检索结果-内蒙古大学图书馆

A domain-adaptive method with cycle perceptual consistency adversarial networks for vehicle target detection in foggy weather

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2022年第7期16卷 971-981页

作者： Guo, Ying Liang, Rui-lin Cui, You-kai Zhao, Xiang-mo Meng, Qiang Changan Univ Joint Lab Internet Vehicles Minist Educ China Mobile Commun Corp Xian Peoples R China Zhejiang Inst Commun Co Ltd Inst Intelligent Transportat Hangzhou Peoples R China Shandong Prov Commun Planning & Design Inst Grp C Inst Intelligent Transportat Jinan Peoples R China

Foggy weather can cause such problems as blurred image information and the loss of image details, which may pose great challenges to road traffic target detection based on images and videos. In this study, we propose a domain-adaptive road vehicle target detection method to implement domain adaptation for the real foggy scene. We firstly constructed a highway vehicle detection dataset with foggy images (HVFD), which contains normal weather images and foggy images and provides a complete data support for vehicle detection based on computer vision. Secondly, by improving CycleGAN we designed an improved generative confrontation network (CPGAN), which realised the style transfer between foggy images and normal weather images. Finally, we formulated a YOLOv4 target detection framework according to the domain adaptation based on the pre-trained YOLOv4 fog vehicle detection model. The experimental results show that the method we put forward can effectively improve vehicle detection performance and reduce the work of manually labelling a large number of foggy image tags, which has a strong generalisation ability for computer vision-based applications in low-visibility weather.

关键词： foggy image tags vehicle detection performance road vehicles improved generative confrontation network deep learning (artificial intelligence) blurred image information YOLOv4 target detection framework foggy scene highway vehicle detection dataset domain-adaptive method road traffic object detection domain-adaptive road vehicle target detection method cycle perceptual consistency adversarial networks computer vision foggy images foggy weather feature extraction domain adaptation image details normal weather images road traffic target detection low-visibility weather traffic engineering computing pre-trained YOLOv4 fog vehicle detection model

来源：评论

学校读者我要写书评

暂无评论

Method for predicting the remaining mileage of electric vehicles based on dimension expansion and model fusion

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2022年第8期16卷 1074-1091页

作者： Tian, Sheng Li, Chengwei Lv, Qing Li, Jia South China Univ Technol Sch Civil Engn & Transportat Guangzhou Peoples R China

Accurately predicting the remaining mileage of electric vehicles (EVs) can effectively alleviate user's mileage anxiety and develop refinement of energy management strategy. However, traditional prediction methods not only consume time and resources, but also accumulate errors and lack interpretability. In this paper, we proposed a model based on dimension expansion and model fusion strategy, which uses the extreme gradient boosting (XGBoost) algorithm to directly predict the remaining mileage of EVs. After pre-processing the real running data of EVs, we constructed the field of remaining driving range and analyzed the relationship between features and remaining driving range, and then directly predicted the remaining driving mileage. Compared with other machine learning methods, XGBoost model has the highest accuracy. Then dimensional extended data set was obtained based on prior knowledge and symbol conversion, which improved the model performance. Finally, the model fusion strategy was adopted to further improve the generalization ability and stability of the model. The experimental results show that the Bootstrap aggregating (Bagging) fusion model has the highest predictive performance on the test set and outperformed other methods. The maximum RAE is not more than 3.5%, RMSE is less than 3km and MAE is about 2 km.

关键词： energy management RMSE electric vehicles RAE model fusion dimension expansion supervised learning data handling dimensional extended data set neural nets generalisation (artificial intelligence) machine learning predictive performance remaining driving range symbol conversion remaining driving mileage bagging user mileage anxiety bootstrap aggregating fusion model extreme gradient boosting prior knowledge remaining mileage prediction XGBoost algorithm generalization ability energy management systems traffic engineering computing mean square error methods

来源：评论

学校读者我要写书评

暂无评论

CAE-GAN: A hybrid model for vehicle trajectory prediction

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2022年第12期16卷 1682-1696页

作者： Chen, Long Zhou, Qiyang Cai, Yingfeng Wang, Hai Li, Yicheng Jiangsu Univ Automot Engn Res Inst Zhenjiang 212013 Jiangsu Peoples R China Jiangsu Univ Sch Automot & Traff Engn Zhenjiang Jiangsu Peoples R China

