This paper proposes a novel two-level route layout mode for the autonomous modular bus (AMB) based urban public transportation corridor. The two-level route consists of two components: mainline and feeder routes. The mainline covers all stations along the corridor, while feeder routes serve only specific consecutive stations. The departure and terminal stations of the feeder route are set up as special stations where AMBs can combine and split. The feeder route layout scheme remains fixed during an operating period but can be adjusted across different periods. We establish a mixed-integer programming model to concurrently determine the optimal special station deployment scheme, AMB timetable, and vehicle formation plan under the two-level route layout mode. The objectives are to minimize the average passenger waiting time, AMB energy consumption, and the number of special stations. Subsequently, the model is addressed by employing the multi-objective boxing match algorithm. A case study is conducted utilizing a real-world urban transportation corridor, comparing the proposed method against an optimization approach that concentrates on AMB timetabling and vehicle formation planning without considering feeder routes. The results underscore the superiority of our method, which reduces average passenger waiting time by 13.00% and vehicle energy consumption by 2.53%, without increasing the number of required special stations.

Autonomous vehicles (AVs) offer a promising glimpse into a future where transportation is smarter, safer, and more streamlined. Nevertheless, as AVs continue to interact with conventional vehicles (CVs), the potential for increased complexities and challenges cannot be overlooked, such as the frozen robot problem. This study proposes a regret-based model for motion planning responsibilities, encompassing self-respect and courtesy for conflicting personal interests. By incorporating these reciprocal responsibilities, socially compatible driving behaviors are promoted, and uncertainties in behavior are also reduced. A Self-Respect-Courtesy (SR-C) plane is further introduced, illustrating the interaction intensity and tendency. To navigate the trade-offs of responsibilities in varying situations, the concept of environmental niche is provided. Niches help to characterize the outcomes of specific actions with the resulting conditions to fulfill responsibilities. Finally, a hierarchical multi-agent inverse reinforcement learning algorithm is designed to calibrate the proposed model with NGSIM highway lane-changing cases. We found that the proposed model can significantly improve the calibration results and reduce the predictions error of mandatory lane changes by up to 20%. Moreover, the cross-entropy error also significantly decreases in a stable stage, indicating that responsible actions can safely reduce the behavior uncertainties of interactions. Our research revealed that drivers prioritize courtesy responsibility in discretionary lane changes with more consistency, whereas their self-respect preferences are stronger but show more variability in mandatory lane changes. These findings provide valuable insights into the underlying mechanism of interactions.

Accurately predicting the trajectories of surrounding agents is critically important for the safe operation of autonomous vehicles. However, complex environments not only increase the computational cost but also introduce multimodality in the agent’s future motions, presenting significant challenges for motion prediction. Therefore, we propose FSGA, a trajectory prediction framework that constructs a relationship graph between agents and lanes in future spacetime and utilizes a graph attention mechanism to analyze the interactions between these future spacetime agents and lanes. We employ a vectorized representation method to model the traffic scene as a graph. Through this vectorized scene representation, FSGA extracts spatio-temporal features from agents’ historical trajectories, including interactions between agents, and between agents and lanes. To reduce computational costs, a lane scoring mechanism is employed, allowing the model to focus only on the lanes that have a significant impact on the current agent’s motion. To address complexity and multimodality of future trajectories, we introduce a branching tree goal prediction structure that models multiple possible motion trends by representing waypoints and endpoints at different temporal stages. Based on these predicted goals, we construct a relationship graph that captures the associations between agents and lanes in future spacetime and obtains the interactions between agents and lanes, improving trajectory prediction accuracy. Experiments show that FSGA achieves state-of-the-art performance on Argoverse1.

The emerged integrated chassis is playing an increasingly pivotal role in enhancing vehicle stability, which typically needs estimation of the stability envelope and proper coordination strategy for chassis controller design. The conventional stability envelope determined in phase plane ensure vehicle self-stability, while it neglects the influence of chassis actuators. To play out full performance of integrated chassis, this study proposes a novel extended stability envelope analysis method to quantify the effect of different chassis actuator combinations on vehicle lateral dynamic. In energy phase plane, the state change direction with influence of control inputs are employed to judge whether a given state will go beyond the tire’s grip limit. Accordingly extended stability envelopes of some typical actuator combinations, i.e., active front steering (AFS) and active rear steering (ARS), AFS and torque vectoring control (TVC), ARS and TVC, are determined. Meanwhile, principle to quantify the overlapping effectiveness range and unique effectiveness range between ARS and TVC is introduced, accordingly effectiveness factors used for coordination control are discussed. In summary, the proposed method in this article has considerable potential for coordination control of integrated chassis, by providing comprehensive analysis method to estimate the extended stability boundary and quantify the effectiveness range with influence of actuators taken into account.

