Intelligent Vehicle Navigation Research Articles

Stable training via elastic adaptive deep reinforcement learning for autonomous navigation of intelligent vehicles

The uncertain stability of deep reinforcement learning training on complex tasks impedes its development and deployment, especially in intelligent vehicles, such as intelligent surface vessels and self-driving cars. Complex and varied environmental states puzzle training of decision-making networks. Here we propose an elastic adaptive deep reinforcement learning algorithm to address these challenges and achieve autonomous navigation in intelligent vehicles. Our method trains the decision-making network over the function and optimization learning stages, in which the state space and action space of autonomous navigation tasks are pruned by choosing classic states and actions to reduce data similarity, facilitating more stable training. We introduce a task-adaptive observed behaviour classification technique in the function learning stage to divide state and action spaces into subspaces and identify classic states and actions. In which the classic states and actions are accumulated as the training dataset that enhances its training efficiency. In the subsequent optimization learning stage, the decision-making network is refined through meticulous exploration and accumulation of datasets. The proposed elastic adaptive deep reinforcement learning enables the decision-making network to effectively learn from complex state and action spaces, leading to more efficient training compared to traditional deep reinforcement learning approaches. Simulation results demonstrate the remarkable effectiveness of our method in training decision-making networks for intelligent vehicles. The findings validate that our method provides reliable and efficient training for decision-making networks in intelligent vehicles. Moreover, our method exhibits stability in training other tasks characterized by continuous state and action spaces.

Intelligent vehicle navigation systems and autonomous driving technology: A comprehensive analysis

This paper conducts a comprehensive study and analysis of intelligent vehicle navigation systems and autonomous driving technology. We review the historical development of autonomous driving technology, discuss key concepts such as perception, decision-making, and control, explore various types of autonomous vehicles, and examine various aspects of intelligent vehicle navigation systems. Additionally, we investigate safety, reliability, legal frameworks in the field of autonomous driving, as well as future trends and ethical considerations. Finally, we summarize the main findings of the research and provide recommendations for future studies.

Intelligent vehicle localization and navigation based on intersection fingerprint roadmap (IRM) in underground parking lots

The localization and navigation of underground parking lots presents a ‘last mile’ problem for intelligent vehicles. To address the issues of poor localization and navigation efficiency caused by the lack of global navigation satellite system signals in this scenario, this study presents a low-cost and computationally efficient real-time localization and navigation algorithm. The algorithm begins by constructing a node fingerprint based on the characteristics of underground parking lot intersections and proposes an intersection fingerprint roadmap (IRM). To address the problem of difficult initial position computation during navigation, a node-level localization method based on scene matching and pose calculation is proposed and then combined with the IRM. Finally, the IRM is used to search for routes on the scene plan and perform path smoothing to obtain a globally feasible driving route. The algorithm proposed in this study can achieve fast vehicle localization and real-time navigation with low computational power, and the proposed localization method can be applied not only to underground parking lots but also to other outdoor scenes. The method is evaluated on datasets collected from different environments, and the experimental results show that the algorithm has an average localization error of approximately 20 cm and provides faster navigation speed and smoother navigation in comparison with other algorithms.

Research on optimal path planning technology for vehicle positioning and navigation system

Abstract Intelligent Transportation System (ITS) is created in the context of the full development and progress of contemporary science and technology, aiming to help people make the best choice, drive the best vehicle, and control the best road network traffic. Intelligent vehicle positioning and navigation system is a high-tech integrated system that applies automatic vehicle positioning technology, geographic information system and database technology, computer technology, multimedia technology, and modern communication technology. In this paper, based on the research of GPS and GIS technology, by analyzing and comparing the advantages and disadvantages of different optimal path planning algorithms, we propose to optimize the A* algorithm by using the minimum heap sorting method for nodes and combining the twoway search and hierarchical search strategies to complete the improvement of the A* algorithm. The experimental results show that the path search speed of the improved algorithm has been significantly improved and the path design optimization is obvious. The improvement algorithm proposed in this paper is feasible and reasonable, which is good for upgrading the in-vehicle navigation system and relieving urban traffic congestion.

