Enhancing lane changing trajectory prediction on highways: A heuristic attention-based encoder-decoder model

Lane changing maneuvers could have a substantial influence on traffic flow characteristics as a result of their interfering effect on surrounding vehicles. The interference effect of lane changing is more pronounced when heavy vehicles change lanes compared to when passenger cars undertake the same maneuver. This is due to the physical effects that the heavy vehicles impose on surrounding traffic. This paper investigates and compares the traffic flow characteristics which influence the lane changing behavior of heavy vehicle and passenger car drivers on freeways under heavy traffic conditions. A trajectory data set comprising 28 heavy vehicle and 28 passenger car lane changing maneuvers is analyzed in this study. The results suggest a substantial difference exists between the traffic characteristics influencing the lane changing behavior of heavy vehicle and passenger car drivers. Heavy vehicles' speed changes little during a lane changing maneuver. Heavy vehicle drivers mainly move into the slower lanes to prevent obstructing the fast moving vehicles which approach from the rear. However, passenger car drivers increase their speed according to the speeds of the lead and lag vehicles in the target lane. They more commonly move into the faster lanes to gain speed advantages.

Lane change (LC) is a maneuver that allows drivers to enter into a lane that suits their requirements and comfort. The LC process requires the driver to assess its neighborhood traffic in its original and target lanes before undertaking the maneuver. Other vehicles in the neighborhood also need to adjust for safe lane change. The LC trajectory is determined by the accuracy of these subjective assessments as well as the state of traffic. An erroneous assessment by LC vehicle or neighboring vehicles or an incorrect maneuver can cause collision. The collision can be prevented if the LC trajectory can be predicted and the feasibility of LC can be communicated to different vehicles involved in this process. . In the present paper, neural network is used for long term forecast of the lane change trajectory and for short term near future positions of the LC vehicle. The neural network is trained using past LC trajectories of different vehicles. The trained network is then used for long and short term forecast of the vehicle’s positions during LC. Simulation results with actual filed data observed data indicates that neural network is able to learn LC maneuvers and is able to perform short term prediction with sufficient accuracy.

Along with the rapid development of autonomous driving technology, autonomous vehicles are showing a trend of practicality and popularity. Autonomous vehicles perceive environmental information through sensors to provide a basis for the decision making of vehicles. Based on this, this paper investigates the lane-changing decision-making behavior of autonomous vehicles. First, the similarity between autonomous vehicles and moving molecules is sought based on a system-similarity analysis. The microscopic lane-changing behavior of vehicles is analyzed by the molecular-dynamics theory. Based on the objective quantification of the lane-changing intention, the interaction potential is further introduced to establish the molecular-dynamics lane-changing model. Second, the relationship between the lane-changing initial time and lane-changing completed time, and the dynamic influencing factors of the lane changing, were systematically analyzed to explore the influence of the microscopic lane-changing behavior on the macroscopic traffic flow. Finally, the SL2015 lane-changing model was compared with the molecular-dynamics lane-changing model using the SUMO platform. SUMO is an open-source and multimodal traffic experimental platform that can realize and evaluate traffic research. The results show that the speed fluctuation of autonomous vehicles under the molecular-dynamics lane-changing model was reduced by 15.45%, and the number of passed vehicles was increased by 5.93%, on average, which means that it has better safety, stability, and efficiency. The molecular-dynamics lane-changing model of autonomous vehicles takes into account the dynamic factors in the traffic scene, and it reasonably shows the characteristics of the lane-changing behavior for autonomous vehicles.

Key feature selection and risk prediction for lane-changing behaviors based on vehicles’ trajectory data

<div>Single lane changing is one of the typical scenarios in vehicle driving. Planning an appropriate lane change trajectory is crucial in autonomous and semi-autonomous vehicle research. Existing polynomial trajectory planning mostly uses cubic or quintic polynomials, neglecting the lateral jerk constraints during lane changes. This study uses seventh-degree polynomials for lane change trajectory planning by considering the vehicle lateral jerk constraints. Simulation results show that the utilization of the seventh-degree method results in a 41% reduction in jerk compared to the fifth-degree polynomial. Furthermore, this study also proposes lane change trajectory schemes that can cater to different driving styles (e.g., safety, efficiency, comfort, and balanced performance). Depending on the driving style, the planned lane change trajectory ensures that the vehicle achieves optimal performance in one or more aspects during the lane change process. For example, with the trajectory that provides the best comprehensive performance under given constraints (initial speed of 20 m/s, lane width of 3.5 m, and a longitudinal distance of 50 m to the obstacle in front), the four-wheel steering model predictive control can effectively track the planned trajectory, with the maximum jerk value being 6.4 m/s<sup>3</sup> and the longitudinal speed after lane change being approximately 12.6 m/s. Although this study assumes specific longitudinal displacement before and after the lane change, the methodology is applicable to other scenarios. For example, it can determine the shortest longitudinal displacement and the optimal lane change trajectory given predefined vehicle speeds and maximum lateral acceleration conditions. The lane change trajectories developed in this study can be directly applied to the system design of autonomous vehicles.</div>

