Search

We present a Vehicle Model (VM) that has 17 degrees of freedom and includes nonlinear tire and powertrain subsystems. Implemented as a relatively small piece of C++ code, the model runs vehicle dynamics 2000 times faster than real time at a simulation time step of $1 \times 10^{-3}, \text {s}$ on a single core of a commodity CPU. When executed on the GPU, one can perform 300000 vehicle simulations in real-time. These properties make the model a good candidate for reinforcement learning, model predictive control, model predictive path integral control, path planning, state estimation, and traffic simulation tasks. The model is expressive in that it can capture the dynamics of vastly different vehicles. This is demonstrated by first calibrating the model to mimic the dynamics of a 1/ $6^{th}$ scale vehicle called the Autonomy Research Testbed (ART) vehicle, which has a mass of approximately 5.8 kg. Subsequently, the model is calibrated to mimic the dynamics of a heavy-duty High Mobility Multipurpose Wheeled Vehicle (HMMWV), which has a mass of 2097 kg. The Bayesian calibration process discussed can $(i)$ handle real-life measurement noise, and $(ii)$ provide model parameter probability distributions. The vehicle model, which is open source and freely available in a public repository, can also be imported into Python owing to SWIG wrapping.

This article proposes a nonlinear observer for estimating a powertrain transmitted torque of vehicles equipped with dual-clutch transmissions. A new dynamical model of vehicle driveline is presented. It introduces a single characteristic state based on the dynamic friction model to scale the clutch torque capacity, which is available from the actuator position-to-torque map. Based on this model, the generated torque on the clutch disk can be effectively estimated in real time, and the sensitivity of the estimated value caused by high gain feedback can be reduced in the observer. Also, the engine torque uncertainty is estimated from the integrated driveline model. The Lyapunov-based analysis is presented to prove the convergence of the proposed observer. The effectiveness of the proposed method is shown by experimental validations on a test vehicle that allows for real measurements of the input shaft torque of transmission. [ABSTRACT FROM AUTHOR]

Vehicle Re-Identification (ReID) is of great significance for public security and intelligent transportation. Large and comprehensive datasets are crucial for the development of vehicle ReID in model training and evaluation. However, existing datasets in this field have limitations in many aspects, including the constrained capture conditions, limited variation of vehicle appearances, and small scale of training and test set, etc. Hence, a new, large, and challenging benchmark for vehicle ReID is urgently needed. In this paper, we propose a large vehicle ReID dataset, called VERI-Wild 2.0, containing 825,042 images. It is captured using a city-scale surveillance camera system, consisting of 274 cameras covering a very large area over 200 $km^2$ k m 2 . Specifically, the samples in our dataset present very rich appearance diversities thanks to the long time span collecting settings, unconstrained capturing viewpoints, various illumination conditions, diversified background environments, and different weather conditions. Furthermore, to facilitate more practical benchmarking, we define a challenging and large test set containing about 400K vehicle images that do not have any camera overlap with the training set. VERI-Wild 2.0 is expected to be able to facilitate the design, adaptation, development, and evaluation of different types of learning models for vehicle ReID. Besides, we also design a new method for vehicle ReID. We observe that orientation is a crucial factor for feature matching in vehicle ReID. To match vehicle pairs captured from similar orientations, the learned features are expected to capture specific detailed differential information for discriminating the visually similar yet different vehicles. In contrast, features are desired to capture the orientation invariant common information when matching samples captured from different orientations. Thus a novel disentangled feature learning network (DFNet) is proposed. It explicitly considers the orientation information for vehicle ReID, and concurrently learns the orientation specific and orientation common features that thus can be adaptively exploited via an adaptive matching scheme when dealing with matching pairs from similar or different orientations. The comprehensive experimental results show the effectiveness of our proposed method. [ABSTRACT FROM AUTHOR]

Due to the uncertainty of vehicle model parameters, modeling errors and external disturbances, the performance of path tracking control system is poor, especially under high velocity and large curvature extreme conditions. To address this issue, this paper presents a novel gain-scheduled robust control strategy based on linear parameter varying system with model predictive control (MPC) and H∞. Firstly, fully considering the influence of the time-varying characteristics of vehicle velocity and tire cornering stiffness on the path tracking system model, a novel linear parameter varying system model is built for path tracking control of autonomous vehicle. Then, a path tracking robust controller is designed based on gain-scheduled approach, and the linear matrix inequality (LMI) is applied to solve optimization problem, in which MPC and H∞ robust control theory are applied to the controller design process. Finally, the simulation experiments have verified that the proposed novel robust control strategy can improve the path tracking accuracy and ensure the vehicle lateral and roll stability, especially under high velocity and large curvature extreme conditions. Meanwhile, the proposed robust controller shows superiority to suppress the parameters uncertainty, modeling error and external disturbance. [ABSTRACT FROM AUTHOR]

