Development Of Intelligent Vehicles Research Articles

Research on Image Processing Resource Reconstruction Based on Load Balancing Strategy

With the development of intelligent vehicles, the vehicle image processing system has put forward increasing demand for computing resource utilization efficiency and real-time processing. However, the traditional information processing method that binds software and hardware severely restricts the efficient use of image processing resources. In order to solve this problem, this paper proposes a resource reconstruction scheme based on a load balancing strategy, which can realize unified management and dynamic allocation of image processing resources by establishing a system resource view. Then, this paper builds a physical verification platform and constructs a comparative verification experiment to verify the effectiveness of the resource reconstruction scheme. Experimental results prove that this resource reconstruction scheme can effectively improve the resource utilization efficiency of multi-core digital signal processing (DSP), realize software-defined hardware functions, and optimize the real-time performance of parallel processing.

Driver Assistant Using Deep Learning

Abstract: In the development of intelligent vehicles, accurate traffic sign detection and recognition are critical. This project proposes an improved algorithm for traffic sign detection and recognition, aiming to address limitations of traditional methods such as environmental sensitivity and poor real-time performance of deep learning-based approaches. The algorithm employs HSV-based spatial threshold segmentation for effective traffic sign detection based on shape features. Furthermore, the algorithm enhances the classical convolutional neural network model by utilizing Gabor kernel for initial convolution, incorporating batch normalization after pooling, and employing the Adam optimizer algorithm. The proposed algorithm is evaluated using the German Traffic Sign Recognition Benchmark, achieving a favorable prediction and accurate recognition rate of 99.75%, with an average processing time of 5.4ms per frame. Compared to other algorithms, the proposed approach demonstrates superior accuracy, real-time performance, generalization ability, and training efficiency. The findings of this project are expected to contribute to reducing accident rates and enhancing road traffic safety through improved traffic sign recognition in intelligent vehicles.

Driver Fatigue and Distracted Driving Detection Using Random Forest and Convolutional Neural Network

Driver fatigue and distracted driving are the two most common causes of major accidents. Thus, the on-board monitoring of driving behaviors is key in the development of intelligent vehicles. In this paper, we propose an approach which detects driver fatigue and distracted driving behaviors using vision-based techniques. For driver fatigue detection, a single shot scale-invariant face detector (S3FD) is first used to detect the face in the image and then the face alignment network (FAN) is utilized to extract facial features. After that, the facial features are used to determine the driver’s yawns, head posture, and the opening or closing of their eyes. Finally, the random forest technique is used to analyze the driving conditions. For distracted driving detection, a convolutional neural network (CNN) is used to classify various distracted driving behaviors. Also, Adam optimizer is used to reinforce optimization performance. Compared with existing methods, our approach is more accurate and efficient. Moreover, distracted driving can be detected in real-time running on the embedded hardware.

An Ontology-Based Dynamic Attack Graph Generation Approach for the Internet of Vehicles

With the development of automobile intelligence, the security of the Internet of Vehicles has become a key factor that affects the development of intelligent vehicles. However, existing security risk analysis methods for the IoV either focus only on certain levels, such as the component level, or perform only a static analysis. This paper proposes a dynamic attack graph generation method for the IoV to identify and visually display the security risks caused by the associated vulnerabilities in an IoV system. First, using the actual architecture of the IoV, this paper shows how to model the security elements and their relationships in the IoV system and proposes a network security ontology model for this system. Second, it shows how to construct a reasoning rule base according to the causal relationship between the vulnerabilities using the Semantic Web Rule Language Finally, in view of the rapid change in the network topology of the IoV, a dynamic attack graph generation algorithm based on an ontology reasoning engine is proposed, which can effectively reduce the overhead caused by the changes in the attack graph. The effectiveness of the algorithm is demonstrated through an actual security event scenario and a constructed scenario. The experimental results show that the algorithm can dynamically and accurately display the network attack graph of the IoV. The proposed method is helpful in globally analyzing the threat caused by the combined exploitation of the vulnerabilities in an IoV system and risk management.

