A journal of IEEE and CAA , publishes high-quality papers in English on original theoretical/experimental research and development in all areas of automation

Vol. 10,  No. 9, 2023

Display Method:
How Generative Adversarial Networks Promote the Development of Intelligent Transportation Systems: A Survey
Hongyi Lin, Yang Liu, Shen Li, Xiaobo Qu
2023, 10(9): 1781-1796. doi: 10.1109/JAS.2023.123744
Abstract(692) HTML (88) PDF(1438)
In current years, the improvement of deep learning has brought about tremendous changes: As a type of unsupervised deep learning algorithm, generative adversarial networks (GANs) have been widely employed in various fields including transportation. This paper reviews the development of GANs and their applications in the transportation domain. Specifically, many adopted GAN variants for autonomous driving are classified and demonstrated according to data generation, video trajectory prediction, and security of detection. To introduce GANs to traffic research, this review summarizes the related techniques for spatio-temporal, sparse data completion, and time-series data evaluation. GAN-based traffic anomaly inspections such as infrastructure detection and status monitoring are also assessed. Moreover, to promote further development of GANs in intelligent transportation systems (ITSs), challenges and noteworthy research directions on this topic are provided. In general, this survey summarizes 130 GAN-related references and provides comprehensive knowledge for scholars who desire to adopt GANs in their scientific works, especially transportation-related tasks.
Adaptive Multi-Step Evaluation Design With Stability Guarantee for Discrete-Time Optimal Learning Control
Ding Wang, Jiangyu Wang, Mingming Zhao, Peng Xin, Junfei Qiao
2023, 10(9): 1797-1809. doi: 10.1109/JAS.2023.123684
Abstract(554) HTML (71) PDF(71)
This paper is concerned with a novel integrated multi-step heuristic dynamic programming (MsHDP) algorithm for solving optimal control problems. It is shown that, initialized by the zero cost function, MsHDP can converge to the optimal solution of the Hamilton-Jacobi-Bellman (HJB) equation. Then, the stability of the system is analyzed using control policies generated by MsHDP. Also, a general stability criterion is designed to determine the admissibility of the current control policy. That is, the criterion is applicable not only to traditional value iteration and policy iteration but also to MsHDP. Further, based on the convergence and the stability criterion, the integrated MsHDP algorithm using immature control policies is developed to accelerate learning efficiency greatly. Besides, actor-critic is utilized to implement the integrated MsHDP scheme, where neural networks are used to evaluate and improve the iterative policy as the parameter architecture. Finally, two simulation examples are given to demonstrate that the learning effectiveness of the integrated MsHDP scheme surpasses those of other fixed or integrated methods.
Position Errors and Interference Prediction-Based Trajectory Tracking for Snake Robots
Dongfang Li, Yilong Zhang, Ping Li, Rob Law, Zhengrong Xiang, Xin Xu, Limin Zhu, Edmond Q. Wu
2023, 10(9): 1810-1821. doi: 10.1109/JAS.2023.123612
Abstract(585) HTML (71) PDF(190)
This work presents a trajectory tracking control method for snake robots. This method eliminates the influence of time-varying interferences on the body and reduces the offset error of a robot with a predetermined trajectory. The optimized line-of-sight (LOS) guidance strategy drives the robot’s steering angle to maintain its anti-sideslip ability by predicting position errors and interferences. Then, the predictions of system parameters and viscous friction coefficients can compensate for the joint torque control input. The compensation is adopted to enhance the compatibility of a robot within ever-changing environments. Simulation and experimental outcomes show that our work can decrease the fluctuation peak of the tracking errors, reduce adjustment time, and improve accuracy.
Transformer-Based Macroscopic Regulation for High-Speed Railway Timetable Rescheduling
Wei Xu, Chen Zhao, Jie Cheng, Yin Wang, Yiqing Tang, Tao Zhang, Zhiming Yuan, Yisheng Lv, Fei-Yue Wang
2023, 10(9): 1822-1833. doi: 10.1109/JAS.2023.123501
Abstract(322) HTML (67) PDF(127)
Unexpected delays in train operations can cause a cascade of negative consequences in a high-speed railway system. In such cases, train timetables need to be rescheduled. However, timely and efficient train timetable rescheduling is still a challenging problem due to its modeling difficulties and low optimization efficiency. This paper presents a Transformer-based macroscopic regulation approach which consists of two stages including Transformer-based modeling and policy-based decision-making. Firstly, the relationship between various train schedules and operations is described by creating a macroscopic model with the Transformer, providing the better understanding of overall operation in the high-speed railway system. Then, a policy-based approach is used to solve a continuous decision problem after macro-modeling for fast convergence. Extensive experiments on various delay scenarios are conducted. The results demonstrate the effectiveness of the proposed method in comparison to other popular methods.