Research on the Lane Recognition Method Based on Computer Vision

The lane line is the most important traffic sign in road traffic and plays a significant function in restricting and guaranteeing the running of vehicles. Whether in the vehicle safety driving system or in the intelligent vehicle navigation based on machine vision, lane detection and recognition is a basic and necessary function module. This can enable future in-depth studies on intelligent transportation while also lowering the likelihood of traffic accidents.

A Security and Privacy Preserved Intelligent Vehicle Navigation System

With emergence of smart city concept, intelligent transportation becomes one of its key enablers, with navigation as an indispensable component. Vehicular ad hoc network (VANET) based systems emerge to provide navigation services to vehicle users, which is characterized by employing Roadside Units (RSUs) to collect road condition information from nearby vehicles. The querying vehicle exploits the information retrieved through the RSUs to determine an optimal route. We focus on addressing two key issues in such system. First, while querying, the destination of the vehicle user should not be leaked to the queried RSU and trusted authority (TA), which is likely to be misused. Second, the navigation process should be independent of participation of the centralized TA, which can be a single point of failure. To address these issues, this article proposes a security and privacy preserved intelligent vehicle navigation system, adopting a distributed approach, to retrieve navigation without revealing the destination. The security of the system has been thoroughly evaluated using an adversary model, BAN Logic and AVISPA tool. In addition, the performance of the proposed system has been evaluated, and compared with the existing systems. The comparison illustrates that the proposed system is computationally efficient than the existing systems.

Deep Network-Enabled Haze Visibility Enhancement for Visual IoT-Driven Intelligent Transportation Systems

The Internet of Things (IoT) has recently emerged as a revolutionary communication paradigm where a large number of objects and devices are closely interconnected to enable smart industrial environments. The tremendous growth of visual sensors can significantly promote the traffic situational awareness, traffic safety management, and intelligent vehicle navigation in intelligent transportation systems (ITSs). However, due to the absorption and scattering of light by the turbid medium in atmosphere, the visual IoT inevitably suffers from imaging quality degradation, e.g., contrast reduction, color distortion, etc. This negative impact can not only reduce the imaging quality, but also bring challenges for the deployment of several high-level vision tasks (e.g., object detection, tracking, recognition, etc.) in the ITS. To improve imaging quality under the hazy environment, we propose a deep network-enabled three-stage dehazing network (termed TSDNet) for promoting the visual IoT-driven ITS. In particular, the proposed TSDNet mainly contains three parts, i.e., multiscale attention module for estimating the hazy distribution in the RGB image domain, two-branch extraction module for learning the hazy features, and multifeature fusion module for integrating all characteristic information and reconstructing the haze-free image. Numerous experiments have been implemented on synthetic and real-world imaging scenarios. Dehazing results illustrated that our TSDNet remarkably outperformed several state-of-the-art methods in terms of both qualitative and quantitative evaluations. The high-accuracy object detection results have also demonstrated the superior dehazing performance of the TSDNet under hazy atmosphere conditions. The source code is available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/gy65896/TSDNet</uri> .

Adaptive Kalman filter based on integer ambiguity validation in moving base RTK

In high-precision dynamic positioning, it is necessary to ensure the positioning accuracy and reliability of the navigation system, especially for safety–critical applications, such as intelligent vehicle navigation. In the face of a complex observation environment, when the global navigation satellite system (GNSS) uses carrier phase observations for high-precision relative positioning, ambiguity resolution will be affected, and it is difficult to estimate all ambiguities. In addition, when the GNSS signal quality and measurement noise level are difficult to predict in an environment with many occlusions, the received satellite observations are prone to very large errors, resulting in apparent deviations in the positioning solution. However, traditional positioning algorithms assume that the measurement noise is constant, which is unrealistic. This will cause incorrect ambiguity resolution, lead to meter-level positioning errors, reduce the reliability of the system, and increase the integrity risk of the system. We proposed an innovative adaptive Kalman filter based on integer ambiguity validation (IAVAKF) to improve the efficiency of ambiguity resolution (AR) and positioning accuracy. The partial ambiguity resolution (PAR) method is applied to solve the integer ambiguities. Then, the accuracy of the fixed ambiguity is verified by the ambiguity success rate. Taking the ambiguity success rate as a dynamic adjustment factor, the measurement noise matrix and variance–covariance matrix of the state estimation is adaptively adjusted at each time interval in the Kalman filter to provide a smoothing effect for filtering. The optimal Kalman filter gain matrix is obtained to improve positioning accuracy and reliability. As a result, the static and dynamic vehicle experiments show that the positioning accuracy of the proposed IAVAKF is improved by 26% compared with the KF. Through the IAVAKF, a more realistic PL can be obtained and applied to evaluate the integrity of the navigation system in the position domain. It can reduce the false alarm rate by 2.45% and 1.85% in the horizontal and vertical directions, respectively.