Trajectory prediction of surrounding vehicles is a crucial capability of intelligent driving vehicles. In a scene, a vehicle and its surrounding vehicles constitute an integral system, and the vehicle's future motion trajectory is affected by the actions of surrounding vehicles. The influencing mode and degree are hidden in the relevant historical information of the vehicle and its neighbour vehicle. The existing trajectory prediction methods either do not consider the confidence of the predicted trajectory, or the accuracy requirement is ignored when considering the confidence of the predicted trajectory. In order to address this problem, a mixed Conditional AutoEncoder Generative Adversarial Network (CAE-GAN) model based on the multi-loss function is proposed. The proposed model uses the encoder-decoder structure with a convolutional social pool to extract general features. The generative adversarial networks (GANs) are used to extract the confidence features of the generated trajectories, which enables the proposed model to generate trajectories that are close to the real trajectories. In addition, a classifier structure based on an LSTM network is added to output the probability that the predicted trajectory belongs to a particular lateral maneuver so that the generated trajectory lateral maneuver of the model is consistent with the real trajectory lateral maneuver. The proposed model is evaluated using the publicly available NGSIM US-101 and I-80 datasets, and results show that the accuracy of the proposed model is higher than that of the existing methods. The proposed model achieves an average accuracy improvement of 16.34% in comparison to the most advanced existing models.

关键词： convolutional social pool generative adversarial networks deep learning (artificial intelligence) trajectory prediction methods future motion trajectory I-80 dataset LSTM network road traffic intelligent driving vehicles NGSIM US-101 dataset multiloss function encoder-decoder structure recurrent neural nets neighbour vehicle generated trajectory lateral maneuver feature extraction vehicle trajectory prediction mixed conditional autoencoder generative adversarial network model CAE-GAN traffic engineering computing

来源：评论

学校读者我要写书评

暂无评论

A new theory of driver vision pressure energy field and its application in driver behaviour decision-making model

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2022年第1期16卷 1-12页

作者： Li, Yi Yu, Bo Chen, Yuren Hu, Zhihua Shanghai Maritime Univ Logist Res Ctr 1550 Haigang Ave Shanghai Peoples R China Tongji Univ Minist Educ Key Lab Rd & Traff Engn 4800 Caoan Hwy Shanghai 201804 Peoples R China

The interpretation of driver behaviour decisions is an essential part of driver behaviour research. Unlike previous studies that use driver's vision indicators or behaviour indicators as the basis for behaviour decision-making models, this paper proposes a new concept of vision pressure energy field to describe driver's vision perception. Driver behaviours are regarded as the results of the energy fluctuation of the "Potential Energy-Field Energy-Kinetic Energy" cycle. The energy field model and corresponding classification method of driving risk level are presented. The micro-effect and macro-effect of driver behaviour decisions are considered in the decision effect evaluation process. These models are integrated into an RNN (Recurrent Neural Network) framework. After the field test data training, the model results show that the decision-making framework with a hidden layer can successfully describe the car-following and lane-changing behaviours. The phenomenon of continuous behaviour change can be explained by the prediction result of decision effect level. The vision pressure energy field theory integrates the driver behaviour into the physical energy field theory. It presents a new way to interpret driver's vision perception results. The driver behaviour changes can also be successfully predicted through this theory.

关键词： physical energy field theory driver behaviour decision-making model energy fluctuation decision effect level driver behaviour decisions Potential Energy-Field Energy-Kinetic Energy cycle driver vision pressure energy field risk analysis recurrent neural nets energy field model field test data training decision effect evaluation process decision making continuous behaviour change vision pressure energy field theory behaviour indicators behavioural sciences computing behaviour decision-making models decision making framework traffic engineering computing

来源：评论

学校读者我要写书评

暂无评论

Signal timing parameters inference method at intersections using license plate recognition data

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2022年第8期16卷 1092-1107页

作者： Lin, Qinghai Chen, Jinyong Li, Guilong He, Zhaocheng SUN YAT SEN Univ Sch Intelligent Syst Engn Guangzhou 5106 Peoples R China Guangdong Prov Key Lab Intelligent Transportat Sy Guangzhou Peoples R China Guangdong Polytech Normal Univ Sch Automobile & Transportat Engn Guangzhou Peoples R China

Signal timing parameters are essential components in traffic signal control (TSC). It affects not only traffic management, but also traffic safety. However, due to the confidential issues of the traffic management department, or lack of data integration from different signal manufacturers, it is intractable to obtain the city-scale signal timing data. In the previous studies, some existing estimation methods focused on a single parameter and fixed-timing scheme. To tackle this issue, this study attempts to develop an integrated parameters inference method based on license plate recognition (LPR) data, considering phase weight, average phase duration information and the overall phases of the intersections. In particular, the proposed method includes phase sequence inference model, cycle length inference model and phase duration inference model. To testify the performance of the proposed method, a real-world LPR dataset from Guangzhou, China, is applied. Numerical results show that the proposed method performs more efficiently on parameter inferences than the state-of-the-art (SOTA) approach. For instance, in the given research time period, the mean absolute error (MAE) of each phase duration is 2 s averagely (6.32 s in the SOTA approach), and the mean relative error (MRE) of cycle length is 0.91% (11.67% in the SOTA approach).