Future Directions of Intelligent Vehicles: Potentials, Possibilities, and Perspectives

This is the brief report of the first IEEE Distributed/Decentralized Hybrid Workshop on Future Directions of Intelligent Vehicles (IEEE DHW-FDIV), part of the IEEE Distributed/Decentralized Hybrid Symposia on Intelligent Vehicles (IEEE DHS-IV) organized by the IEEE Transactions on Intelligent Vehicles (TIV). This DHW was conducted through two events on January 12 and February 7, 2022 with 23 and 12 participants from Asia, Europe, and North America, respectively. Various issues related to the current state of IEEE TIV and potential topics for future research and development of intelligent vehicles are addressed. Based on the suggestion of Professor Fei-Yue Wang, the new Editor-in-Chief of TIV, the first report of DHW-FDIV focuses on meta-vehicles and metaverses for smart mobility and intelligent transportation.

Effect of Cognitive Distraction on Physiological Measures and Driving Performance in Traditional and Mixed Traffic Environments

Distracted driving is a dominant cause of traffic accidents. In addition, with the rapid development of intelligent vehicles, mixed traffic environments are expected to become more complicated with multiple types of intelligent vehicles sharing the road, thereby increasing the opportunities for distracted driving. However, the existing research on detecting driver distraction in mixed traffic environments is limited. Therefore, in this study, we analysed the effect of cognitive distraction on the driver physiological measures and driving performance in traditional and mixed traffic environments and compared the parameters extracted in the two environments. Sixty drivers were involved in the data collection, which included normal driving and two distracting tasks while driving in a simulator. Repeated-measures analysis of variance (ANOVA) was performed to examine the effect of cognitive distraction and traffic environments on all parameters. The results indicate that the effects of the pupil diameter, standard deviations (SDs) of the horizontal and vertical fixation angles, blink frequency, speed, SD of the lane positioning (SDLP), SD of the steering wheel angle (SDSWA), and steering entropy (SE) were significant. These findings provide a theoretical foundation for identifying the most appropriate parameters to detect cognitive distraction in traditional and mixed traffic environments to help reduce traffic accidents.

Can driving condition prompt systems improve passenger comfort of intelligent vehicles? A driving simulator study

The main trend in the development of intelligent vehicles has been on ensuring comfort, safety, efficiency, and environmental sustainability. However, current research focuses primarily on the safety and energy saving of intelligent vehicles, and a comfortable driving experience through a human–machine interaction system has not been sufficiently investigated. This study used a high-fidelity 6-degree-of-freedom driving simulator to evaluate the impact of an independently-designed vehicle driving condition prompt (DCP) systems on subjective passenger comfort and motion sickness. The experiment showed that when future driving information is obtained through the vehicle DCP systems, the passengers' subjective comfort is improved, motion sickness is alleviated, and the degree of passenger posture instability is reduced. These conclusions contribute toward improving the comfort of autonomous vehicles and providing a reference for the future design of human–machine interaction systems for intelligent vehicles.

Comparative Analysis of Machine Learning-Based Approaches for Anomaly Detection in Vehicular Data

The rapid growth of demand for transportation, both for people and goods, as well as the massive accumulation of population in urban centers has augmented the need for the development of smart transport systems. One of the needs that have arisen is to efficiently monitor and evaluate driving behavior, so as to increase safety, provide alarms, and avoid accidents. Capitalizing on the evolution of Information and Communication Technologies (ICT), the development of intelligent vehicles and platforms in this domain is getting more feasible than ever. Nowadays, vehicles, as well as highways, are equipped with sensors that collect a variety of data, such as speed, acceleration, fuel consumption, direction, and more. The methodology presented in this paper combines both advanced machine learning algorithms and open-source based tools to correlate different data flows originating from vehicles. Particularly, the data gathered from different vehicles are processed and analyzed with the utilization of machine learning techniques in order to detect abnormalities in driving behavior. Results from different suitable techniques are presented and compared, using an extensive real-world dataset containing field measurements. The results feature the application of both supervised univariate anomaly detection and unsupervised multivariate anomaly detection methods in the same dataset.