STMGCN: Mobile Edge Computing-Empowered Vessel Trajectory Prediction Using Spatio-Temporal Multigraph Convolutional Network

The revolutionary advances in machine learning and data mining techniques have contributed greatly to the rapid developments of maritime Internet of Things (IoT). In maritime IoT, the spatio-temporal vessel trajectories, collected from the hybrid satellite-terrestrial automatic identification system (AIS) base stations, are of considerable importance for promoting traffic situation awareness and vessel traffic services, etc. To guarantee traffic safety and efficiency, it is essential to robustly and accurately predict the AIS-based vessel trajectories (i.e., the future positions of vessels) in maritime IoT. In this work, we propose a spatio-temporal multigraph convolutional network (STMGCN)-based trajectory prediction framework using the mobile edge computing (MEC) paradigm. Our STMGCN is mainly composed of three different graphs, which are, respectively, reconstructed according to the social force, the time to closest point of approach, and the size of surrounding vessels. These three graphs are then jointly embedded into the prediction framework by introducing the spatio-temporal multigraph convolutional layer. To further enhance the prediction performance, the self-attention temporal convolutional layer is proposed to further optimize STMGCN with fewer parameters. Owing to the high interpretability and powerful learning ability, STMGCN is able to achieve superior prediction performance in terms of both accuracy and robustness. The reliable prediction results are potentially beneficial for traffic safety management and intelligent vehicle navigation in MEC-enabled maritime IoT.

Theoretical and experimental research on valve-less piezoelectric pump with arc-shaped tubes

In order to apply a valve-less piezoelectric pump in navigation equipment, this work first revealed a valve-less piezoelectric pump with arc-shaped tubes and then researched the relationship between the main structural parameters of arc-shaped tubes and the output flow rate. According to the velocity difference between clockwise and counterclockwise directions of liquid in circular arc-shaped tubes, the flow characteristics of liquid in circular arc-shaped tubes are analyzed in this paper, and the velocity difference model and flow equation of a piezoelectric pump are established. By using three-dimensional printing technology, five types of prototypes with different parameters were made, and the experimental research was carried out. The results show that at the optimal frequency, the pump output flow is the largest. The output flow is proportional to the driving voltage, the base circle radius, and the cross section area of the pipe. The low-cost and miniaturized navigation equipment made by using the piezoelectric pump has potential application value in unmanned aerial vehicle and intelligent vehicle navigation.

Vision-based mapping of lane semantics and topology for intelligent vehicles

High-definition map is an essential tool for route measurement, planning and navigation of intelligent vehicles. Yet its creation is still a persisting challenge, especially in creating the semantic and topology layer of the map based on visual sensing. However, current semantic mapping approaches do not consider the map applicability in navigation tasks while the topology mapping approaches face the issues of limited location accuracy or expensive hardware cost. In this paper, we propose a joint mapping framework for both semantic and topology layers, which are learned in a lane-level and based on a monocular camera sensor and an on-board GPS positioning device. A map management approach “RoadSegDict” is also proposed to support the efficient updating of semantic map in a crowdsourced manner. Moreover, a new dataset is proposed, which includes a variety of lane structures with detailed semantic and topology annotations.

Research on control strategy of multi-source data fusion solar intelligent vehicle based on image recognition

Abstract Nowadays, the global energy and environmental problems are becoming more and more serious, which promotes the development and utilization of renewable and clean energy in various countries. Intelligent car involves many subjects such as electronic technology, artificial intelligence, automatic control technology, sensor technology and computer technology and has become an important part of the application of artificial intelligence. Solar cell is a necessary part of the normal operation of the solar intelligent car, which can provide clean energy for the intelligent car. In this paper, the image recognition technology is used to design the intelligent vehicle control system. According to the intelligent vehicle path recognition, the scale invariant feature transform (SIFT) algorithm is improved to improve the accuracy of intelligent vehicle recognition. Data fusion is used to process the data detected by multi-sensor, and the running state of intelligent vehicle is studied. An evaluation method of intelligent vehicle navigation parameters based on association rules and belief network is proposed. The maximum power point tracking control is realized by using the interference observation method to ensure that the intelligent vehicle can track the maximum power point of the solar cell.