A Length-Adaptive Non-Dominated Sorting Genetic Algorithm for Bi-Objective High-Dimensional Feature Selection
Yanlu Gong, Junhai Zhou, Quanwang Wu, MengChu Zhou, Junhao Wen
2023, 10(9): 1834-1844. doi: 10.1109/JAS.2023.123648
Abstract(331) HTML (49) PDF(122)
As a crucial data preprocessing method in data mining, feature selection (FS) can be regarded as a bi-objective optimization problem that aims to maximize classification accuracy and minimize the number of selected features. Evolutionary computing (EC) is promising for FS owing to its powerful search capability. However, in traditional EC-based methods, feature subsets are represented via a length-fixed individual encoding. It is ineffective for high-dimensional data, because it results in a huge search space and prohibitive training time. This work proposes a length-adaptive non-dominated sorting genetic algorithm (LA-NSGA) with a length-variable individual encoding and a length-adaptive evolution mechanism for bi-objective high-dimensional FS. In LA-NSGA, an initialization method based on correlation and redundancy is devised to initialize individuals of diverse lengths, and a Pareto dominance-based length change operator is introduced to guide individuals to explore in promising search space adaptively. Moreover, a dominance-based local search method is employed for further improvement. The experimental results based on 12 high-dimensional gene datasets show that the Pareto front of feature subsets produced by LA-NSGA is superior to those of existing algorithms.
Innovative Services for Electric Mobility Based on Virtual Sensors and Petri Nets
Agostino Marcello Mangini, Michele Roccotelli
2023, 10(9): 1845-1859. doi: 10.1109/JAS.2023.123699
Abstract(235) HTML (57) PDF(83)
About 60% of emissions into the earth’s atmosphere are produced by the transport sector, caused by exhaust gases from conventional internal combustion engines. An effective solution to this problem is electric mobility, which significantly reduces the rate of urban pollution. The use of electric vehicles (EVs) has to be encouraged and facilitated by new information and communication technology (ICT) tools. To help achieve this goal, this paper proposes innovative services for electric vehicle users aimed at improving travel and charging experience. The goal is to provide a smart service to allow drivers to find the most appropriate charging solutions during a trip based on information such as the vehicle’s current position, battery type, state of charge, nearby charge point availability, and compatibility. In particular, the drivers are supported so that they can find and book the preferred charge option according to time availability and the final cost of the charge points (CPs). To this purpose, two virtual sensors (VSs) are designed, modeled and simulated in order to provide the users with an innovative service for smart CP searching and booking. In particular, the first VS is devoted to locate and find available CPs in a preferred area, whereas the second VS calculates the charging cost for the EV and supports the driver in the booking phase. A UML activity diagram describes VSs operations and cooperation, while a UML sequence diagram highlights data exchange between the VSs and other electromobility ecosystem actors (CP operator, EV manufacturer, etc.). Furthermore, two timed Petri Nets (TPNs) are designed to model the proposed VSs, functioning and interactions as discrete event systems. The Petri Nets are synchronized by a single larger TPN that is simulated in different use cases and scenarios to demonstrate the effectiveness of the proposed VSs.
Cyclic-Pursuit-Based Circular Formation Control of Mobile Agents with Limited Communication Ranges and Communication Delays
Boyin Zheng, Cheng Song, Lu Liu
2023, 10(9): 1860-1870. doi: 10.1109/JAS.2023.123576
Abstract(330) HTML (93) PDF(129)
This article addresses the circular formation control problem of a multi-agent system moving on a circle in the presence of limited communication ranges and communication delays. To minimize the number of communication links, a novel distributed controller based on a cyclic pursuit strategy is developed in which each agent needs only its leading neighbour’s information. In contrast to existing works, we propose a set of new potential functions to deal with heterogeneous communication ranges and communication delays simultaneously. A new framework based on the admissible upper bound of the formation error is established so that both connectivity maintenance and order preservation can be achieved at the same time. It is shown that the multi-agent system can be driven to the desired circular formation as time goes to infinity under the proposed controller. Finally, the effectiveness of the proposed method is illustrated by some simulation examples.
