Loading...

Table of Content

    28 October 2021, Volume 26 Issue 5 Previous Issue    Next Issue

    Editorial
    Intelligent Connected Vehicle
    For Selected: Toggle Thumbnails
    Editorial
    Intelligent Connected Vehicle as the New Carrier Towards the Era of Connected World
    ZHUANG Hanyang (庄瀚洋), QIAN Yeqiang (钱烨强), YANG Ming(杨 明)
    2021, 26 (5):  559-560. 
    Abstract ( 362 )   PDF (83KB) ( 177 )  
    Human beings have been kept pursuing of higher efficiency and better safety to move people and things around since thousands of years ago. In modern soci-ety, vehicles are therefore invented and utilized to boost the speed and enhance the safety. In recent years, rapid development of information technology has brought hu-man into a new era of connected world. Internet and smartphones have made it extremely easy to get ac-cess to anyone from anywhere any time. In this back-ground, intelligent connected vehicles (ICVs) have been proposed and investigated. In the similar manner as the smartphones, ICVs are expected to be the next gener-ation carrier for people to get connected to the world. ICVs are equipped with novel sensors, controllers, and actuators to understand the environment, make decisions, and take actions, respectively. The word “intelligent” indicates that the vehicle should be able to handle unexpected events on the road. The word “connected” means that the information of each vehicle should be shared and considered globally. Full auton-omy and full connection are the ultimate goals of ICV industry. Unfortunately, we are still far away from this goal; therefore, continuous efforts shall be made to step further to this destination. As the ICV consists of multiple subsystems and is across different disciplines, the overall improvement re-quires the innovation in each aspect. Under this cir-cumstance, the Special Issue on Intelligent Connected Vehicle at Journal of Shanghai Jiao Tong University (Science) has been organized to broaden the perspec-tive, promote the interdisciplinary collaboration, and report the state-of-the-art works.
    Related Articles | Metrics
    Intelligent Connected Vehicle
    Camera-Radar Fusion Sensing System Based on Multi-Layer Perceptron
    YAO Tong (姚 彤), WANG Chunxiang(王春香), QIAN Yeqiang(钱烨强)
    2021, 26 (5):  561-568.  doi: 10.1007/s12204-021-2345-x
    Abstract ( 343 )   PDF (1189KB) ( 214 )  
    Environmental perception is a key technology for autonomous driving. Owing to the limitations of a single sensor, multiple sensors are often used in practical applications. However, multi-sensor fusion faces some problems, such as the choice of sensors and fusion methods. To solve these issues, we proposed a machine learning-based fusion sensing system that uses a camera and radar, and that can be used in intelligent vehicles. First, the object detection algorithm is used to detect the image obtained by the camera; in sequence, the radar data is preprocessed, coordinate transformation is performed, and a multi-layer perceptron model for correlating the camera detection results with the radar data is proposed. The proposed fusion sensing system was verified by comparative experiments in a real-world environment. The experimental results show that the system can effectively integrate camera and radar data results, and obtain accurate and comprehensive object information in front of intelligent vehicles.
    References | Related Articles | Metrics
    Wavelet Transform-Based High-Definition Map Construction From a Panoramic Camera
    ZHUANG Hanyang (庄瀚洋), ZHOU Xuejun (周学军), WANG Chunxiang (王春香), QIAN Yuhan (钱宇晗)
    2021, 26 (5):  569-576.  doi: 10.1007/s12204-021-2346-9
    Abstract ( 260 )   PDF (1242KB) ( 137 )  
    High-definition (HD) maps are key components that provide rich topologic and semantic information for decision-making in vehicle autonomous driving systems. A complete ground orthophoto is usually used as the base image to construct the HD map. The ground orthophoto is obtained through inverse perspective transformation and image mosaicing. During the image mosaicing, multiple consecutive orthophotos are stitched together using pose information and image registration. In this study, wavelet transform is introduced to the image mosaicing process to alleviate the information loss caused by image overlapping. In the orthophoto wavelet transform, high-frequency and low-frequency components are fused using different strategies to form a complete base image with clearer local details. Experimental results show that the accuracy of the orthophotos generated using this method is improved.