We present a GPU-accelerated collision detection method for the navigation of vehicles in enclosed spaces represented using large point clouds. Our approach takes a CAD model of a vehicle, converts it to a volumetric representation or voxels, and computes the collision of the voxels with a point cloud representing the environment to identify a suitable path for navigation. We perform adaptive and efficient collision of voxels with the point cloud without the need for mesh generation. We have developed a GPU-accelerated voxel Minkowski sum algorithm to perform a clearance analysis of the vehicle. Finally, we provide theoretical bounds for the accuracy of the collision and clearance analysis. Our GPU implementation is linked with Unreal Engine to provide flexibility in performing the analysis. [ABSTRACT FROM AUTHOR]

Cooperative vehicle safety (CVS) systems have been receiving significant attention as key enablers of important intelligent transportation system (ITS) applications, such as cooperative collision warning, emergency braking, and auto driving. In CVS systems, vehicles periodically broadcast short packets that include various types of vehicular information, e.g., position and speed. In this paper, we propose an optimization method for the broadcast rate in vehicle-to-vehicle (V2V) broadcast communications at an intersection based on theoretical analysis. We consider a model in which the locations of vehicles are modeled separately as queuing, and running segments and derive key performance metrics of V2V broadcast communications through a stochastic geometry approach. We develop closed-form approximate formulas for the theoretical expressions because they are mathematically intractable. Based on approximate analysis, we optimize the broadcast rate such that the interference at an intersection is mitigated and the overall performance of V2V communications is maximized. Because of the closed-form approximation, the optimal rate can be used as a guideline for a real-time control-method, which cannot be achieved through time-consuming simulations. We evaluate our method through numerical examples and demonstrate its effectiveness. [ABSTRACT FROM AUTHOR]

Vehicle model recognition is a typical fine-grained classification task that has a wide range of application prospects in safe cities and constitutes a research hotspot in the field of computer vision. Vehicles in images can appear at various angles, resulting in large differences in appearance. The existence of “multiviews” renders vehicle model recognition challenging. Recent research on vehicle model recognition has not fully explored the pose information of vehicles in different images, resulting in low model performance. In this study, we use vehicle pose information to solve the multiview vehicle model recognition (MV-VMR) problem and design a convolutional neural network (CNN) model with embedded vehicle pose information, known as the embedding pose CNN (EP-CNN). The proposed model includes two subnetworks: the pose estimation subnetwork (PE-SubNet) and vehicle model classification subnetwork (VMC-SubNet). PE-SubNet extracts the vehicle pose information, including the pose features and vehicle viewpoint. In VMC-SubNet, considering the scale variation of vehicles, an improved squeeze-and-excitation (SE) block, named the MultiSE block is implemented. We embed the vehicle viewpoint into the MultiSE block, which reweighs each channel such that the extracted features elicit different responses to different viewpoints. Subsequently, the pose features and classification features are integrated for classification. Experiments are conducted on the benchmark CompCars web-nature and Stanford Cars datasets. The results demonstrate that the proposed EP-CNN method can achieve higher recognition accuracy than most classic CNN models and several state-of-the-art fine-grained vehicle model classification algorithms. Code has been made available at: https://github.com/HFUT-CV/EP-CNN. [ABSTRACT FROM AUTHOR]

The aim of the paper is to define a limit performance of highly efficient battery electric quadricycles for urban mobility. The vehicle is employed for an energy efficiency competition in which urban concepts compete. A multi-physics (thermo-electro-mechanical) Tank To Wheels (TTW) model has been developed and validated. Given the information on the track route (track map and elevation) and on the throttle input command, the model computes the vehicle power demand and cruising speed. The model is validated by means of both indoor and outdoor experimental tests. The validated model is employed for the optimization of the transmission gear ratio and of the driving strategy to minimize the overall energy consumption on a given track. Design variables are related to the transmission gear ratio and to the throttle command profile. The algorithm aims to minimize the energy consumption, including constraints on the maximum current provided by the battery and on the maximum time available to complete the lap. Comparison with some common driving strategies confirmed the effectiveness of the proposed solution, with a foreseen 9% reduction of the vehicle overall energy demand. The single seater ultra-efficient electric vehicle, compared to other urban quadricycles, has a battery capacity ten times lower and a 50% longer driving range. [ABSTRACT FROM AUTHOR]