关键词： traffic safety phase sequence inference model integrated parameters inference method Guangzhou China city-scale signal timing data traffic signal control inference mechanisms object recognition MAE data integration road safety road traffic control average phase duration information traffic engineering computing phase duration inference model mean absolute error LPR phase weight TSC confidential issues signal timing parameters inference method cycle length inference model mean relative error license plate recognition data MRE traffic management department

来源：评论

学校读者我要写书评

暂无评论

A sharing deep reinforcement learning method for efficient vehicle platooning control

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2022年第12期16卷 1697-1709页

作者： Lu, Sikai Cai, Yingfeng Chen, Long Wang, Hai Sun, Xiaoqiang Jia, Yunyi Jiangsu Univ Automot Engn Res Inst Zhenjiang 212013 Jiangsu Peoples R China Jiangsu Univ Sch Automot & Traff Engn Zhenjiang Jiangsu Peoples R China Clemson Univ Int Ctr Automot Reseatch CU ICAR Dept Automot Engn Greenville SC USA Zhenjiang City Jiangsu Univ Engn Technol Res Inst Zhenjiang Jiangsu Peoples R China

The combination of reinforcement learning and platooning control has been widely studied, which has been considered to be altruistic on the basis of safety. However, the platooning control method based on reinforcement learning is not mature. This paper proposes a platoon sharing deep deterministic policy gradient algorithm (PSDDPG) for a multi-vehicle network which overcomes the problem of low efficiency of continuous action space exploration. On the basis of the noise of deep deterministic policy gradient (DDPG) algorithm, additional platoon noise is added to enhance the diversity of training samples during exploration and avoid using a single-vehicle data, which could weaken model robustness. The method of using the time sequence information as an input and replay buffer backup is proposed to prevent the problems of insufficient exploration and low sampling efficiency from deteriorating the training effect. The robustness of the proposed algorithm is verified by tests using the Carla simulator. Also, the platoon merging, overtaking, cruising following and obstacle avoidance control of a vehicle platoon is realized under typical traffic flow conditions. The experimental results show that the PSDDPG algorithm can provide an efficient control strategy and thus has the potential to reduce energy consumption and improve road efficiency.

关键词： obstacle avoidance control mobile robots replay buffer backup low sampling efficiency road vehicles efficient vehicle platooning control energy consumption deep learning (artificial intelligence) road efficiency control engineering computing learning (artificial intelligence) vehicle platoon deep deterministic policy gradient algorithm efficient control strategy training samples single-vehicle data platooning control method platoon merging PSDDPG algorithm road traffic sharing deep reinforcement learning method model robustness time sequence information additional platoon noise reinforcement learning insufficient exploration collision avoidance gradient methods multivehicle network traffic engineering computing continuous action space exploration

来源：评论

学校读者我要写书评

暂无评论

The impact of ride-hailing services on the use of traditional taxis: Evidence from Chinese urban panel data

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2022年第11期16卷 1611-1622页

作者： Zhong, Jun Zhou, Huan Lin, Yan Ren, Fangxiao Hunan Univ Technol Sch Business Zhuzhou 412007 Peoples R China Dalian Maritime Univ Dept Management Sci & Engn Dalian Peoples R China

Currently, opinions on whether and how ride-hailing services have impacted the use of traditional taxis are divided, and no consensus has been reached. This paper used panel data from 44 cities in China from 2010 to 2016 to estimate the impact of ride-hailing services on the use of traditional taxis by applying the multiperiod difference-in-difference (multiperiod DID) method and conducting a series of robustness tests. The results show that ride-hailing services have a large, negative impact on the use of traditional taxis, and this negative impact is more severe in eastern cities. These results suggest that in China it is necessary for traditional taxi operators to make changes and innovations to become more sustainable in response to pressure from ride-hailing services. Additionally, it is suggested that city governments in China pay greater attention to the impact of emerging ride-hailing services on traditional taxi services, embrace emerging modes of mobility, and create appropriate policies to coordinate the development of both new and traditional services.

关键词： public transport ride-hailing services city governments multiperiod DID method taxi services Chinese urban panel data taxi operators multiperiod difference-in-difference method government policies sustainable development traffic engineering computing town and country planning

来源：评论

学校读者我要写书评

暂无评论

The influence of landscape intervention used as an alertness maintaining 'tool' on driving behaviour

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2022年第3期16卷 394-407页

作者： Yu-Le Zhang Shou-Lin Zhu Inner Mongolia Agr Univ Coll Energy & Transportat Engn Hohhot 010000 Peoples R China

Monotony, caused by deficiencies in the visual reference, can increase the risk of road accidents. This paper focuses on using landscape intervention as an alertness maintaining 'tool' to counter monotony. Therefore, a driving simulator experiment with four different scenes was designed. One is the control group with no landscape. The other three are landscapes with different spatial patterns. The spatial patterns were treatments combing the different graphic shapes and different enclosure degrees. Twenty-six subjects were required to drive in the centre of the right lane with a speed limit of 80 km/h as an alertness task. To assess the impact of different landscapes on driving behaviour, speed, steering wheel angle, and lane position were analysed. The results indicated a significant advantage of alertness maintaining of semi-open chevron landscape. Speeding and lane crossing has been significantly improved (p < 0.05). Also, the adequate length of the semi-open chevron landscape based on the alertness threshold of steering wheel angle SampEn is longer than the other spatial patterns. This work can help to develop more efficient monotony countermeasures with landscape intervention. It also provides valuable information for methods of adequate duration determination in alertness maintaining 'tool' based on driving behaviour.