    References | Related Articles | Metrics
    Lightweight Method for Vehicle Re-identification Using Reranking Algorithm Based on Topology Information of Surveillance Network
    ZOU Yue (邹 悦), LI Lin (李 霖), YANG Xubo (杨旭波)
    2021, 26 (5):  577-586.  doi: 10.1007/s12204-021-2347-8
    Abstract ( 271 )   PDF (1353KB) ( 228 )  
    As an emerging visual task, vehicle re-identification refers to the identification of the same vehicle across multiple cameras. Herein, we propose a novel vehicle re-identification method that uses an improved ResNet-50 architecture and utilizes the topology information of a surveillance network to rerank the final results. In the training stage, we apply several data augmentation approaches to expand our training data and increase their diversity in a cost-effective manner. We reform the original RestNet-50 architecture by adding non-local blocks to implement the attention mechanism and replacing part of the batch normalization operations with instance batch normalization. After obtaining preliminary results from the proposed model, we use the reranking algorithm, whose core function is to improve the similarity scores of all images on the most likely path that the vehicle tends to appear to optimize the final results. Compared with most existing state-of-the-art methods, our method is lighter, requires less data annotation, and offers competitive performance.
    References | Related Articles | Metrics
    Intelligent Analysis of Abnormal Vehicle Behavior Based on a Digital Twin
    LI Lin (李 霖), HU Zeyu(胡泽宇), YANG Xubo (杨旭波)
    2021, 26 (5):  587-597.  doi: 10.1007/s12204-021-2348-7
    Abstract ( 289 )   PDF (2627KB) ( 88 )  
    Analyzing a vehicle’s abnormal behavior in surveillance videos is a challenging field, mainly due to the wide variety of anomaly cases and the complexity of surveillance videos. In this study, a novel intelligent vehicle behavior analysis framework based on a digital twin is proposed. First, detecting vehicles based on deep learning is implemented, and Kalman filtering and feature matching are used to track vehicles. Subsequently, the tracked vehicle is mapped to a digital-twin virtual scene developed in the Unity game engine, and each vehicle’s behavior is tested according to the customized detection conditions set up in the scene. The stored behavior data can be used to reconstruct the scene again in Unity for a secondary analysis. The experimental results using real videos from traffic cameras illustrate that the detection rate of the proposed framework is close to that of the state-of-the-art abnormal event detection systems. In addition, the implementation and analysis process show the usability, generalization, and effectiveness of the proposed framework.
    References | Related Articles | Metrics
    Efficient Online Vehicle Tracking for Real–Virtual Mapping Systems
    CHEN Jiacheng (陈佳诚), LI Lin(李 霖), YANG Xubo (杨旭波)
    2021, 26 (5):  598-606.  doi: 10.1007/s12204-021-2349-6
    Abstract ( 143 )   PDF (558KB) ( 53 )  
    Multi-object tracking is a vital problem as many applications require better tracking approaches. Although learning-based detectors are becoming extremely powerful, there are few tracking methods designed to work with them in real time. We explored an effcient fiexible online vehicle tracking-by-detection framework suitable for real–virtual mapping systems, which combines a non-recursive temporal window search with delayed output and produces stable trajectories despite noisy detection responses. Its computation speed meets the real-time requirements, whereas its performance is comparable with that of state-of-the-art online trackers on the DETRAC dataset. The trajectories from our approach also contain the target class and color information important for virtual vehicle motion reconstruction.