Path planning and path following are key technologies for autonomous vehicles. Strong nonlinearity, coupling characteristics, parametric uncertainties, external disturbances, and complicated driving scenarios put forward great challenges in estimation, control, and optimization of autonomous vehicles. Multi-dimension information utilization offers an effective solution to cope with the foregoing challenges. The presented research proposes a path planning and tracking framework for optimal design and regulation of control variables for collision avoidance and active guidance system, aiming to improve tracking performance and system robustness. For path planning approach, a virtual harmful potential field in 3 dimension (3D) is built to provide a collision-free trajectory as a reference path for path following, depending on vehicle kinematic model and boundary conditions of roads. A linear parameter-varying, polytopic vehicle lateral model is developed to address the issues of time-variable longitudinal speeds and preview distance on vehicle lateral stability management. Further, a strong robust gain-scheduling path following control approach based on linear matrix inequality methodology is proposed to tackle problems relating to characteristics of temporal variations, parametric uncertainties and external disturbances. Comparisons study under the TruckMaker/Xpack4-RapidECU joint HIL platform demonstrate that both tracking precision and high robustness to parametric uncertainties and external disturbances are superior than model predictive control method. The presented path following method provides an important insight into exploiting the robust control method's favorable merits while guaranteeing its computing efficiency. [ABSTRACT FROM AUTHOR]

This article deals with the control chassis dynamics of a light electric vehicle. In the frame of Global Chassis Control, it proposes a hierarchical control approach focuses on compensating the chassis dynamics against driver disturbances. It is based on a supervisory structure consists of two main levels, a global controller for regulating chassis variables and a local controller to each suspension system. Both controllers are designed using a fractional-order robust controller, namely CRONE controller, to provide an optimized solution for the performance-robustness tradeoff. The main objective is to improve ride comfort for passengers while respecting road holding and handling criteria. In contrast to other studies, random road disturbances are considered to investigate the limits of the proposed approach. Analysis in frequency and time domain has been done to evaluate the performance and robustness of the designed controller. Simulation results based on a full nonlinear 14 degree of freedom vehicle model show that the proposed strategy can effectively improve the ride quality in the field of interest where a significant performance in driver disturbance rejection is reported. [ABSTRACT FROM AUTHOR]

Model predictive control (MPC) in automotive active safety applications has gained great success in recent years. Detailed vehicle models along with a receding horizon control scheme seek the optimal control distribution among actuators for enhanced performances while satisfying system constraints. To promote the re-usability and scalability of MPC-based vehicular active safety systems, an agent-based MPC (AMPC) is proposed in our previous study for a modularized control architecture. In this paper, implementation of the control scheme is demonstrated on an all-wheel-drive test platform. Both centralized and agent-based MPCs for vehicular stability are compared for their control performances as well as computational costs on embedded hardware. It is shown from experimental results that agent-based MPC is more flexible and computationally efficient in handling vehicle active safety challenges while gives no compromise to control performances compared to their holistically formulated counterparts. [ABSTRACT FROM AUTHOR]

In performance tests, replacing humans with robotic drivers has many advantages, such as high efficiency and high security. To realize the vehicle speed tracking control with a robotic driver, this article proposes a novel deep reinforcement learning (DRL) approach based on deep deterministic policy gradient (DDPG). Specifically, the design of the approach includes state space, action space, reward function, and control algorithm. Then, to shorten the training time, the proposed approach utilizes the basic fundamental relationship between vehicle speed and pedal opening to intervene in network exploration. Furthermore, to solve speed fluctuations in low-speed sections, the replay buffer is optimized by adding weighted training samples. Experiments are conducted on fifteen cars, and results show that the proposed algorithm can effectively control the vehicle speed. Generally, it only needs three or four episodes of training to meet the requirements. Compared with the Segment-PID method, the proposed method has a smoother speed and fewer overbound times. Note to Practitioners—This article was motivated by the algorithm of deep reinforcement learning (DRL), which can be applied to automatic control including vehicle speed tracking control with a robotic driver. With the complexity, the delay of the vehicle model, and the dead zone of the pedal, it is difficult for existing classic control theories to achieve satisfactory vehicle speed tracking. This article analyzes the mathematical model based on the DRL theory and the vehicle dynamic model, then builds the speed tracking controller based on the deep deterministic policy gradient (DDPG) algorithm, and makes corresponding optimizations for the problems in the experiment. Through the control software integrated with the DDPG algorithm, the operator can easily realize the training and testing of the neural network, which greatly reduces the time and cost of vehicle testing and is of great significance to the emission certification of new vehicles. [ABSTRACT FROM AUTHOR]

