Vehicle Lane Change Research Articles

Dynamic Planning of Optimally Safe Lane-change Trajectory for Autonomous Driving on Multi-lane Highways Using a Fuzzy Logic–based Collision Estimator

The collision avoidance system in an autonomous vehicle, intended to address traffic safety issues, has a crucial function called collision estimation. It accomplishes this by identifying potential dangers and notifying the drivers in advance or by using autonomous control to navigate safely. In this work, a novel approach is proposed for generating and selecting a lane-change trajectory for the vehicle in a driving scenario where two vehicles are simultaneously executing lane-change processes on highways and approaching the same target lane. Moreover, a novel fuzzy logic estimator based on time-to-collision (TTC) and time-to-gap (TTG) is designed to estimate the collision risk. In the collision-avoidance process, the proposed estimator is utilized to determine the risk of a collision with polynomial function-based generation of possible lane-change trajectories. The safest lane-change trajectory is then provided to the motion controller so it can navigate the vehicle safely through such a challenging lane-change scenario. This work also investigates Stanley and Pure Pursuit controllers to follow the optimized trajectory. The simulation experiment results demonstrate that the proposed approach for dynamic trajectory generation during the lane-change process can successfully handle this type of challenging situation and prevent a potential collision. Experimental results also indicate that monitoring the movement of the nearby lane-changing vehicle is crucial for safe lane-change execution and that the proposed approach successfully handles the challenging situation, preventing potential collision.

Expressway lane change strategy of autonomous driving based on prior knowledge and data-driven

Automatic driving in expressways is generally considered to be the easiest for commercial landing, and vehicle behavior decision-making is the core of automatic driving technology, which directly affects the safety and comfort of vehicle driving. In this paper, an automatic lane change strategy based on prior knowledge and data-driven is proposed for expressway. Our method decouples autonomous driving decisions into two processes. Firstly, we propose an optimization model of lane change features based on prior knowledge, which not only provides a basis for the selection of vehicle features but also effectively reduces the interference of invalid features. The driving risk field is used to screen the vehicle targets that have potential influence on the lane change of the ego vehicle, and the maximal information coefficient evaluates the effectiveness of the features of the vehicle targets through the maximal information coefficient to optimize the lane change features. Then, a data-driven lane-changing model is proposed based on Attention-BiLSTM (Bi-directional Long Short-Term Memory), which transforms vehicle lane-changing into time series prediction, fully explores the relationship between lane-changing features in the context information, and effectively improves the anti-interference ability of the system against accidental errors. In addition, the attention mechanism can adaptively adjust the weight of lane-changing features and effectively capture the important information that needs attention in different lane-changing states, further improving the accuracy of the model. Finally, the validity of the model is verified by remote sensing data sets and real vehicle experiments. The test results show that the test accuracy of our method is 95.1% in data sets and 94.2% in real vehicle experiments, which fully meets the requirements of lane change decisions about automatic driving.

Vehicle Driving Behavior Analysis and Unified Modeling in Urban Road Scenarios

To improve the simulation accuracy and efficiency of microscopic urban traffic, a unified modeling method considering the behavioral characteristics of vehicle drivers is proposed by considering the lane-changing vehicles on the inlet lanes of signalized intersections and their approach following vehicles on the target lanes as research objects. Based on the driver’s multidirectional, multi-vehicle anticipation ability and introducing lateral vehicle influence coefficients, the full velocity difference car-following model was extended to microscopic traffic models that consider the driver’s capacity for multi-directional, multi-vehicle anticipation. The extended model can describe longitudinal movements of lane changing and car followers using lateral vehicle influential parameters. The influences of traffic control signals and the type of lane change on drivers’ decisions were integrated into the model by reformulating the optimal velocity function of the basic car following the model. Similar modeling methods and components were applied to formulate four groups of experimental models and one group of test models. Vehicle trajectory data and manual observations were collected on urban arteries to calibrate and evaluate the research models, experimental models, and test models. The results show that the car-following behavior is more sensitive to the variation in the status of the lateral moving vehicle and change of lane-changing type compared to lane-changing behavior during the lane-changing process. In addition, when lane changing gradually encroaches on the target lane, the vehicle observes the driving conditions and adjusts its driving behaviors differently. This research helps to analyze travel characteristics and influence mechanisms of vehicles on urban roads, which is a guide for the future development of sustainable transportation and self-driving vehicles and promoting the efficient operation of urban transportation systems.