    References | Related Articles | Metrics
    Multi-Object Tracking Strategy of Autonomous Vehicle Using Modified Unscented Kalman Filter and Reference Point Switching
    WANG Muyuan∗ (王木塬), WU Xiaodong (吴晓东)
    2021, 26 (5):  607-614.  doi: 10.1007/s12204-021-2350-0
    Abstract ( 336 )   PDF (1070KB) ( 151 )  
    In this study, a multi-object tracking (MOT) scheme based on a light detection and ranging sensor was proposed to overcome imprecise velocity observations in object occlusion scenarios. By applying real-time velocity estimation, a modified unscented Kalman filter (UKF) was proposed for the state estimation of a target object. The proposed method can reduce the calculation cost by obviating unscented transformations. Additionally, combined with the advantages of a two-reference-point selection scheme based on a center point and a corner point, a reference point switching approach was introduced to improve tracking accuracy and consistency. The state estimation capability of the proposed UKF was verified by comparing it with the standard UKF in single-target tracking simulations. Moreover, the performance of the proposed MOT system was evaluated using real traffic datasets.
    References | Related Articles | Metrics
    Intelligent-Assist Algorithm for Remote Shared-Control Driving Based on Game Theory
    QIAO Bangjun∗ (乔邦峻), LI Huanghe (李黄河), WU Xiaodong (吴晓东)
    2021, 26 (5):  615-625.  doi: 10.1007/s12204-021-2351-z
    Abstract ( 172 )   PDF (759KB) ( 356 )  
    Contemporary autonomous-driving technology relies on good environmental-perception systems and high-precision maps. For unknown environments or scenarios where perception fails, a human-in-the-loop remote-driving system can effectively complement common solutions, although safety remains an issue for its application. A haptic shared-control algorithm based on non-cooperative game theory is presented in this paper. The algorithm generates collision-free reference paths with model predictive control and predicts the driver’s path using a two-point preview model. Man-machine torque interaction is modeled as a Nash game, and the assist system’s degree of intervention is regulated in real time, according to assessments of collision risk and the driver’s concentration. Simulations of several representative scenarios demonstrate how the proposed method improves driving safety, while respecting driver decisions.
    References | Related Articles | Metrics
    Stochastic Model Predictive Control Approach to Autonomous Vehicle Lane Keeping
    ZHANG Chenzhi (张晨之), ZHUANG Cheng (庄 诚), ZHENG Xueke (郑学科), CAI Runze (蔡润泽), LI Mian (李 冕)
    2021, 26 (5):  626-633.  doi: 10.1007/s12204-021-2352-y
    Abstract ( 299 )   PDF (702KB) ( 114 )  
    In real-world scenarios, the uncertainty of measurements cannot be handled effciently by traditional model predictive control (MPC). A stochastic MPC (SMPC) method for handling the uncertainty of states in autonomous driving lane-keeping scenarios is presented in this paper. A probabilistic system is constructed by considering the variance of states. The probabilistic problem is then transformed into a solvable deterministic optimization problem in two steps. First, the cost function is separated into mean and variance components. The mean component is calculated online, whereas the variance component can be calculated offline. Second, Cantelli’s inequality is adopted for the deterministic reformulation of constraints. Consequently, the original probabilistic problem is transformed into a quadratic programming problem. To validate the feasibility and effectiveness of the proposed control method, we compared the SMPC controller with a traditional MPC controller in a lane-keeping scenario. The results demonstrate that the SMPC controller is more effective overall and produces smaller steady-state distance errors.
