Object detection in autonomous driving scenarios has become a popular task in recent years. Due to the high-speed movement of vehicles and the complex changes in the surrounding environment, objects of different scales need to be detected, which places high demands on the performance of the network model. Additionally, different driving devices have varying performance capabilities, and a lightweight model is needed to ensure the stable operation of devices with limited computing power. To address these challenges, we propose a lightweight network called BiGA-YOLO based on YOLOv5. We design the Ghost-Hardswish Conv module to simplify the convolution operations and incorporate spatial coordinate information into feature maps using Coordinate Attention. We also replace the PANet structure with the BiFPN structure to enhance the expression ability of features through different weights during the process of fusing multi-scale feature maps. Finally, we conducted extensive experiments on the KITTI dataset, and our BiGA-YOLO achieved a mAP@0.5 of 92.2% and a mAP@0.5:0.95 of 68.3%. Compared to the baseline model YOLOv5, our proposed model achieved improvements of 1.9% and 4.7% in mAP@0.5 and mAP@0.5:0.95, respectively, while reducing the model size by 15.7% and the computational cost by 16%. The detection speed was also increased by 6.3 FPS. Through analysis and discussion of the experimental results, we demonstrate that our proposed model is superior, achieving a balance between detection accuracy, model size, and detection speed.

To scientifically and effectively evaluate the service capacity of expressway service areas (ESAs) and improve the management level of ESAs, we propose a method for the recognition of vehicles entering ESAs (VeESAs) and estimation of vehicle dwell times using electronic toll collection (ETC) data. First, the ETC data and their advantages are described in detail, and then the cleaning rules are designed according to the characteristics of the ETC data. Second, we established feature engineering according to the characteristics of VeESA and proposed the XGBoost-based VeESA recognition (VR-XGBoost) model. Studied the driving rules in depth, we constructed a kinematics-based vehicle dwell time estimation (K-VDTE) model. The field validation in Part A/B of Yangli ESA using real ETC transaction data demonstrates that the effectiveness of our proposal outperforms the current state-of-the-art. Specifically, in Part A and Part B, the recognition accuracies of VR-XGBoost are 95.9% and 97.4%, respectively, the mean absolute errors (MAEs) of dwell time are 52 and 14 s, respectively, and the root mean square errors (RMSEs) are 69 and 22 s, respectively. In addition, the confidence level of controlling the MAE of dwell time within 2 min is more than 97%. This work can effectively recognize the VeESA and accurately estimate the dwell time, which can provide a reference idea and theoretical basis for the service capacity evaluation and layout optimization of the ESA.

Toll stations are bottlenecks in the traffic flow of expressways, and the evaluation of their capacity is essential for the operation of the expressway. Traditional capacity studies are mostly based on theoretical modelling of traffic engineering or simulation, with a focus on parameter tuning and idealized hypotheses, resulting in poor reliability. In view of the coexistence of electronic toll collection lanes and compound toll collection lanes at toll stations of expressways in China, the present study analyses the capacity of entrance and exit lanes of toll stations under mixed traffic conditions using a real toll data-driven approach. Firstly, the service time of a single vehicle during the saturation period was taken as the starting point for the capacity estimates. Secondly, the variation in service time for multiple categories of vehicles is modelled using lognormal distribution. Finally, the capacity of the two types of toll lanes at the designated toll station is determined. The important outcome of this study is the calculation of authentic capacity at the toll stations and the discussion of individual special toll lanes. Accordingly, it contributes to the development of appropriate policies to manage the operation of the toll plaza effectively.

Electronic toll collection (ETC) data mining has become one of the hotspots in the research of intelligent expressway extension applications. Ensuring the integrity of ETC data stands as a critical measure in upholding data quality. ETC data are typical structured data, and although deep learning holds great potential in the ETC data restoration field, its applications in structured data are still in the early stages. To address these issues, we propose an expressway ETC missing transaction data restoration model considering multi-attribute features (MAF). Initially, we employ an entity embedding neural network (EENN) to automatically learn the representation of categorical features in multi-dimensional space, subsequently obtaining embedding vectors from networks that have been adequately trained. Then, we use long short-term memory (LSTM) neural networks to extract the changing patterns of vehicle speeds across several continuous sections. Ultimately, we merge the processed features with other features as input, using a three-layer multilayer perceptron (MLP) to complete the ETC data restoration. To validate the effectiveness of the proposed method, we conducted extensive tests using real ETC datasets and compared it with methods commonly used for structured data restoration. The experimental results demonstrate that the proposed method significantly outperforms others in restoration accuracy on two different datasets. Specifically, our sample data size reached around 400,000 entries. Compared to the currently best method, our method improved the restoration accuracy by 19.06% on non-holiday ETC datasets. The MAE and RMSE values reached optimal levels of 12.394 and 23.815, respectively. The fitting degree of the model to the dataset also reached its peak ($R^2$ = 0.993). Meanwhile, the restoration stability of our method on holiday datasets increased by 5.82%. An ablation experiment showed that the EENN and LSTM modules contributed 7.60% and 9% to the restoration accuracy, as well as 4.68% and 7.29% to the restoration stability. This study indicates that the proposed method not only significantly improves the quality of ETC data but also meets the timeliness requirements of big data mining analysis.

In Electronic Toll Collection (ETC) systems, accurate gantry topology data are crucial for fair and efficient toll collection. Currently, inaccuracies in the topology data can cause tolls to be based on the shortest route rather than the actual distance travelled, contradicting the ETC system’s purpose. To address this, we adopt a novel Gradient Boosting Decision Tree (GBDT) algorithm, Light Gradient Boosting Machine (LightGBM), to dynamically update ETC gantry topology data on highways. We use ETC gantry and toll booth transaction data from a province in southeast China, where ETC usage is high at 72.8%. From this data, we generate a candidate topology set and extract five key characteristics. We then use Amap API and QGIS map analysis to annotate the candidate set, and, finally, apply LightGBM to train on these features, generating the dynamic topology. Our comparison of LightGBM with 14 other machine learning algorithms showed that LightGBM outperformed the others, achieving an impressive accuracy of 97.6%. This methodology can help transportation departments maintain accurate and up-to-date toll systems, reducing errors and improving efficiency.

Due to the challenge of limited line of sight in the perception system of intelligent driving vehicles (cameras, radar, body sensors, etc.), which can only perceive threats within a limited range, potential threats outside the line of sight cannot be fed back to the driver. Therefore, this article proposes a safety perception detection method for beyond the line of sight for intelligent driving. This method can improve driving safety, enabling drivers to perceive potential threats to vehicles in the rear areas beyond the line of sight earlier and make decisions in advance. Firstly, the electronic toll collection (ETC) transaction data are preprocessed to construct the vehicle trajectory speed dataset; then, wavelet transform (WT) is used to decompose and reconstruct the speed dataset, and lightweight gradient noosting machine learning (LightGBM) is adopted to train and learn the features of the vehicle section speed. On this basis, we also consider the features of vehicle type, traffic flow, and other characteristics, and construct a quantitative method to identify potential threat vehicles (PTVs) based on a fuzzy set to realize the dynamic safety assessment of vehicles, so as to effectively detect PTVs within the over-the-horizon range behind the driver. We simulated an expressway scenario using an ETC simulation platform to evaluate the detection of over-the-horizon PTVs. The simulation results indicate that the method can accurately detect PTVs of different types and under different road scenarios with an identification accuracy of 97.66%, which verifies the effectiveness of the method in this study. This result provides important theoretical and practical support for intelligent driving safety assistance in vehicle–road collaboration scenarios.