To investigate highway vehicle lane-changing behavior, we utilized the publicly available naturalistic driving dataset, HighD, to extract the movement data of vehicles involved in lane changes and their proximate counterparts. We employed univariate and multivariate Cox proportional hazards models alongside random survival forest models to analyze the influence of various factors on lane change duration, assess their statistical significance, and compare the performance of multiple random survival forest models. Our findings indicate that several variables significantly impact lane change duration, including the standard deviation of lane-changing vehicles, lane-changing vehicle speed, distance to the following vehicle in the target lane, lane-changing vehicle length, and distance to the following vehicle in the current lane. Notably, the standard deviation and vehicle length act as protective factors, with increases in these variables correlating with longer lane change durations. Conversely, higher lane-changing vehicle speeds and shorter distances to following vehicles in both the current and target lanes are associated with shorter lane change durations, indicating their role as risk factors. Feature variable selection did not substantially improve the training performance of the random survival forest model based on our findings. However, validation set evaluation showed that careful feature variable selection can enhance model accuracy, leading to improved AUC values. These insights lay the groundwork for advancing research in predicting lane-changing behaviors, understanding lane-changing intentions, and developing pre-emptive safety measures against hazardous lane changes.

Investigating model and impacts of lane-changing execution process based on CA model

With the development of vehicle-to-vehicle (V2V) communication, it is possible to share information among multiple vehicles. However, the existing research on automated lane changes concentrates only on the single-vehicle lane change with self-detective information. Cooperative lane changes are still a new area with more complicated scenarios and can improve safety and lane-change efficiency. Therefore, a multi-vehicle cooperative automated lane-change maneuver based on V2V communication for scenarios of eight vehicles on three lanes was proposed. In these scenarios, same-direction and intersectant-direction cooperative lane changes were defined. The vehicle that made the cooperative decision obtained the information of surrounding vehicles that were used to cooperatively plan the trajectories, which was called cooperative trajectory planning. The cooperative safety spacing model was proposed to guarantee and improve the safety of all vehicles, and it essentially developed constraints for the trajectory-planning task. Trajectory planning was treated as an optimization problem with the objective of maximizing safety, comfort, and lane-change efficiency under the constraints of vehicle dynamics and the aforementioned safety spacing model. Trajectory tracking based on a model predictive control method was designed to minimize tracking errors and control increments. Finally, to verify the validity of the proposed maneuver, an integrated simulation platform combining MATLAB/Simulink with CarSim was established. Moreover, a hardware-in-the-loop test bench was performed for further verification. The results indicated that the proposed multi-vehicle cooperative automated lane-change maneuver can achieve lane changes of multiple vehicles and increase lane-change efficiency while guaranteeing safety and comfort.

Fine-grained highway autonomous vehicle lane-changing trajectory prediction based on a heuristic attention-aided encoder-decoder model

In order to accurately predict the lane‐changing trajectory of the vehicle and improve the driving safety of the vehicle, a lane‐changing trajectory prediction model based on the combination of convolutional neural network (CNN) and bidirectional long short‐term memory (BiLSTM) neural network is proposed by comprehensively considering the historical driving behavior, the spatial characteristics of surrounding vehicles and the bidirectional time sequence information of the vehicle trajectory. Firstly, the vehicle trajectory data are filtered and smoothed, and it is divided into three categories: left lane change, right lane change, and straight driving, and a lane change trajectory sample set is constructed. Secondly, CNN‐BiLSTM model is constructed to identify the sample set of lane‐changing trajectory. Considering the interaction between vehicles in the driving process, the information of predicted vehicle, and surrounding vehicles is taken as the input of the model. The extracted feature vector is input to the BiLSTM layer for prediction after the CNN layer feature extraction, and the horizontal and vertical coordinates of the target vehicle at the next time are output. Thirdly, the trajectory data of the US‐101 dataset in NGSIM is selected to verify the performance of the CNN‐BiLSTM model, and at the same time, it is compared with models such as CNN‐LSTM, long short‐term memory (LSTM), BiLSTM, and CNN‐GRU‐Att. Finally, the verification result shows that the overall fitting degree of the vehicle lane change trajectory prediction of the proposed model reaches 99.50%, and the mean square error and mean absolute error are 0.0003076 and 0.01417, which are improved compared with other models. In the meanwhile, the research on multistep trajectory prediction in different prediction time domains is carried out. It was found that the longer the prediction time domain is, the lower the prediction performance of the model decreases, but the prediction accuracy still reached more than 96%, and it was able to accurately predict the lane change trajectory.