    References | Related Articles | Metrics
    Cooperative Adaptive Cruise Control Using Delay-Based Spacing Policy: A Robust Adaptive Non-Singular Terminal Sliding Mode Approach
    WANG Weiyang (王维旸), CUI Ke (崔 科), GU Lizhong(顾立忠), LU¨ Xinjun (吕新军)
    2021, 26 (5):  634-646.  doi: 10.1007/s12204-021-2353-x
    Abstract ( 280 )   PDF (1152KB) ( 71 )  
    This study proposes two speed controllers based on a robust adaptive non-singular terminal sliding mode control approach for the cooperative adaptive cruise control problem in a connected and automated vehicular platoon. The delay-based spacing policy is adopted to guarantee that all vehicles in the platoon track the same target velocity profile at the same position while maintaining a predefined time gap. Factors such as nonlinear vehicle longitudinal dynamics, engine dynamics with time delay, undulating road profiles, parameter uncertainties, and external disturbances are considered in the system modeling and controller design. Different control objectives are assigned to the leading and following vehicles. Then, controllers consisting of a sliding mode controller with parameter adaptive laws based on the ego vehicle’s state deviation and linear coupled state errors, and a Smith predictor for time delay compensation are designed. Both inner stability and strong string stability are guaranteed in the case of nonlinear sliding manifolds. Finally, the effectiveness of the proposed controllers and the benefits of 44.73% shorter stabilization time, 11.20% less speed overshoot, and virtually zero steady-state inner vehicle distance deviation are illustrated in a simulation study of a seven-vehicle platoon cooperative adaptive cruise control and comparison experiments with a coupled sliding mode control approach.
    References | Related Articles | Metrics
    Parameter Identification of Magic Formula Tire Model Based on Fibonacci Tree Optimization Algorithm
    FENG Shilin (冯世林), ZHAO Youqun (赵又群), DENG Huifan (邓汇凡), WANG Qiuwei(王秋伟), CHEN Tingting (陈婷婷)
    2021, 26 (5):  647-657.  doi: 10.1007/s12204-021-2354-9
    Abstract ( 377 )   PDF (1130KB) ( 109 )  
    The magic formula (MF) tire model is a semi-empirical tire model that can precisely simulate tire behavior. The heuristic optimization algorithm is typically used for parameter identification of the MF tire model. To avoid the defect of the traditional heuristic optimization algorithm that can easily fall into the local optimum, a parameter identification method based on the Fibonacci tree optimization (FTO) algorithm is proposed, which is used to identify the parameters of the MF tire model. The proposed method establishes the basic structure of the Fibonacci tree alternately through global and local searches and completes optimization accordingly. The global search rule in the original FTO was modified to improve its efficiency. The results of independent repeated experiments on two typical multimodal function optimizations and the parameter identification results showed that FTO was not sensitive to the initial values. In addition, it had a better global optimization performance than genetic algorithm (GA) and particle swarm optimization (PSO). The root mean square error values optimized with FTO were 5.09%, 10.22%, and 3.98% less than the GA, and 6.04%, 4.47%, and 16.42% less than the PSO in pure lateral and longitudinal forces, and pure aligning torque parameter identi?cation. The parameter identification method based on FTO was found to be effective.
    References | Related Articles | Metrics
    Developing High-Precision Maps for Automated Driving in China: Legal Obstacles and the Way to Overcome Them
    ZHANG Taolue∗ (张韬略), TU Huizhao (涂辉招), QIU Wei (邱 炜)
    2021, 26 (5):  658-669.  doi: 10.1007/s12204-021-2355-8
    Abstract ( 261 )   PDF (285KB) ( 157 )  
    A high-precision map (HPM) is the key infrastructure to realizing the function of automated driving (AD) and ensuring its safety. However, the current laws and regulations on HPMs in China can lead to serious legal compliance problems. Thus, proper measures should be taken to remove these barriers. Starting with a complete view of the current legal obstacles to HPMs in China, this study first explains why these legal obstacles exist and the types of legal interests they are trying to protect. It then analyzes whether new technology could be used as an alternative to resolve these concerns. Factors such as national security, AD industry needs, and personal data protection, as well as the ?exibility of applying technology, are discussed and analyzed hierarchically for this purpose. This study proposes that China should adhere to national security and AD industry development, pass new technical regulations that redefine the scope of national security regarding geographic information in the field of HPMs, and establish a national platform under the guidance and monitoring of the government to integrate scattered resources and promote the development of HPMs via crowdsourcing. Regarding the legal obstacles with higher technical plasticity, priority should be given to technical solutions such as “available but invisible” technology. Compared with the previous research, this study reveals the current legal barriers in China that have different levels of relevance to national security and different technical plasticity. It also proposes original measures to remove them, such as coordinating national security with the development of the AD industry, reshaping the boundary of national security and industrial interests, and giving priority to technical solutions for legal barriers that have strong technical plasticity.