Impact of adaptive cruise control (ACC) system on fatality and injury reduction in China

Objectives The road traffic safety situation around the world is not optimistic. The development of intelligent vehicles has become an ideal way to reduce road traffic crashes. The adaptive cruise control (ACC) system is an effective intelligent vehicle active safety system for avoiding certain types of collisions. This study aimed to assess the safety benefits of ACC in China, including the potential maximum impact and realistic impact. Methods This study applies a national-level safety impact evaluation model to assess the safety benefits of ACC in China, including the potential maximum impact and realistic impact. Road traffic fatality and severe injury trends in China, proportion of different collision types in China, effectiveness of collision avoidance, and market penetration rate of ACC are considered in the potential maximum impact scenario. Furthermore, the ACC activation rate and the technology’s technical limitations, including its effectiveness in different weather, light, speed, and road conditions, are discussed in the realistic scenario. Results With a 100% market penetration rate, fatalities could be reduced by 5.48%, and injuries could be reduced by 4.91%. With a large increase in market penetration rate of ACC in the coming future, the reductions in fatalities and severe injuries are 324–957 and 1,035–2,737 in 2025 and 531–1,579 and 1,604–4,242 in 2030. Considering ACC’s activation rate and its 4 main limitations, the adjusted realistic result is approximately one-third of the potential maximum result. Conclusions The result clearly shows that the ACC system can improve road traffic safety in China. Technical limitations have a great impact on ACC’s safety benefits. Of all of the limiting factors, the turn-on rate provides the most room for improvement, and improving the suitability of the ACC system on curved and sloped roads provides the smallest effect.

Electroencephalogram Authentication Access Control to Smart Car

In recent years, with the development of intelligent vehicles, the demand for security will be more and more big. One of the most important solutions is the use of new biometric technology. This paper presents an identity authentication system based on electroencephalogram (EEG) signals. The overall goal of this research is to design a new authentication method and develop the corresponding application. Therefore, we carried out a series of EEG experiments, and analyzed and discussed the experimental results. Based on these results, we study the algorithm of recognizing EEG signal features. Depended on the uniqueness of their EEG signals to be capable of authenticating access control to car, we build and present an Access Control System. The accuracy of authentication system is more than 87.3%.

Online operations of automated electric taxi fleets: An advisor-student reinforcement learning framework

Automation and electrification are inevitable trends in the development of intelligent vehicles. It is envisioned that automated electric taxis (AETs) will play an important role in future transportation systems for serving personalized travel demands. To tackle the operational challenges caused by the high spatiotemporal heterogeneity of customer demands entails novel online strategy to intelligently manage AET fleet. This study proposes an advisor-student reinforcement learning framework to solve the online operations problem of AET fleet through which the taxis are intelligently assigned to serve demands, dispatched to zones with excessive future demands, and forced to get refueled at charging stations. Extensive numerical experiments illustrate the advantages of the proposed framework over myopic and nearest distance greedy strategies, especially when vehicle relocation is highly needed.

A Vision-Based Driver Assistance System with Forward Collision and Overtaking Detection.

One major concern in the development of intelligent vehicles is to improve the driving safety. It is also an essential issue for future autonomous driving and intelligent transportation. In this paper, we present a vision-based system for driving assistance. A front and a rear on-board camera are adopted for visual sensing and environment perception. The purpose is to avoid potential traffic accidents due to forward collision and vehicle overtaking, and assist the drivers or self-driving cars to perform safe lane change operations. The proposed techniques consist of lane change detection, forward collision warning, and overtaking vehicle identification. A new cumulative density function (CDF)-based symmetry verification method is proposed for the detection of front vehicles. The motion cue obtained from optical flow is used for overtaking detection. It is further combined with a convolutional neural network to remove repetitive patterns for more accurate overtaking vehicle identification. Our approach is able to adapt to a variety of highway and urban scenarios under different illumination conditions. The experiments and performance evaluation carried out on real scene images have demonstrated the effectiveness of the proposed techniques.

Design of electroencephalogram authentication access control to smart car.