Car-following behavior is one of the most common driving behaviors. To reduce the impact of driver reaction delay and accumulated errors in predicting long sequences on the accuracy of speed prediction, we propose a deep learning car-following model based on a sequence-to-sequence (seq2seq) architecture with an attention mechanism. Firstly, we analyze the characteristics of driver reaction delay during the car-following process and design an attention mechanism to learn the probability distribution of driver reaction delay. This allows the model to consider more environmental information at the moment when the driver actually makes a decision, rather than just the current environmental information, during the prediction process. By utilizing the seq2seq architecture to model car-following behavior, the model focuses more on reducing the impact of accumulated errors during the training process. Additionally, we propose a temporal consistency constraint loss to improve the robustness and stability of the car-following model’s training method and enhance the prediction results. Finally, we use Gated Recurrent Units (GRU) and the sequence-to-sequence model (seq2seq) as baselines, and simulation results demonstrate that our model achieves more accurate car-following behavior prediction.

Trajectory tracking is a crucial responsibility for autonomous vehicles as they strive to avoid collisions. During combined-slip emergency situations where steering and driving/braking joint control are required, the nonlinearity and coupling of tire forces become increasingly important, rendering a linear tire model-based controller ineffective and leading to degraded path-tracking performance. Such degradation can ultimately jeopardize vehicle stability. To address the aforementioned issue, we establish a hierarchical coordinated controller for four-wheel independent drive (4WID) autonomous vehicles, specifically tailored to handle combined-slip trajectory tracking and yaw stability control, considering variable tire cornering stiffness. At the upper level, a model predictive lateral motion controller is engineered based on a novel combined-slip UniTire-Ctrl model. The predictive model captures the intricate nonlinear and coupling characteristics of tire forces through an analytical expression of effective cornering stiffness. This enables the controller to account for the impact of longitudinal force on lateral motion control and coordinate the front-wheel steering angle and direct yaw moment in an efficient manner. Additionally, a linear quadratic longitudinal motion controller is developed to follow the desired longitudinal speed. The lower-level torque distribution controller is constructed to prioritize vehicle stability by minimizing tire adhesion utilization. Finally, the effectiveness of the controller under combined-slip conditions is validated through the CarSim and Matlab/Simulink co-simulation platforms, which demonstrates that the developed combined-slip motion controller with UniTire-Ctrl model exhibits superior tracking precision and stability under extreme combined-slip conditions.

Eco-driving can reduce vehicle energy consumption and carbon dioxide emissions. However, the effect of eco-driving training can diminish over time. It is necessary to provide the drivers with continuous feedback. Meanwhile, bus operators also need a fair incentive system to encourage drivers to drive ecologically to reduce energy costs. Considering the influence of traffic conditions, ambient temperature, and passenger load on the energy consumption of battery electric buses, a quantitative evaluation method for eco-driving with energy consumption as a single evaluation index is proposed. Specifically, the traffic conditions recognition method based on low-frequency data is constructed and then the division of ambient temperature range is discussed. According to the traffic conditions and ambient temperature, the actual operation data of 19 battery electric buses in one year are divided into 12 control groups and the reference energy consumption of each control group is obtained. The reference energy consumption describes the range of variation in energy consumption changes for different traffic conditions and ambient temperatures. In addition, to describe the impact of passenger load on bus energy consumption, a passenger load conversion factor is proposed. Finally, the eco-driving evaluation method is constructed using the reference energy consumption and the passenger load conversion factor. Since factors of the traffic conditions, ambient temperature, and passenger load are integrated into the eco-driving evaluation method design, the score depends only on the driver’s eco-driving level and the results show that efficient drivers will not get lower scores due to driving in poor driving conditions.

Intelligent tire systems are promising solutions for achieving precise vehicle state estimations, localization, and motion control in the context of autonomous driving. Tire cornering properties, namely, lateral force, aligning moment, and pneumatic trail, are crucial factors that should be accurately estimated for vehicle dynamics control purposes. In this work, a soft sensor for estimating tire cornering properties based on intelligent tire and machine learning is developed. The intelligent tire system is based on a triaxial accelerometer mounted on the inner liner of the tire tread, which provides acceleration measurements from the$x$,$y$, and$z$directions. Partial least squares and variable importance in the projection scores (PLS-VIP) are used in the feature extraction of the acceleration signals over the contact patch. A Gaussian process regression (GPR) model is trained to predict the cornering properties with confidence intervals under different input conditions. Based on the variances in the GPR predictions and minimum mean-square error criterion, a data fusion method for pneumatic trail estimation is proposed. It is demonstrated that the developed GPR models for cornering properties and the data fusion method for pneumatic trail estimation have satisfactory accuracy and reliability. The experimental results show that the soft sensor proposed in this work is a strong candidate for further applications in the development of vehicle state estimation and control algorithms.