AUTOSIM: Automated Urban Traffic Operation Simulation via Meta-Learning
Yuanqi Qin, Wen Hua, Junchen Jin, Jun Ge, Xingyuan Dai, Lingxi Li, Xiao Wang, Fei-Yue Wang
2023, 10(9): 1871-1881. doi: 10.1109/JAS.2023.123264
Abstract(332) HTML (39) PDF(97)
Online traffic simulation that feeds from online information to simulate vehicle movement in real-time has recently seen substantial advancement in the development of intelligent transportation systems and urban traffic management. It has been a challenging problem due to three aspects: 1) The diversity of traffic patterns due to heterogeneous layouts of urban intersections; 2) The nature of complex spatiotemporal correlations; 3) The requirement of dynamically adjusting the parameters of traffic models in a real-time system. To cater to these challenges, this paper proposes an online traffic simulation framework called automated urban traffic operation simulation via meta-learning (AUTOSIM). In particular, simulation models with various intersection layouts are automatically generated using an open-source simulation tool based on static traffic geometry attributes. Through a meta-learning technique, AUTOSIM enables an automated learning process for dynamic model settings of traffic scenarios featured with different spatiotemporal correlations. Besides, AUTOSIM is capable of adapting traffic model parameters according to dynamic traffic information in real-time by using a meta-learner. Through computational experiments, we demonstrate the effectiveness of the meta-learning-based framework that is capable of providing reliable supports to real-time traffic simulation and dynamic traffic operations.
Geometry Flow-Based Deep Riemannian Metric Learning
Yangyang Li, Chaoqun Fei, Chuanqing Wang, Hongming Shan, Ruqian Lu
2023, 10(9): 1882-1892. doi: 10.1109/JAS.2023.123399
Abstract(301) HTML (70) PDF(108)
Deep metric learning (DML) has achieved great results on visual understanding tasks by seamlessly integrating conventional metric learning with deep neural networks. Existing deep metric learning methods focus on designing pair-based distance loss to decrease intra-class distance while increasing inter-class distance. However, these methods fail to preserve the geometric structure of data in the embedding space, which leads to the spatial structure shift across mini-batches and may slow down the convergence of embedding learning. To alleviate these issues, by assuming that the input data is embedded in a lower-dimensional sub-manifold, we propose a novel deep Riemannian metric learning (DRML) framework that exploits the non-Euclidean geometric structural information. Considering that the curvature information of data measures how much the Riemannian (non-Euclidean) metric deviates from the Euclidean metric, we leverage geometry flow, which is called a geometric evolution equation, to characterize the relation between the Riemannian metric and its curvature. Our DRML not only regularizes the local neighborhoods connection of the embeddings at the hidden layer but also adapts the embeddings to preserve the geometric structure of the data. On several benchmark datasets, the proposed DRML outperforms all existing methods and these results demonstrate its effectiveness.
Underwater Cable Localization Method Based on Beetle Swarm Optimization Algorithm
Wenchao Huang, Zhijun Pan, Zhezhuang Xu
2023, 10(9): 1893-1895. doi: 10.1109/JAS.2022.106073
Abstract(338) HTML (32) PDF(163)
MPC-Based Change Management of Supply Chain Under Disruption Risks: The Case of Battery Industry
Yi Yang, Chen Peng
2023, 10(9): 1896-1898. doi: 10.1109/JAS.2023.123294
Abstract(204) HTML (43) PDF(96)
Deep Transfer Ensemble Learning-Based Diagnostic of Lithium-Ion Battery
Dongxu Ji, Zhongbao Wei, Chenyang Tian, Haoran Cai, Junhua Zhao
2023, 10(9): 1899-1901. doi: 10.1109/JAS.2022.106001
Abstract(333) HTML (43) PDF(133)
Competitive Meta-Learning
Boxi Weng, Jian Sun, Gao Huang, Fang Deng, Gang Wang, Jie Chen
2023, 10(9): 1902-1904. doi: 10.1109/JAS.2023.123354
Abstract(304) HTML (33) PDF(113)