    References | Related Articles | Metrics
    Service Caching and Task Offloading for Mobile Edge Computing-Enabled Intelligent Connected Vehicles
    HUANG Mengting (黄梦婷), YI Yuhan (易雨菡), ZHANG Guanglin∗ (张光林)
    2021, 26 (5):  670-679.  doi: 10.1007/s12204-021-2356-7
    Abstract ( 203 )   PDF (635KB) ( 61 )  
    The development of intelligent connected vehicles (ICVs) has tremendously inspired the emergence of a new computing paradigm called mobile edge computing (MEC), which meets the demands of delay-sensitive on-vehicle applications. Most existing studies focusing on the issue of task offloading in ICVs assume that the MEC server can directly complete computation tasks without considering the necessity of service caching. However, this is unrealistic in practice because a large number of tasks require the use of corresponding third-party libraries and databases, that is, service caching. Therefore, we investigate the delay optimization in an MEC-enabled ICVs system with multiple mobile vehicles, resource-limited base stations (BSs), and one cloud server. We aim to determine the optimal service caching and task offloading decisions to minimize the overall system delay using mixed-integer nonlinear programming. To address this problem, we ?rst convert it into a quadratically constrained quadratic program and then propose an effcient semide?nite relaxation-based joint service caching and task offloading (JSCTO) algorithm to obtain the service caching and task o?oading decisions. In the simulations, we validate the e?ciency of our proposed method by setting different numbers of vehicles and the storage capacity of BSs. The results show that our proposed JSCTO algorithm can significantly decrease the total delay of all o?oaded tasks compared with the cloud processing only scheme.
    References | Related Articles | Metrics
    Obstacle Avoidance in Multi-Agent Formation Process Based on Deep Reinforcement Learning
    JI Xiukun (冀秀坤), HAI Jintao (海金涛), LUO Wenguang (罗文广), LIN Cuixia (林翠霞), XIONG Yu(熊 禹), OU Zengkai (殴增开), WEN Jiayan(文家燕)
    2021, 26 (5):  680-685.  doi: 10.1007/s12204-021-2357-6
    Abstract ( 308 )   PDF (675KB) ( 73 )  
    To solve the problems of di?cult control law design, poor portability, and poor stability of traditional multi-agent formation obstacle avoidance algorithms, a multi-agent formation obstacle avoidance method based on deep reinforcement learning (DRL) is proposed. This method combines the perception ability of convolutional neural networks (CNNs) with the decision-making ability of reinforcement learning in a general form and realizes direct output control from the visual perception input of the environment to the action through an end-to-end learning method. The multi-agent system (MAS) model of the follow-leader formation method was designed with the wheelbarrow as the control object. An improved deep Q netwrok (DQN) algorithm (we improved its discount factor and learning e?ciency and designed a reward value function that considers the distance relationship between the agent and the obstacle and the coordination factor between the multi-agents) was designed to achieve obstacle avoidance and collision avoidance in the process of multi-agent formation into the desired formation. The simulation results show that the proposed method achieves the expected goal of multi-agent formation obstacle avoidance and has stronger portability compared with the traditional algorithm.
    References | Related Articles | Metrics
    IoT System for Intelligent Firefighting in the Electric Power Industry
    HE Wei (何 伟)
    2021, 26 (5):  686-689.  doi: 10.1007/s12204-021-2358-5
    Abstract ( 323 )   PDF (96KB) ( 96 )  
    Traditional ?re safety management in the electric power industry has signi?cant drawbacks, including a lack of data, di?culty of maintenance, lack of supervision, and lack of interaction. This type of management lags behind current advanced safety management concepts such as “gate advancement” and “full process man-agement”, and it fails to meet the needs of future energy internet construction and development. In response to these problems, an internet of things system for smart ?re?ghting in the electric power industry was constructed in this study. This system de?nes a centralized information window, trains a power intelligent ?re?ghting brain, establishes a ?re?ghting cloud management and control system, constructs a power ?re?ghting interaction mech-anism, and performs multi-party coordination of ?re?ghting mechanisms to realize concept of “a whole network on one screen and everything in one network” for managing ?res.