In recent years, with the development of intelligent vehicles, the demand for security will be bigger and bigger. One of the most important solutions is the use of new biometric technology. At present, there are still some areas to be improved on biometric technology. For example, diseases will destroy some biological characteristics, some detection methods are too slow, many detection methods do not need living detection, and so on. Electroencephalogram (EEG) is a new biometric tool for living identification. In this Letter, a kind of identity authentication system based on the EEG signal is presented. The overall goal of this research is to design a new authentication method and develop the corresponding application. Therefore, the authors carried out a series of EEG experiments, and analysed and discussed the experimental results. Based on these results, they build and present an access control system based on the uniqueness of their EEG signals to be capable of authenticating access control to the car. The accuracy of the authentication system is >87.3%.

Hardware in the loop framework proposal for a semi-autonomous car architecture in a closed route environment

The development of intelligent vehicles has been increasing at great speed in recent years, which has allowed to improve their capabilities in autonomous driving systems. Many of these features are related to advanced driving assistance systems and autonomous driving systems. This capability improvement, has been achieved because of recent developments of automation oriented software and hardware. Such improvements, allowed the vehicle to achieve a more precise perception of it’s working environment. For this improvement it is important the integration and simulation of the systems in different configurations, such as hardware-in-the-loop, software-in-the-loop and model-in-the-loop. In this paper, we present a framework proposal that allows the design and testing of computer vision and control systems for the partial automation of a vehicle, with the use of hardware and software systems in the loop. This proposal is focused on the rapid experimental development of these systems for the implementation by autonomous vehicle designers and engineers. Our proposal allows faster data capture, either in a real or simulated environment, to improve and optimize data training with machine learning algorithms; this proposal integrates several open source systems and hardware with the necessary capacity for real-time implementation.

Research on Machine Learning Method for Intelligent Driving

Machine learning technology is widely used in intelligent driving of vehicles. Intelligent driving based on machine learning is one of the directions and difficulties in the development of intelligent vehicles. Firstly, this paper analyses the related concepts of machine learning. Then this paper analyses the research of machine learning under intelligent driving. Finally, this paper analyses the research of machine learning algorithm.

An Efficient and Safe Road Condition Monitoring Authentication Scheme Based on Fog Computing

In recent years, with the development of intelligent vehicles and wireless sensor network technology, the research on road safety has attracted much attention in vehicular ad-hoc networks (VANETs). By sensing events on the road, vehicles can broadcast information to inform others of traffic jams or accidents. However, the mobile vehicle network has a large transmission delay, which makes real-time content transmission impossible. In this paper, a new certificateless aggregate signcryption scheme (CLASC) is proposed by using a fog computing framework that supports mobility, low latency, and location awareness. It is combined with online/offline encryption (OOE) technology, which reduces many time-consuming operations and improves the security of vehicle users and the efficiency of message authentication. In addition, the scheme has the characteristics of mutual authentication, anonymity, untraceability, and nondeniability. Based on the difficulty of the discrete logarithm problem (DLP) and the computational Diffie-Hellman (CDH) problem, the scheme is further proved to be unforgeability and confidentiality under the random oracle model. The simulation results show that compared with the existing schemes, this scheme can not only ensure the security requirements of the system but also achieve higher efficiency in computing and communication.

The Negative Impact of Vehicular Intelligence on Energy Consumption

The development of intelligent vehicle will provide the Chinese automotive industry with a strategic opportunity for transformation and upgrading. Vehicular intelligence provides new solutions for energy conservation and emissions mitigation. However, the process of vehicular intelligence is progressive. The saving of energy consumption depends on the high smart car market penetration rate. But one thing that can be confirmed is that intelligent vehicles are equipped with advanced sensors, controllers, and actuators, in combination with connecting communication technologies compared with conventional vehicles, for which the energy consumption of the vehicle will definitely increase. In this study, vehicle fuel consumption cost at different levels of intelligence is calculated, considering the energy consumption of hardware used for automation and connecting functions, the energy consumption cost generated by the quality of the hardware, and the wind resistance. The results reveal that the energy consumption per 100 kilometers of an intelligent vehicle ranges from 0.78L to 1.86L, more than traditional vehicle. The energy consumption cost of automation functions is much higher than that of the connecting functions. Computing platform performance, connection strength, and radar performance are the three main factors that affect energy consumption cost. Based on the analysis, the high energy consumption cost of vehicular intelligence has a profound impact on choosing power platform.