Human Knowledge Enhanced Reinforcement Learning for Mandatory Lane-Change of Autonomous Vehicles in Congested Traffic

Human Knowledge Enhanced Reinforcement Learning for Mandatory Lane-Change of Autonomous Vehicles in Congested Traffic

Research on lane-changing decision and control of autonomous vehicles based on game theory

Designing a secure and reliable decision-planning model for vehicle lane changing is of utmost practical significance because it is one of the most frequent driving behaviors and has a substantial impact on the safety of drivers’ lives and property. First, a Gaussian mixed Hidden Markov model (GMHMM) is trained for lane change intention recognition (LCIR), and the results reveal that the model has a great performance. This will simplify the game process and provide drivers and passengers with warnings. Second, the safety, efficiency, and comfort payoffs of vehicle lane changes are taken into account when building the game model. When building the safety payoff function, temporal collision risk and spatial collision risk of vehicles are two of them that are carefully taken into account. After that, the vehicle’s trajectory tracking control is decoupled into lateral LQR + feedforward control and longitudinal dual proportional integral derivative (PID) control based on the Frenet coordinate system. Finally, a vehicle lane change scenario is built for simulation analysis, and the effects of driving comfort factor and driving efficiency factor on lane change results are considered. The results show that the proposed game theory lane change model ensures lane change safety while satisfying human drivers’ requirements for lane change efficiency and comfort.

Three dimensional tracking of rigid objects in motion using 2D optical flows

Detecting and tracking objects in image sequences is paramount for any video analytic. While object detectors have become increasingly robust, motion estimates based on the straightforward approach of running a detector and linking detections are prone to bounding box noise. An algorithm for the monocular estimation of the 3D motion of rigid objects is presented, combining an object detector, minimal camera calibration, and 2D optical flows observed on the image plane. The algorithm utilizes the 2D Delaunay triangulation over geometrically consistent optical flow tracks to identify regions common to the same object. The algorithm is evaluated on the special case of image sequences of vehicles on roads. Experiments on BrnoCompSpeed dataset show that speed estimates from both detector-based tracking (mean error of 1.62 km/h) and optical flow based tracking (mean error of 2.19 km/h) perform competitively on straight roads. An extensive empirical evaluation is also conducted on a new dataset containing synthetic and real world scenes that also include vehicle trajectories involving rotation. A naive baseline bounding box track based method obtained a mean error of 5.29 km/h for vehicle speed, but optical flow tracks performed significantly better with mean error of 1.65 km/h on the new dataset including rotation. With the new method, video analytics such as vehicle speed estimation and lane change detection can obtain precise information about the 3D trajectory of rigid objects in motion.

Coordinated Decision Control of Lane-Change and Car-Following for Intelligent Vehicle Based on Time Series Prediction and Deep Reinforcement Learning.

Adaptive cruise control and autonomous lane-change systems represent pivotal advancements in intelligent vehicle technology. To enhance the operational efficiency of intelligent vehicles in combined lane-change and car-following scenarios, we propose a coordinated decision control model based on hierarchical time series prediction and deep reinforcement learning under the influence of multiple surrounding vehicles. Firstly, we analyze the lane-change behavior and establish boundary conditions for safe lane-change, and divide the lane-change trajectory planning problem into longitudinal velocity planning and lateral trajectory planning. LSTM network is introduced to predict the driving states of surrounding vehicles in multi-step time series, combining D3QN algorithm to make decisions on lane-change behavior. Secondly, based on the following state between the ego vehicle and the leader vehicle in the initial lane, as well as the relationship between the initial distance and the expected distance with the leader vehicle in the target lane, with the primary objective of maximizing driving efficiency, longitudinal velocity is planned based on driving conditions recognition. The lateral trajectory and conditions recognition are then planned using the GA-LSTM-BP algorithm. In contrast to conventional adaptive cruise control systems, the DDPG algorithm serves as the lower-level control model for car-following, enabling continuous velocity control. The proposed model is subsequently simulated and validated using the NGSIM dataset and a lane-change scenarios dataset. The results demonstrate that the algorithm facilitates intelligent vehicle lane-change and car-following coordinated control while ensuring safety and stability during lane-changes. Comparative analysis with other decision control models reveals a notable 17.58% increase in driving velocity, underscoring the algorithm's effectiveness in improving driving efficiency.

