. These matches estimate a coarse and global registration using weighted Singular Value Decomposition (SVD). A target-guided denoising step is applied to both the obtained matches and latent features to estimate the final fine registration considering the local geometry. All these steps are carried out following an end-to-end approach, which has been shown to perform better than 7 state-of-the-art registration methods in two datasets commonly used for this task (ModelNet40 and the Karlsruhe Institute of Technology and Toyota Technological Institute dataset, KITTI), especially in the case of large transformations.Graphic abstract

A novel road attribute detection system for autonomous vehicles using sensor fusion

A novel road attribute detection system for autonomous vehicles using sensor fusion

SYSAI for System Health Management - a Statistical Framework for the Analysis of Diagnosis Systems

On-board failure diagnosis and health management systems (HMS) are crucial for the operation of complex autonomous aerospace systems. False alarms (false positives, FPs) or false negatives (FNs) can lead to lower system performance or even loss of mission or the autonomous vehicle. Therefore, a careful verification and validation (V&V) is important. Due to the high dimensionality of the system’s state space, however, exhaustive testing of the HMS is usually not possible. In this paper, we present how our SYSAI (System Analysis for Systems with AI components) framework can support intelligent analysis and testing of HMS on the system level. SYSAI’s capabilities to efficiently explore high-dimensional state and parameter spaces and to identify diagnosability regions and their boundaries, makes a comprehensive analysis of the diagnosis system possible and can provide feedback to the designer. We will illustrate our approach using the ADAPT (Advanced Diagnostics and Prognostics Testbed) redundant power storage and distribution system.

Multi-Vehicle Cooperative Decision-Making in Merging Area Based on Deep Multi-Agent Reinforcement Learning

In recent years, reinforcement learning (RL) methods have shown powerful learning capabilities in single-vehicle autonomous driving. However, few studies have focused on multi-vehicle cooperative driving based on RL, particularly in the dynamically changing traffic environments of highway ramp merge zones. In this paper, a multi-agent deep reinforcement learning (MARL) framework for multi-vehicle cooperative decision-making is proposed based on actor–critic, which categorizes vehicles into two groups according to their origins in the merging area. At the same time, the complexity of the network is reduced and the training process of the model is accelerated by utilizing mechanisms such as partial parameter sharing and experience playback. Additionally, a combination of global and individual rewards is adopted to promote cooperation in connected autonomous vehicles (CAVs) and balance individual and group interests. The training performance of the model is compared under three traffic densities, and our method is also compared with state-of-the-art benchmark methods. The simulation results show that the proposed MARL framework can have stronger policy learning capability and stability under various traffic flow conditions. Moreover, it can also effectively improve the speed of vehicles in the merging zone and reduce traffic conflicts.

Wide human-like neural network incorporating driving styles for human-like driving intention analysis

Enhancing the synergy between autonomous and human-driven vehicles at the societal level requires understanding drivers’ behaviors and cognitive patterns, as well as conducting human-like driving intention analysis. To achieve this goal, this study designs a novel framework for analyzing human-like driving intention. Firstly, a spectral clustering method is employed to characterize driving styles. Secondly, a misclassification cost matrix is tailored to different driving needs. Finally, inspired by the complex neural networks found in the human brain, we develop a specialized lightweight neural network, termed the Width Human-like Neural Network (WNN), aimed at realizing personalized cognition and facilitating human-like decision-making in driving intention. Experimental studies and validation based on natural driving trajectory data from Kunming, China, demonstrate that the method accurately infers internal implicit driving intention from external explicit and observable driving behaviors, achieving a prediction accuracy of 99.8%. This framework strategically allocates limited computational resources to critical areas for autonomous vehicles and exemplifies best practices for improving neural network performance in driving intention analysis tasks.

Enhancing underwater thruster anomaly detection with support vector glow encoding description

As the driving system of autonomous underwater vehicles (AUVs), the healthy operation of underwater thrusters is crucial to ensure the performance of AUVs. Due to the challenge of obtaining faulty samples during normal operation, anomaly detection of underwater thrusters is often carried out using only normal samples. To address this, Support Vector Glow Encoding Description (SVGED) is proposed for anomaly detection of underwater thrusters trained with only normal samples. Specifically, the deep features of the operational vibration signals collected from the underwater thruster are extracted by a deep convolutional autoencoder based on a glow model. The glow-encoded data are then used to identify the boundary of normal operation for the underwater thruster. When a new sample enters the trained SVGED model, it can be effectively identified as representing a normal or anomalous condition of the underwater thruster. The developed SVGED was evaluated using AUV experiments. When using the proposed SVGED method, the accuracy reaches 99.81%, Results show that the proposed method can effectively real time detect anomalies in the underwater thruster compared to other methods. It lays the foundation for ensuring the healthy operation of AUVs.

Autonomous Vehicle Prototype Based on Image Processing

Autonomous Vehicle Prototype Based on Image Processing

A knowledge-guided reinforcement learning method for lateral path tracking

Lateral Control algorithms in autonomous vehicles often necessitates an online fine-tuning procedure in the real world. While reinforcement learning (RL) enables vehicles to learn and improve the lateral control performance through repeated trial and error interactions with a dynamic environment, applying RL directly to safety-critical applications in real physical world is challenging because ensuring safety during the learning process remains difficult. To enable safe learning, a promising direction is to make use of previously gathered offline data, which is frequently accessible in engineering applications. In this context, this paper presents a set of knowledge-guided RL algorithms that can not only fully leverage the prior collected offline data without the need of a physics-based simulator, but also allow further online policy improvement in a smooth, safe and efficient manner. To evaluate the effectiveness of the proposed algorithms on a real controller, a hardware-in-the-loop and a miniature vehicle platform are built. Compared with the vanilla RL, behavior cloning and the existing controller, the proposed algorithms realize a closed-loop solution for lateral control problems from offline training to online fine-tuning, making it attractive for future similar RL-based controller to build upon.

Modeling Suburban Freeway Travel Variability considering Connected and Autonomous Vehicles

Modeling Suburban Freeway Travel Variability considering Connected and Autonomous Vehicles

Dispatching prebooking-based shared autonomous vehicles with hybrid ownership: an approach of vehicle trip planning with uncertain travel times

Prebooking-based shared autonomous vehicles hold significant potential in offering convenient and reliable travel services. This study proposes a vehicle trip chain planning model for shared autonomous vehicles, incorporating both company-owned and privately shared vehicles under the prebooking service mode. To address travel time uncertainty, a distributionally robust chance-constrained optimisation model is developed, ensuring feasibility and resilience to time changes, while discussing a computationally safe tractable approximation for ambiguous chance constraints. Additionally, a Divide-Combine Method is proposed to improve optimising efficiency for large-scale scenarios. Validation with New York dataset confirms the effectiveness of the proposed trip chain planning model in meeting demands and reducing operating costs. The integration of privately shared vehicles partly substitutes company-owned vehicles, thus easing the economic burden on operators. Additionally, while a more conservative approach may slightly impact revenue expectations, it ensures smoother execution of trip chains amid unpredictable traffic conditions, underscoring model’s significance for operations.