    References | Related Articles | Metrics
    Curvature Adaptive Control Based Path Following for Automatic Driving Vehicles in Private Area
    SHI Qiang (师 强), ZHANG Jianlin (张建林), YANG Ming∗ (杨 明)
    2021, 26 (5):  690-698.  doi: 10.1007/s12204-021-2359-4
    Abstract ( 458 )   PDF (1349KB) ( 116 )  
    Path following refers to traveling along the desired path with automatic steering control, which is a crucial technology for automatic driving vehicles. Roads in private areas are highly irregular, resulting in a large curvature variation, which reduces the control accuracy of the path following. A curvature adaptive control (CAC) based path-following method was proposed to solve the problem mentioned above. Speci?cally, CAC takes advantage of the complementary characteristics in response to the path curvature ?uctuation of pure pursuit and front-wheel feedback and by combining the two methods further enhances the immunity of the control accuracy in response to a curvature ?uctuation. With CAC, the quantitative indices of the path curvature ?uctuation and control accuracy were constructed. The model between the path curvature ?uctuation and a dynamic parameter was identi?ed using the quantitative index of the control accuracy as the optimization target. The experimental results of a real vehicle indicate that the control accuracy of path following is further enhanced by its immunity in response to curvature ?uctuation improved by the CAC. In addition, CAC is easy to deploy and requires low demand for hardware resources.
    References | Related Articles | Metrics
    Integrated Framework for Test and Evaluation of Autonomous Vehicles
    SU Yimin (苏奕敏), WANG Lin∗ (王 琳)
    2021, 26 (5):  699-712.  doi: 10.1007/s12204-021-2360-y
    Abstract ( 183 )   PDF (795KB) ( 49 )  
    Autonomous vehicles must pass e?ective standard tests to verify their reliability and safety. Accord-ingly, it is very important to establish a complete scienti?c test and evaluation system for autonomous vehicles. A comprehensive framework incorporating the design of test scenarios, selection of evaluation indexes, and estab-lishment of an evaluation system is proposed in this paper. The aims of the system are to obtain an objective and quantitative score regarding the intelligence of autonomous vehicles, and to form an automated process in the future development. The proposed framework is built on a simulation platform to ensure the feasibility of the design and implementation of the test scenarios. The design principle for the test scenarios is also presented. To reduce subjective in?uences, the proposed framework selects objective indexes from four aspects: safety, comfort, driving performance, and standard regulations. The order relation analysis method is adopted to formulate the index weights, and fuzzy comprehensive evaluation is used to quantify the scores. Finally, a numerical example is provided to visually demonstrate the evaluation results for the autonomous vehicles scored by the proposed framework.
    References | Related Articles | Metrics
    Control System of Two-Wheel Self-Balancing Vehicle
    REN Haoa∗ (任 淏), ZHOU Congb (周 聪)
    2021, 26 (5):  713-721.  doi: 10.1007/s12204-021-2361-x
    Abstract ( 334 )   PDF (351KB) ( 98 )  
    This study mainly concerns a motion model and the main control algorithm of two-wheel self-balancing vehicle models. Details of the critical parameters fetching and output value of two-wheel self-balancing vehicle models are introduced, including those concerning balance control, speed control and direction control. An improved cascade coupling control scheme is proposed for two-wheel vehicles, based on a proportional-integral-derivative (PID) control algorithm. Moreover, a thorough comparison between a classic control system and the improved system is provided, and all aspects thereof are analyzed. It is determined that the control performance of the two-wheel self-balancing vehicle system based on the PID control algorithm is reliable, enabling the vehicle body to maintain balance while moving smoothly along a road at a fast average speed with better practical per-formance.