Optimal lane-changing and collision avoidance strategy for intelligent vehicles in high-speed driving environments

This paper proposes an optimal lane-changing and control strategy to solve the collision avoidance of intelligent vehicles in highway driving. A quintic polynomial lane-changing trajectory model is established in the Frenet frame, and the trajectory planning problem is transformed into the optimisation problem of lane-changing target coordinates. To consider the multiple environmental factors on vehicle lane-changing, the feasible region of the lane-changing target point was generated based on the lateral acceleration limitation and the safe distance constraints. The performance evaluation indexes were established, considering efficiency, comfort, lateral slip, and lane offset. The multi-objective optimisation problem is solved in the feasible region, and the optimisation results were applied to the collision detection process. In addition, considering the nonlinear characteristics of tire force caused by sudden lateral movements of vehicles at high-speed, a vehicle dynamics model based on the Magic Formula is established, and Model Predictive Control is applied to track the expected trajectory. Simulation results show that this strategy can plan suitable lane-changing trajectories in high-speed working conditions and achieve vehicle collision avoidance.

Vehicle Lane Change Multistep Trajectory Prediction Based on Data and CNN_BiLSTM 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.

Energy-optimal Speed Regulation for CAVs in Urban Mixed Traffic with Induced Lane Changes

Existing urban energy-optimal vehicle speed regulation (eco-driving) studies were mostly carried out on single-lane roads or did not consider lane changes, delivering overidealized eco-driving results. In addition to lane changes that every vehicle may encounter on roads, the eco-driving connected automated vehicles (CAVs) are subject to induced types of lane changes of manually-driven vehicles (MVs), a cut-in of an MV to the front of a CAV from an adjacent lane, and an escape of an MV from behind a CAV to an adjacent lane. So far, very few studies have considered the impact of lane changes while designing the ecodriving strategies and evaluated the eco-driving performance with respect to lane-changing disturbances in credible simulation environments; and almost no existing work has addressed both lateral and longitudinal disturbances for eco-driving. This paper has developed a generic and deployable eco-driving strategy for CAVs under a unified rolling-horizon optimal control framework that can deal with both longitudinal and lateral disturbances of MVs without assuming communications between CAVs and MVs. The performance of this disturbance-rejection eco-driving strategy was thoroughly evaluated with regard to a multilane urban network, using the microscopic traffic simulator SUMO. The strategy was demonstrated very effective and robust, and similar results were little reported before.

A heuristic-action-involved safe lane-change of autonomous vehicles with multiple constraints on road centerline and speed under highway environment

Lane-change (LC) is one of the most important topics in autonomous vehicles (AVs) on highways. To enhance the implementation of effective LC in AVs, this paper proposes a framework based on deep reinforcement learning, which takes into account heuristic actions and multiple constraints related to the centerline of the road and speed, to improve the overall performance of LC in AVs. Firstly, the influence of unreasonable vehicle actions on the algorithm training process is studied. To improve the rationality of the to-be-trained actions, a novel reasonable action screening mechanism is proposed. Secondly, to keep the vehicle on the centerline of the lane and avoid the collision with other vehicles, a method is designed to calculate the center position of the vehicle. Thirdly, a segmented speed reward mechanism is proposed to constrain vehicle speed. Subsequently, a dynamic reward function is established to train the control algorithm. Lastly, the proposed strategy is evaluated in two simulation scenarios of highways. The simulation results show that the proposed method can increase the number of reasonable actions by more than 30% and improve the success rate of obstacle avoidance with the increase of over 52% in both static and dynamic scenarios compared with the benchmark algorithms.

Study on Anthropomorphic Lane Changing Decision Making for Smart Trucks Based on Driving Behavior

In order to improve the applicability of the lane change decision model to different styles of drivers, especially for trucks with less flexibility and stability of their own, safe and reasonable lane change decisions are more important. In this paper, based on the research of integrating personalized driver styles in automatic lane change control technology, a vehicle lane change decision model considering driver styles is established. Firstly, the HighD dataset is screened and the drivers are classified by driving style using principal component analysis and K-means (K-means) cluster analysis. Secondly, we propose a lane change decision model that takes into account the rationality and safety of lane change by learning human drivers' driving behavior experience through Long Short-Term Memory (LSTM) method, so as to improve drivers' acceptance and satisfaction of smart cars. Finally, the joint simulation by Simulink/CarSim/PreScan software proves that the lane change decision method based on the driver's style proposed in this study can realize the autonomous lane change task of smart cars.