A major aspect in the development of advanced driving assistance systems (ADASs) is the research in developing human driving behavior prediction and recognition models. Recent contributions focus on developing these models for estimating different driving behaviors like lane or speed change. Thus, the models are incorporated into the ADAS to generate warnings and hints for safe maneuvers. Driving behavior recognition and prediction models are generally developed based on machine learning (ML) algorithms and are proven to generate accurate estimations. Previous review research contributions tend to focus on ML-based models for the prediction and recognition of speed change, trajectory change, and even driving styles. Due to high number of driving errors occurring during a lane change, a state-of-art review of different ML-based models for lane-changing behavior prediction and recognition is helpful to present a comparison between different models in terms of structure, influencing input variables, and performance. This enables the integration of the most efficient model for the development of ADASs to avoid accidents during a lane change. First, definitions and terms related to the model’s task and evaluation metrics used to evaluate the model’s performance are described to improve the readability. Then, the different input variables of the models affecting the lane-changing behaviors are presented. Next, a review of the models developed based on well-known approaches, such as artificial neural network (ANN), hidden Markov model (HMM), and support vector machine (SVM), using different input variables is given. Three lane-changing behaviors are focused on here: left/right lane change and lane keeping. The advantages and disadvantages of the different ML models with a comparison are summarized as well. Finally, the improvements required in the future are discussed.

More than 90% of traffic accidents are caused by driver behavior, with lane change behavior being a major contributor. Recently, driving assistance systems are being introduced on vehicles to reduce traffic accidents, and a reliable vehicle lane change collision detection system is a key component of these systems. Besides, the foundation of the vehicle lane change detection system is the effective vehicle lane change detection model. In this paper, based on the support vector machine, we propose a model for detecting driver lane change maneuvers and take into account the real-time vehicle dynamic features transmitted via Vehicle to X (V2X) Communication. The accuracy is ideal for lane keep and lane change situations, and it is also robust for zigzag driving situations, according to tests conducted using the NGSIM real traffic dataset. The detection accuracy for left and right lane change maneuvers is 97.5% and 99.09%, respectively, while the false alarm rate is 8.56%. Additionally, the average advance detection time is 1.7 s, which is suitable for actual driving application scenarios.

The expressway traffic flow characteristics of off-ramp have an important impact on traffic congestion. By analyzing the vehicle car-following and lane-changing behavior, the cellular automata model of off-ramp area was built. When analyzing lane-changing behavior, the behavior of vehicle changing lane from main lane to off-ramp buffer lane was considered, and the information interaction process between the vehicle ready to changing lane and the rear vehicle on the target lane was considered, the new lane-changing rules were built. In the new lane-changing rules, free lane-changing and compulsory lane-changing were included. By analyzing the simulation results, when traffic flow density is lower, the behavior of vehicles in the main lane changing lane into off-ramp buffer lane has less effect on traffic capacity of the road section, but when traffic flow density rising to a certain extent, the behavior of vehicles in the main lane changing lane into off-ramp buffer lane will bring an adverse effect on traffic capacity of the road section.

Lane-changing behavior is one of the most common driving behaviors while driving. Due to the complexity of its operation, vehicle collision accidents are prone to occur when changing lanes. Under the environment of vehicle networking, drivers can obtain more accurate traffic information in time, which can be of great help in terms of improving lane-changing safety. This paper analyzes the core factors that affect the safety of vehicles changing lanes, establishes the weight model of influencing factors of lane-changing behavior using the analytic hierarchy process (AHP), and obtains the calculation method of lane-changing behavior factors (LCBFs). Based on the fuzzy reasoning theory, the headway between the lane-changing vehicle and adjacent vehicles in the target lane was examined, and fuzzy logic lane-changing models were established for both situations (i.e., change to the left and change to the right lane). The fuzzy logic lane-changing models were tested via simulation experiments, and the test results showed that the models have a better warning effect on lane changing (LCBF = 1.5), with an accuracy of more than 90%. Thus, the established model in this paper can provide theoretical support for safety warnings when changing lanes and theoretical support for the sustainable development of transportation safety.

Nowadays, with more and more attention being paid to the characteristics and experience of drivers, a large number of driver classification algorithms have emerged. However, these methods basically cannot be adjusted independently to each driver. Therefore, this paper proposes a self-learning lane change motion planning system considering the driver’s personality. Firstly, the method of driver data acquisition and processing is determined to obtain and extract the lane change data. Then, the planning system built in this paper is explained from two aspects: lane change trigger and lane change trajectory. According to the artificial potential field theory, an obstacle driving risk field is established to evaluate the acceptance of environmental risks of different drivers, and to achieve personalized lane change triggers through online statistics. At the same time, the safety of lane change is ensured by establishing the safety distance model of the target lane. On the other hand, the driver characteristic coefficient Jc and the driver reaction and operation time td are introduced into the traditional Gaussian-distributed model to establish a personalized lane change trajectory planning model, in which the parameters are obtained from offline and online learning. Offline learning is based on DTW for trajectory matching, and uses AP clustering to obtain the generalized parameters; Online learning uses LSTM to achieve personalized updates. Finally, this paper selected 15 drivers for verification, and the results show that the motion planning system can well reproduce the lane change behavior of the driver.