Cognitive Competence Improvement for Autonomous Vehicles: A Lane Change Identification Model for Distant Preceding Vehicles

Wang, Chang; Sun, Qinyu; Li, Zhen; Zhang, Hongjia; Ruan, Kaili

doi:10.1109/access.2019.2924557

Cited by 18 publications

(14 citation statements)

References 65 publications

(63 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…If the distance from the ego vehicle to the lane is equal to 0, the vehicle has crossed the lane. Therefore, discerning the LC state after vehicles have crossed the lanes using the identification model is insignificant [59]. In summary, the time window should be less than the average time required for the vehicle to cross the lane.…”

Section: B Evaluation Of the Xgboost-based Lcd Modelmentioning

confidence: 99%

A Novel Lane-Changing Decision Model for Autonomous Vehicles Based on Deep Autoencoder Network and XGBoost

Han

2020

IEEE Access

View full text Add to dashboard Cite

Lane-changing (LC) is a critical task for autonomous driving, especially in complex dynamic environments. Numerous automatic LC algorithms have been proposed. This topic, however, has not been sufficiently addressed in existing on-road manoeuvre decision methods. Therefore, this paper presents a novel LC decision (LCD) model that gives autonomous vehicles the ability to make human-like decisions. This method combines a deep autoencoder (DAE) network with the XGBoost algorithm. First, a DAE is utilized to build a robust multivariate reconstruction model using time series data from multiple sensors; then, the reconstruction error of the DAE trained with normal data is analysed for LC identification (LCI) and training data extraction. Then, to address the multi-parametric and nonlinear problem of the autonomous LC decision-making process, an XGBoost algorithm with Bayesian parameter optimization is adopted. Meanwhile, to fully train our learning model with large-scale datasets, we proposed an online training strategy that updates the model parameters with data batches. The experimental results illustrate that the DAE-based LCI model is able to accurately identify the LC behaviour of vehicles. Furthermore, with the same input features, the proposed XGBoost-based LCD model achieves better performance than other popular approaches. Moreover, a simulation experiment is performed to verify the effectiveness of the decision model. INDEX TERMS Autonomous vehicle, lane-changing identification, lane-changing decision-making, deep autoencoder network, XGBoost.

show abstract

Section: B Evaluation Of the Xgboost-based Lcd Modelmentioning

confidence: 99%

A Novel Lane-Changing Decision Model for Autonomous Vehicles Based on Deep Autoencoder Network and XGBoost

Han

2020

IEEE Access

View full text Add to dashboard Cite

show abstract

“…At the same time, the target lists recognized by LiDAR data and images are matched to maximize the detection speed and achieve an average detection accuracy of 99.16% for pedestrian detection. Reference [114] used a stereo camera and LiDAR to detect the lane change behavior of the front vehicle. They used the neural network model based on particle swarm optimization to classify the distance, radial speed and horizontal speed of the vehicle to recognize the lane change behavior, and the final comprehensive recognition rate reached more than 88%.…”

Section: Fusion Strategy Based On Target Attributesmentioning

confidence: 99%

Multi-Sensor Fusion in Automated Driving: A Survey

2020

View full text Add to dashboard Cite

With the significant development of practicability in deep learning and the ultra-highspeed information transmission rate of 5G communication technology will overcome the barrier of data transmission on the Internet of Vehicles, automated driving is becoming a pivotal technology affecting the future industry. Sensors are the key to the perception of the outside world in the automated driving system and whose cooperation performance directly determines the safety of automated driving vehicles. In this survey, we mainly discuss the different strategies of multi-sensor fusion in automated driving in recent years. The performance of conventional sensors and the necessity of multi-sensor fusion are analyzed, including radar, LiDAR, camera, ultrasonic, GPS, IMU, and V2X. According to the differences in the latest studies, we divide the fusion strategies into four categories and point out some shortcomings. Sensor fusion is mainly applied for multi-target tracking and environment reconstruction. We discuss the method of establishing a motion model and data association in multi-target tracking. At the end of the paper, we analyzed the deficiencies in the current studies and put forward some suggestions for further improvement in the future. Through this investigation, we hope to analyze the current situation of multi-sensor fusion in the automated driving process and provide more efficient and reliable fusion strategies.

show abstract

“…But, under special circumstances (such as obstacles ahead, the target lane can only be reached after changing lanes, and a more suitable lane needs to be selected), the lane change behavior will be performed. In mixed traffic, autonomous vehicles will affect speed and will also effectively reduce the occurrence of lanechanging behavior [1][2][3][4][5]. It is important to understand the impact of autonomous vehicles on the speed and lanechanging behavior in a 4-lane closed road in mixed traffic.…”

Section: Introductionmentioning

confidence: 99%

Influence of Lane-Changing Behavior on Traffic Flow Velocity in Mixed Traffic Environment

Xie

Ren

Zheng

2022

Journal of Advanced Transportation

View full text Add to dashboard Cite

In mixed traffic with autonomous vehicles, the relationship between speed and lane-changing behavior is an important basis for mixed traffic control. In this study, we use empirical, simulation, and data-driven methods to study the relationship between speed and lane change rates in mixed traffic under different autonomous vehicle penetration rates. We use the empirical data to establish the corresponding road simulation models. Based on the simulation model, the traffic flow simulation experiments under the conditions of 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, and 90% penetration rate of autonomous vehicles were carried out. The analysis of the simulation results found that: (1) the penetration of autonomous vehicles into the road has a positive impact on the lanes far away from the entrance and exit, while the impact on the lanes closer to the entrance and exit is not obvious. (2) Lane-changing behavior has effectively decreased with the penetration of autonomous vehicles, but it is not obvious when the penetration rate exceeds 10%, and there is no significant drop in the lane connecting the entrance and exit. (3) There is a linear relationship between speed and lane-changing rate. Under different penetration rates, the data-driven analysis is used to perform multiple linear regressions, and the regression formula fits are all above 0.7. Based on the above findings, the linear formula of the fitting is proposed, and the value interval of the parameters in different states is given as well. Due to the small changes in the parameter values under different permeability conditions, the model has a certain degree of stability. The speed-lane change rate model proposed in this study can better describe the relationship between the speed of the ring-shaped urban expressway and the lane-changing behavior in the mixed traffic environment with the larger traffic flow.

show abstract

Cognitive Competence Improvement for Autonomous Vehicles: A Lane Change Identification Model for Distant Preceding Vehicles

Cited by 18 publications

References 65 publications

A Novel Lane-Changing Decision Model for Autonomous Vehicles Based on Deep Autoencoder Network and XGBoost

A Novel Lane-Changing Decision Model for Autonomous Vehicles Based on Deep Autoencoder Network and XGBoost

Multi-Sensor Fusion in Automated Driving: A Survey

Influence of Lane-Changing Behavior on Traffic Flow Velocity in Mixed Traffic Environment

Contact Info

Product

Resources

About