Automated Speed and Lane Change decision-making Model using Support Vector Machine

One of the major obstacles that the auto industry must overcome is the rise of autonomous vehicles. The study of lane-changing is an important part of this problem. Previous studies on autonomous vehicle lane changes have predominantly focused on lane change path planning and path monitoring, with limited attention given to the autonomous vehicle's lane change decision-making process. This paper introduces a novel Lane Change Decision-Making Model for autonomous vehicles using the Support Vector Machine (SVM) method. The suggested model employs real-time sensor data to assess whether or not a lane change is possible, taking into account the proximity of other vehicles (cars, buses, motorbikes), and adjusting speed as necessary to ensure a seamless transition. Researching the various facets of lane changes in autonomous vehicles allows for decision-making that is grounded in utility, safety, and tolerance. The implementation of a support vector machine (SVM) technique with Bayesian parameter optimization is used to deal with the non-linearity and complexity of the process of autonomous lane change decision-making. Finally, we compare the suggested SVM model against a rule-based lane change model using the test data. The SVM-based strategy is shown to improve lane change decision-making in a comprehensive simulation exercise, which in turn improves the safety and efficiency of autonomous driving systems. The experiment also use a real vehicle to gauge the efficacy of the underlying decision-making model.

Empirical evaluation of change in crash risk due to lane marking reallocation: A case study in Kochi City, Japan

Lane markings are considered an essential component of a road system. As highlighted in previous studies, they are directly linked to efficiency and safety. The rearrangement or reallocation of lane markings can be an economical way to improve efficiency. However, such changes could influence driver behavior. Thus, there is a tradeoff between efficiency and safety. Through a case study in Kochi City, Japan, this study evaluated the change in crash risk caused by a lane marking reallocation. Video data were collected before and after the implementation of a new road layout (achieved by reallocating lane markings) that was intended to mitigate traffic congestion at a signalized intersection. Based on the video data, PICUD (Possibility Index for Collision with Urgent Deceleration), a surrogate safety index used to estimate collision risk, was estimated for lane changes and conflicts between leading and following vehicles in the through lane. In particular, it was confirmed that the collision risk between a lane-changing vehicle and a leading vehicle in the through lane was reduced due to the reduction in traffic density caused by the new road layout. In addition, the results indicate that the PICUD value tends to decrease (i.e., the crash risk tends to increase) with increasing speed of the following vehicle relative to the leading vehicle. Overall, the improvement in safety after the implementation of the new road layout was marginal and statistically insignificant. Therefore, this study highlights the necessity of incorporating speed control measures, such as speed limits, along with congestion alleviation measures in order to enhance safety.

Vehicle Lane Change Models—A Historical Review

Lane changing is a complex operation that has a significant impact on traffic safety. The accurate identification and assessment of potential risks in the driving environment before lane changing is crucial for the safe and smooth completion of a lane change. In this paper, the research status of vehicle lane change models is reviewed. Firstly, various factors affecting lane change models are analyzed. Different drivers will be affected by vehicle dynamic parameters, vehicle driving states, and driver characteristics under various road environments. Secondly, the vehicle lane change models are divided into four types: the empirical model of lane changing, the physical model of lane changing, the cognitive model of lane changing, and the mixed model of lane changing. The advantages and disadvantages of different types of lane change models are analyzed, and the key problems to be solved by different lane change models are expounded, respectively, from the aspects of input variables and reasoning algorithms. Finally, according to the advantages and disadvantages of different lane change models, a future research direction is proposed.

Cooperative decision-making of multiple autonomous vehicles in a connected mixed traffic environment: A coalition game-based model

Advances in vehicle-networking technologies have enabled vehicles to cooperate in mixed traffic. However, realizing the cooperative decision-making of multiple connected autonomous vehicles (CAVs) when influenced by the presence of connected manual vehicles (CMVs) is a challenging area in current research. In this study, we propose a coalition game-based (CG-based) model for multi-CAV cooperative decision-making in a connected mixed traffic environment. First, the model integrates the perceived risk field theory, quantifying the driving risk from the perspective of different CMVs; this risk is used to determine the uncertainty of the motion state of CMVs. Second, the model can identify the conflicts caused by multiple lane-changing vehicles and decouple the conflict problem into multiple two-vehicle lane-changing games, including a cooperative game between two CAVs and a non-cooperative game between a CAV and a CMV. To test the proposed model, four scenarios that blocked the passage of multiple CAVs were set up; in these scenarios, the average speed of the CG-based model was 21.05, 16.76, 23.17, and 12.55% higher than that of the LC2013 model. The simulation results showed that the CG-based model could improve the efficiency of multiple CAVs while ensuring safety in a mixed traffic flow.