Lane Detection Algorithm for Intelligent Vehicles in Complex Road Conditions and Dynamic Environments.

Lane detection is an important foundation in the development of intelligent vehicles. To address problems such as low detection accuracy of traditional methods and poor real-time performance of deep learning-based methodologies, a lane detection algorithm for intelligent vehicles in complex road conditions and dynamic environments was proposed. Firstly, converting the distorted image and using the superposition threshold algorithm for edge detection, an aerial view of the lane was obtained via region of interest extraction and inverse perspective transformation. Secondly, the random sample consensus algorithm was adopted to fit the curves of lane lines based on the third-order B-spline curve model, and fitting evaluation and curvature radius calculation were then carried out on the curve. Lastly, by using the road driving video under complex road conditions and the Tusimple dataset, simulation test experiments for lane detection algorithm were performed. The experimental results show that the average detection accuracy based on road driving video reached 98.49%, and the average processing time reached 21.5 ms. The average detection accuracy based on the Tusimple dataset reached 98.42%, and the average processing time reached 22.2 ms. Compared with traditional methods and deep learning-based methodologies, this lane detection algorithm had excellent accuracy and real-time performance, a high detection efficiency and a strong anti-interference ability. The accurate recognition rate and average processing time were significantly improved. The proposed algorithm is crucial in promoting the technological level of intelligent vehicle driving assistance and conducive to the further improvement of the driving safety of intelligent vehicles.

Privacy-Preserving Content Dissemination for Vehicular Social Networks: Challenges and Solutions

Vehicular social networks (VSNs), viewed as the integration of traditional vehicular networks and social networks, are promising communication platforms based on the development of intelligent vehicles and deployment of intelligent transportation systems. Passengers can obtain information by searching over Internet or querying vehicles in proximity through intra-vehicle equipment. Hence, the performance of content dissemination in VSNs heavily relies on inter-vehicle communication and human behaviors. However, privacy preservation always conflicts with the usability of individual information in VSNs. The highly dynamic topology and increasing kinds of participants lead to potential threats for communication security and individual privacy. Therefore, the privacy-preserving solutions for content dissemination in VSNs have become extremely challenging, and numerous researches have been conducted recently. Compared with related surveys, this article provides the unique characteristics of privacy-preserving requirements and solutions for content dissemination in VSNs. It focuses on: 1) a comprehensive overview of content dissemination in VSNs; 2) the privacy issues and potential attacks related to content dissemination; and 3) the corresponding solutions based on privacy consideration. First, the characteristics of VSNs, content dissemination and its solutions in VSNs are revealed. Second, the privacy issues for content dissemination in the current VSN architecture are analyzed and classified according to their features. Various privacy-preserving content dissemination schemes, attempting to resist distinct attacks, are also discussed. Finally, the research challenges and open issues are summarized.

The reliability of Ag wedge bonding with various bonding pads for power devices

Abstract Recent years, with the development of intelligent vehicles, power devices would be widely used. But traditional Si power devices have problems under high temperature. Now, there is a tendency that using SiC instead of Si in power devices, as SiC has a higher band gap 3.25 eV comparing with 1.12 eV of Si. So that SiC power devices can withstand higher temperature and voltage. Also SiC power devices have the advantages of lower parasitic parameters, smaller device size and shorter response time. In this study, to investigate the reliability of Ag wedge bonding, Ag wires were bonded on Al pad and Au-Ni pad. And then we did shear test after high temperature storage life test (HTSL). Finally, we used energy dispersive X-ray spectroscopy (EDS) to do cross-section observation. The results show that, for Al pad, the shear strength has decreased after annealing at 300°C with mold packages, cracks and corrosion were observed. For Au-Ni pad, the shear strength has increased after annealing, and no cracks or corrosion formed. So Ag bonding wire is proposed as an alternative to Al bonding wire for selected metal pads in power devices.