关键词： alertness threshold road accidents steering systems road vehicles driver information systems lane alertness maintaining tool different landscapes different enclosure degrees road traffic semiopen chevron landscape different graphic shapes driving simulator experiment driving behaviour road safety different spatial patterns steering wheel angle SampEn wheels landscape intervention alertness task efficient monotony countermeasures traffic engineering computing

来源：评论

学校读者我要写书评

暂无评论

Research on fatigue detection based on visual features

引用

IET IMAGE PROCESSING 2022年第4期16卷 1044-1053页

作者： Zhao, Guangzhe He, Yanqing Yang, Hanting Tao, Yong Beijing Univ Civil Engn & Architecture Coll Elect & Informat Engn Beijing 102600 Peoples R China Coll Beihang Univ Beijing Peoples R China

The high incidence of traffic accidents brings immeasurable losses to life. In order to avoid such crises, researchers and automakers have used many methods to solve this problem. Among them, technology based on visual features is widely used in driver fatigue detection. As fatigue detection plays a vital role in the driving process, the high accuracy of fatigue monitoring is very important. This paper focuses on the method based on convolutional neural network to detect driver fatigue. First, in the face detection part, the Single-Shot Multi-Box Detector algorithm is used to improve the speed and accuracy of face detection to extract the eye and mouth regions;second, the VGG16 network is used to learn fatigue features, which is performed on the NTHU-Drowsy Driver Detection (NTHU-DDD) data set and the other two modified data sets Training test. The main result of this work is that the accuracy of fatigue monitoring is higher than other methods including the original method, with an accuracy rate of over 90%. And it has better generalization ability than the multi-physical feature fusion detection method. At the same time, we propose the fatigue detection method based on convolutional neural network to improve the advanced driver assistance system (ADAS) to make it more robust and reliable decision making.

关键词： VGG16 network convolutional neural network visual features fatigue detection method road safety advanced driver assistance system convolutional neural nets fatigue features road accidents face detection part fatigue monitoring learning (artificial intelligence) automakers modified data sets Training test feature extraction multiphysical feature fusion detection method decision making Single-Shot MultiBox Detector algorithm NTHU-Drowsy Driver Detection data object detection fatigue driver information systems traffic engineering computing driver fatigue detection

来源：评论

学校读者我要写书评

暂无评论

Prediction model for short-term traffic flow based on a K-means-gated recurrent unit combination

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2022年第5期16卷 675-690页

作者： Sun, Zhaoyun Hu, Yuanjiao Li, Wei Feng, Shaowei Pei, Lili Changan Univ Sch Informat Engn Xian 710064 Shaanxi Peoples R China

Short-term forecasting of traffic flow is an indispensable part of easing traffic pressure. Considering that different traffic flow patterns will affect the short-term traffic flow prediction results, a combined method based on the K-means clustering algorithm and gated recurrent unit (GRU) is proposed to build a short-term traffic flow prediction model to overcome the above problems. The K-means algorithm is used to cluster historical traffic flow data to establish different traffic flow pattern libraries. The K-nearest neighbour (KNN) classification algorithm is used to determine the historical traffic flow pattern most similar to the traffic flow change trend of the date to be predicted. All historical traffic flow data in this category is used training samples to make targeted predictions. The traffic flow data of performance measurement system (PeMS) in California, USA is used to verify the performance of the proposed model. Compared with the GRU network, stacked auto encoders (SAEs), random forest (RF), and support vector machine regression (SVR), the results show that the proposed combination model K-means-GRU considers the diversity of traffic flow patterns and improves the prediction accuracy, it can better solve the short-term traffic flow prediction problem.

关键词： pattern clustering short-term traffic flow prediction model intelligent transportation systems random forests short-term traffic flow prediction results different traffic flow patterns historical traffic flow pattern road traffic recurrent unit combination prediction accuracy traffic pressure recurrent neural nets targeted predictions different traffic flow pattern libraries combination model K-means-GRU traffic flow change trend regression analysis short-term traffic flow prediction problem pattern classification support vector machines cluster historical traffic flow data neighbour classification algorithm short-term forecasting traffic engineering computing

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：