Intelligent Vehicle Navigation System Based on Visual Detection and Positioning

This design is a set of intelligent vehicle navigation system which relies on visual detection and positioning. The image is processed and the coordinates of beacon and car body are extracted by the camera hanging above. The coordinate information is transmitted to the intelligent vehicle and fused with the intelligent vehicle data to determine its own environment and running state. The system integrates image processing, wireless communication and intelligent path planning to realize the task of accurate recognition and automatic tracing.

A linear Kalman filter-based integrity monitoring considering colored measurement noise

Apart from positioning accuracy, the reliability of a navigation system is also significant, especially for safety–critical applications, such as intelligent vehicle navigation. Generally, the Global Navigation Satellite System (GNSS) positioning algorithm assumed that measurement noise is uncorrelated white noise. However, the appearance of colored noise, which comes from various noise sources, does not follow the Gaussian white noise assumption in the Kalman filter and would degrade both the accuracy and reliability of positioning. To deal with this problem, we propose a linear Kalman filter-based integrity monitoring method, which is based on a linear colored Kalman filter (CKF) considering measurement time correlation colored noise by a first-order Gauss–Markov model. Both simulated and real dynamic experiments were conducted to test the proposed algorithm. The results proved that CKF is capable of modeling time correlation with a realistic filter covariance in both simulated static and in-field dynamic tests. Furthermore, the performance of an integrity monitoring system with a protection level based on the CKF has proved to be more feasible and effective to bound position errors. Besides, the simulated results demonstrate that it can typically reduce false alarm by 24.81% in the horizontal direction and 39.47% in the vertical direction in a simulated experiment. The dynamic test shows that the proposed method can reduce the false alarm rate by 22.11% and 15.62% in the horizontal and vertical directions, respectively.

Behavior Prediction of Traffic Actors for Intelligent Vehicle Using Artificial Intelligence Techniques: A Review

Intelligent vehicle technology has made tremendous progress due to Artificial Intelligence (AI) techniques. Accurate behavior prediction of surrounding traffic actors is essential for the safe and secure navigation of the intelligent vehicle. Minor misbehavior of these vehicles on the busy roads may lead to an accident. Due to this, there is a need for vehicle behavior research work in today’s era. This research article reviews traffic actors’ behavior prediction techniques for intelligent vehicles to perceive, infer, and anticipate other vehicles’ intentions and future actions. It identifies the key strategies and methods for AI, emerging trends, datasets, and ongoing research issues in these fields. As per the authors’ knowledge, this is the first systematic literature review dedicated to the vehicle behavior study examining existing academic literature published by peer review venues between 2011 and 2021. A systematic review was undertaken to examine these papers, and five primary research questions have been addressed. The findings show that using sophisticated input representation that includes traffic rules and road geometry, artificial intelligence-based solutions applied to behavior prediction of traffic actors for intelligent vehicles have shown promising success, particularly in complex driving scenarios. Finally, the paper summarizes the most widely used approaches in behavior prediction of traffic actors for intelligent vehicles, which the authors believe serves as a foundation for future research in behavior prediction of surrounding traffic actors for secure and accurate intelligent vehicle navigation.

Road obstacles positional and dynamic features extraction combining object detection, stereo disparity maps and optical flow data

One of the most relevant tasks in an intelligent vehicle navigation system is the detection of obstacles. It is important that a visual perception system for navigation purposes identifies obstacles, and it is also important that this system can extract essential information that may influence the vehicle's behavior, whether it will be generating an alert for a human driver or guide an autonomous vehicle in order to be able to make its driving decisions. In this paper we present an approach for the identification of obstacles and extraction of class, position, depth and motion information from these objects that employs data gained exclusively from passive vision. We performed our experiments on two different data-sets and the results obtained shown a good efficacy from the use of depth and motion patterns to assess the obstacles' potential threat status.