On gap matching for lane-changing vehicles: Coordinated strategies to improve traffic flow at weave bottlenecks

The weaving section on the freeway will be an active bottleneck due to vehicles’ mandatory or discretionary lane-changing (LC) conflicts. In the early stages of the development of intelligent connected transportation, destination information collection of weaving vehicles is possible, which provides an opportunity to improve performance in weaving traffic. In this paper, a simulation system with coordinated strategies embedded to satisfy the LC needs is presented under the weaving traffic flow. First, an algorithm is proposed to evaluate the balance between the demand for LC vehicles and the available gaps provided in weaving traffic, which can be adjusted by variable speed limit (VSL) and ramp metering strategies. Second, to validate the proposed algorithm, a typical weave section with a merge and a diverge connected through a single auxiliary lane is researched, and a simulation platform based on cellular automata has been developed. The optimal control strategies are obtained by model predictive control (MPC) and taking the minimum travel time as an objective. Third, the advantages of coordinated control strategies are studied and discussed in traffic efficiency improvement and the success of LC proportions under different combinations of weaving ratios and large vehicle traffic mix.

Optimized deep extreme learning machine for traffic prediction and autonomous vehicle lane change decision-making

Deep learning occupies a prominent position in the field of machine learning, where researchers have introduced autoencoder into extreme learning machine (ELM) to establish deep extreme learning machine (DELM), with the aim of enabling autonomous feature learning from data. This approach has found widespread applications in various domains, including finance and electrochemistry. However, the network accuracy of DELM is sensitive to the initial weighting and bias configurations, which can be negatively impacted by random initialization. To mitigate the adverse effects of such random initialization on network performance, this study employs two swarm intelligence optimization algorithms, particle swarm optimization (PSO) and harris hawks optimization (HHO), to optimize the input weights and biases of DELM. In addition, we apply the variational mode decomposition (VMD) technique to decompose the original data, enabling a deeper exploration of latent information within the data. Subsequently, two integrated models are constructed: VMD-HHO-DELM and VMD-PSO-DELM. We apply these models to address road traffic condition prediction and autonomous lane change decision-making problems, with a specific focus on the transportation domain. Through extensive experimentation on two real-world datasets, we observe that deep learning models based on swarm intelligence optimization algorithms exhibit excellent performance across various scenarios. In the context of road traffic condition prediction, our experiments reveal superior results under different time intervals and traffic volumes compared to traditional models, all while maintaining higher operational efficiency. For the autonomous lane change decision experiment, we establish a driver dissatisfaction model that aids in formulating more reasonable lane change decisions, thus reducing the frequency of abrupt lane changes.

Deep Q-Network-Enabled Platoon Merging Approach for Autonomous Vehicles

Platoon merging control of autonomous vehicles driving in a platoon formation can improve traffic efficiency. However, current platoon merging control approaches primarily rely on rules, making it challenging to achieve optimal control. In this study, we propose a platoon merging control approach based on a deep Q-network (DQN). First, we specify the state and action of the vehicle and establish a set of reward functions to ensure safe driving. Then, the DQN algorithm is used to train a neural network suitable for merging the controls of connected and automated vehicles (CAVs) to continuously approach the state-value function. Finally, we compare the proposed approach with a rule-based (RB) vehicle merging approach using a MATLAB simulation. In particular, CAVs are driven simultaneously using the two approaches in a random environment. The simulation results show that the proposed DQN-based vehicle merging approach requires less merging travel time and fewer vehicle lane change times than the RB approach. Additionally, merging can result in improved capacity in medium and high traffic densities compared with no-merging: the higher the CAV penetration rate, the larger the improvement. We verify the effectiveness of the proposed approach for different initial conditions. We suggest that the proposed method is a safe and robust method for CAV platoon merging, and that it can be applied to increase the capacity of freeways and roads.

Predictive Analysis of Vehicular Lane Changes: An Integrated LSTM Approach

In the rapidly advancing domain of vehicular traffic management and autonomous driving, accurate lane change predictions are paramount for ensuring safety and optimizing traffic flow. This study introduces a comprehensive two-stage prediction model that harnesses the capabilities of long short-term memory (LSTM) for anticipating vehicular lane changes. Initially, we employed a variety of models, such as regression methods, SVMs, and a multilayer perceptron, to categorize lane change behaviors. The dataset was then segmented based on vehicle trajectories and lane change patterns. In the subsequent phase, we utilized the superior classification outcomes from LinearSVC to curate our training data. We developed two dedicated LSTM networks tailored to specific datasets: the lane-keeping LSTM (LK-LSTM) and the lane-changing LSTM (LC-LSTM). By integrating insights from both models, we achieved a comprehensive prediction of vehicular lane changes. Our results indicate that the unified prediction model markedly enhances prediction precision. Accurate lane change predictions offer valuable contributions to advanced driver-assistance systems (ADAS), with the potential to minimize traffic mishaps and enhance traffic fluidity. As we transition to a more autonomous automotive era, refining these predictions becomes essential in seamlessly merging human and automated driving experiences.