The concept of intelligent tires has drawn the attention of researchers in the areas of autonomous driving, advanced vehicle control, and artificial intelligence. The focus of this paper is on intelligent tires and the application of machine learning techniques to tire force estimation. We present an intelligent tire system with a tri-axial acceleration sensor, which is installed onto the inner liner of the tire. Neural Network techniques are used for real-time processing of the sensor data. The accelerometer is capable of measuring the acceleration in x,y, and z directions. When the accelerometer enters the tire contact patch, it starts generating signals until it fully leaves it. Simultaneously, by using MTS Flat-Trac test platform, tire actual forces are measured. Signals generated by the accelerometer and MTS Flat-Trac testing system are used for training three different machine learning techniques with the purpose of online prediction of tire forces. It is shown that the developed intelligent tire in conjunction with machine learning is effective in accurate prediction of tire forces under different driving conditions. The results presented in this work will open a new avenue of research in the area of intelligent tires, vehicle systems, and tire force estimation.

Tire states and capacity monitoring is critical for vehicle and wheel stabilization controls in automated driving and active safety systems. Tire capacity, which represents the performance margin of tire forces from its limits, determines the operational range for vehicle control systems and their actuation through steering or torques at each tire to maintain stability while performing trajectory following. This paper presents a generic tire capacity identification framework that can handle different normal loads, road surface friction, and combined-slip driving scenarios, which are challenging for stabilization and tracking control programs in automated driving systems. A novel measuring method for generating force-training data is designed by combining the indoor tire test procedure and tread rubber friction test rig, in order to obtain adequate and high-quality benchmark datasets. The results from large data sets from road experimenting and indoor tire test facilities, including pure- and combined-slip conditions, confirm effectiveness of the developed learning-based tire capacity estimation which utilizes notions from the model description with bounded uncertainty. More importantly, the proposed method can provide reliable tire properties ranging from the linear to the sliding regions. Further validation is performed on a real test car with on-board sensory measurements, and the results confirm accuracy of the proposed method for various free rolling and hard launch/brake scenarios.

Understanding the dynamic behavior of tires and their interactions with roads plays an important role in designing integrated vehicle control strategies. Accordingly, having access to reliable information about tire–road interactions through tire-embedded sensors is desirable for developing enhanced vehicle control systems. Thus, the main objectives of this research are: 1) to analyze data from an experimental accelerometer-based intelligent tire acquired over a wide range of maneuvers, with different vertical loads, velocities, and high slip angles and 2) to develop a lateral force predictor based on a machine learning tool, more specifically, the Gaussian process regression (GPR) technique. It is determined that the proposed intelligent tire system can provide reliable information about the tire–road interactions even in the case of high slip angles. In addition, lateral force models based on GPR can predict forces very well, outperforming other machine learning models and providing levels of uncertainty that can be useful for designing vehicle control strategies.

Taking advantage of available measurement in Internet of Things (IoT) for intelligent transportation systems, a sideslip angle estimation method for autonomous vehicles is presented and experimentally verified by fusing global navigation satellite system (GNSS) and inertial measurement unit (IMU), and by constructing an observability index (OI). The correlation between the vehicle sideslip error and the inertial navigation system (INS) heading error is presented first. Then, the observability for the heading error in a velocity-based Kalman filter is discussed and a novel index is defined to check the observability of the heading error. The course from a single antenna GNSS in an autonomous vehicle is augmented to estimate the heading error when the observability of the heading error is low. To reject the course measurement for scenarios that include sideslip movement, a binary hypothesis test approach is applied to indicate whether the vehicle is sidesliping. In addition, based on the OI and the sideslip indicator, a hybrid feedback strategy is designed for the heading error correction. To improve the convergence rate of the heading error in the velocity-based Kalman filter, a tuning strategy is presented. The stochastical observability of the designed Kalman observer is investigated for known and stochastic initial conditions. Finally, the proposed sideslip angle estimator is experimentally validated through a vehicle test platform in critical driving scenarios. The results confirm that the proposed OI can effectively identify when the heading error is observable, and also corroborate the effectiveness of the hybrid feedback strategy and adaptation method in the Kalman observer.