Real-time deep reinforcement learning based vehicle navigation

Traffic congestion has become one of the most serious contemporary city issues as it leads to unnecessary high energy consumption, air pollution and extra traveling time. During the past decade, many optimization algorithms have been designed to achieve the optimal usage of existing roadway capacity in cities to leverage the problem. However, it is still a challenging task for the vehicles to interact with the complex city environment in a real time manner. In this paper, we propose a deep reinforcement learning (DRL) method to build a real-time intelligent vehicle routing and navigation system by formulating the task as a sequence of decisions. In addition, an integrated framework is provided to facilitate the intelligent vehicle navigation research by embedding smart agents into the SUMO simulator. Nine realistic traffic scenarios are simulated to test the proposed navigation method. The experimental results have demonstrated the efficient convergence of the vehicle navigation agents and their effectiveness to make optimal decisions under the volatile traffic conditions. The results also show that the proposed method provides a better navigation solution comparing to the benchmark routing optimization algorithms. The performance has been further validated by using the Wilcoxon test. It is found that the achieved improvement of our proposed method becomes more significant under the maps with more edges (roads) and more complicated traffics comparing to the state-of-the-art navigation methods.

Speed profile optimisation for intelligent vehicles in dynamic traffic scenarios

In the autonomous navigation of intelligent vehicles, collision avoidance is essential for driving safety. Similar to the driving preference of human, the driving path and speed can be determined separately. This paper is concerned with speed profile optimisation problem for dynamic obstacle avoidance given the reference path. The optimisation consists of smoothness, risk, and efficiency terms with obstacle constraints. For task formulation, the s-t motion space is constructed to describe the motion of the ego vehicle and obstacles. Then the high-dimensional trajectory space is mapped to the low-dimensional s-t space for computational efficiency. The speed optimisation problem is transformed into a path searching problem considering collision avoidance and searching efficiency. RRT-based algorithm is proposed to search for the optimal speed profile in the s-t space asymptotically. In each searching step, node extension strategy is designed for the space exploring efficiency; then the tree structure is locally refined for asymptotic optimisation. The optimal speed profile is generated after the searching process converges and the speed profile is planned periodically. For performance evaluation, simulation tests in typical traffic conditions are conducted based on the SUMO (Simulation of Urban MObility) platform. Results show the effectiveness and efficiency of this method.

Passive vision road obstacle detection: a literature mapping

ABSTRACT Obstacle detection is one of the main tasks in intelligent vehicle navigation systems. Several research works focused on the use of passive vision (cameras) to accomplish this task have been published. In this paper we present a literature mapping of the state of the art in road obstacle detection using frontal images. This mapping is based upon a systematic literature review that took into consideration papers published between 2007 and 2019. We analyze approaches based upon methods such as image segmentation, stereo vision, optical flow and neural networks and classify them accordingly to their characteristics and detection targets, such as vehicles, pedestrians or obstacles in general. We also inspect if they are performing pavement defects detection, such as potholes, puddles or other types of damage. The detection of pavement problems is important for the reality of in-development countries, which in many cases do not present well-maintained roads and may represent a threat to vehicular navigation. With this mapping we can identify the current state in this research area and also discuss future steps.

Research on colorized physical terrain modeling for intelligent vehicle navigation

Colorized physical terrain models are needed in many applications, such as intelligent navigation, military strategy planning, landscape architecting, and land-use planning. However, current terrain elevation information is stored as digital elevation model file format, and terrain color information is generally stored in aerial images. A method is presented to directly convert the digital elevation model file and aerial images of a given terrain to the colorized virtual three-dimensional terrain model, which can be processed and fabricated by color three-dimensional printers. First, the elevation data and color data were registered and fused. Second, the colorized terrain surface model was created by using the virtual reality makeup language file format. Third, the colorized three-dimensional terrain model was built by adding a base and four walls. Finally, the colorized terrain physical model was fabricated by using a color three-dimensional printer. A terrain sample with typical topographic features was selected for analysis, and the results demonstrated that the colorized virtual three-dimensional terrain model can be constructed efficiently and the colorized physical terrain model can be fabricated precisely, which makes it easier for users to understand and make full use of the given terrain.