    References | Related Articles | Metrics
    Real-Time Trajectory Planning for On-road Autonomous Tractor-Trailer Vehicles
    SHEN Qiyue (沈琦越), WANG Bing (王 冰), WANG Chunxiang∗ (王春香)
    2021, 26 (5):  722-730.  doi: 10.1007/s12204-021-2362-9
    Abstract ( 380 )   PDF (1546KB) ( 150 )  
    Tractor-trailer vehicles, which are composed of a car-like tractor towing a passive trailer, have been widely deployed in the transportation industry, and trajectory planning is a critical step in enabling such a system to drive autonomously. Owing to the properties of being highly nonlinear and nonholonomic with complex dynamics, the tractor-trailer system poses great challenges to the development of motion-planning algorithms. In this study, an indirect trajectory planning framework for a tractor-trailer vehicle under on-road driving is presented to deal with the problem that the traditional planning framework cannot consider the feasibility and quality simultaneously in real-time trajectory generation of the tractor-trailer vehicle. The indirect planning framework can easily handle complicated tractor-trailer dynamics and generate high-quality, obstacle-free trajectory using quintic polynomial spline, speed pro?le optimization, forward simulation, and properly designed cost functions. Simulations under di?erent driving scenarios and trajectories with di?erent driving requirements are conducted to validate the performance of the proposed framework.
    References | Related Articles | Metrics
    Collision-Free Path Planning with Kinematic Constraints in Urban Scenarios
    WANG Liang (王 亮), WANG Bing (王 冰), WANG Chunxiang∗ (王春香)
    2021, 26 (5):  731-738.  doi: 10.1007/s12204-021-2363-8
    Abstract ( 307 )   PDF (2199KB) ( 117 )  
    In urban driving scenarios, owing to the presence of multiple static obstacles such as parked cars and roadblocks, planning a collision-free and smooth path remains a challenging problem. In addition, the path-planning problem is mostly non-convex, and contains multiple local minima. Therefore, a method for combining a sampling-based method and an optimization-based method is proposed in this paper to generate a collision-free path with kinematic constraints for urban scenarios. The sampling-based method constructs a search graph to search for a seeding path for exploring a safe driving corridor, and the optimization-based method constructs a quadratic programming problem considering the desired state constraints, continuity constraints, driving corridor constraints, and kinematic constraints to perform path optimization. The experimental results show that the proposed method is able to plan a collision-free and smooth path in real time when managing typical urban scenarios.
    References | Related Articles | Metrics
    Iterative-Reweighting-Based Robust Iterative-Closest-Point Method
    ZHANG Jianlin (张建林), ZHOU Xuejun (周学军), YANG Ming (杨 明)
    2021, 26 (5):  739-746.  doi: 10.1007/s12204-021-2364-7
    Abstract ( 322 )   PDF (1614KB) ( 58 )  
    In point cloud registration applications, noise and poor initial conditions lead to many false matches. False matches signi?cantly degrade registration accuracy and speed. A penalty function is adopted in many robust point-to-point registration methods to suppress the in?uence of false matches. However, after applying a penalty function, problems cannot be solved in their analytical forms based on the introduction of nonlinearity. Therefore, most existing methods adopt the descending method. In this paper, a novel iterative-reweighting-based method is proposed to overcome the limitations of existing methods. The proposed method iteratively solves the eigenvectors of a four-dimensional matrix, whereas the calculation of the descending method relies on solving an eight-dimensional matrix. Therefore, the proposed method can achieve increased computational e?ciency. The proposed method was validated on simulated noise corruption data, and the results reveal that it obtains higher e?ciency and precision than existing methods, particularly under very noisy conditions. Experimental results for the KITTI dataset demonstrate that the proposed method can be used in real-time localization processes with high accuracy and good e?ciency.
    References | Related Articles | Metrics