This article proposes a novel gain-scheduling control method to improve vehicle lateral stability based on a variable-parameter-dependent approach. A time-varying velocity-dependent model to describe the vehicle lateral dynamic characteristics is constructed under cornering stiffness uncertainties and controller saturation. A novel Lyapunov function with variable-parameter-dependent Lyapunov matrix is established to develop the conditions. Using the proposed Lyapunov function, a general condition of the variable-parameter-dependent saturated robust yaw moment controller is provided in terms of time-varying parameter-dependent matrix inequalities. In order to obtain a tractable solution, a condition is further developed with finite linear matrix inequalities. Moreover, an optimal distribution method is adopted to generate the desired yaw moment based on torque allocation. These torques are computed by optimizing the distribution errors and tire workloads. Simulations results under both J-turn and double-lane changing scenarios are used to illustrate the merits of the proposed method. The control synthesis approach is also applicable to other applications involving time-varying parameters. [ABSTRACT FROM AUTHOR]

This paper analyzes priority-based distributed access control in vehicular networks. Characterizing the spatial correlation between vehicles, users, and roads, this paper uses Cox point processes to model vehicles and users, respectively. Then, this paper proposes that each vehicle transmits if it has the highest priority among all of its neighbors. Using the stochastic geometry framework, this paper derives the probability that a typical vehicle is retained to transmit. Then, this paper derives a second-order moment measure of the retained point process to evaluate the signal-to-interference (SIR) distribution of a typical user. The derived SIR distribution is compared to the system-level simulation of the SIR distribution of a typical user. [ABSTRACT FROM AUTHOR]

Longitudinal velocity, lateral velocity, and front wheel steering angle are crucial states for vehicle active safety features. However, direct measurement of these variables requires expensive measurement instruments and can be easily affected by vehicle nonlinear dynamics, outliers and noise pollution. Moreover, unknown inputs bring great challenges to their accurate estimation. Therefore, this paper develops a novel robust unscented M-estimation-based filter (RUMF) for vehicle state estimation with unknown driver steering torque. The nonlinear system model is constructed based on the vehicle dynamics model. Unscented transformation (UT) and statistical linearization are implemented to transform the nonlinear process and measurement function into a linear-like regression form with data redundancy to achieve outlier suppression. Then, an M-estimation-based iterated algorithm is designed to address process uncertainty and innovation and observation outliers for robust vehicle state estimation. The iteratively reweighted least-squares (IRLS) method based on the M-estimation methodology is utilized for unknown input estimation. Simulations and experimental results compared with extended Kalman filter (EKF) and particle filter (PF) have verified the effectiveness and robustness of the proposed algorithm. [ABSTRACT FROM AUTHOR]

Appropriate vehicle lateral stability control is the key to ensure vehicle driving safety, whereas accurate lateral stability recognition can help improve the performance of vehicle control. In this article, the vehicle stability recognition and coordinated control are studied. Firstly, the vehicle dynamic model is established, through vehicle simulation software Carsim, the attribute dataset representing the vehicle lateral stability is further obtained. Then, clustering by fast search and find of density peaks method (CFSFDP) based procedure for the classification of the lateral stability as ‘Absolutely stability’, ‘Nearly stability’ and ‘Hardly stability’ is applied. The brain emotional learning network combined with genetic algorithm (GA-BEL) model is used to train datasets to recognize vehicle stability categories during driving. For the different recognition results, a coordinated control strategy based on active front steering control (AFS) and direct yaw moment control (DYC) is designed finally. Through the Carsim/Simulink co-simulations and Hardware-in-the-Loop tests under several typical driving conditions, the superiorities of the stability recognition method and the coordinated control strategy proposed in this paper are verified. [ABSTRACT FROM AUTHOR]

This article focuses on the 3-D guidance strategy for a reusable launch vehicle (RLV) during terminal area energy management (TAEM) phase. Based on sliding-mode and shrinking-horizon techniques, the proposed scheme consists of trajectory generation and correction mechanisms, which can enhance the guidance precision and robustness against disturbances. The RLV guidance model, in the form of a set of highly nonlinear differential equations in the time domain, is recast as an altitude-domain model. By this means, the main characteristics of TAEM gliding motion are extracted. The altitude-domain model is thereby used for trajectory generation. A sliding surface and a guidance law are proposed. Hybrid TAEM constraints can be fully satisfied when the proposed guidance law drives the altitude-domain vehicle model to the designated altitude. Using the proposed guidance law as the input of the altitude-domain model, a constrained TAEM trajectory is generated, leading to TAEM guidance commands simultaneously. The commands are utilized to drive the time-domain model to the terminal target. In an attempt to compensate for model uncertainties and initial deviations, the guidance commands are modified periodically by the shrinking-horizon correction mechanism according to current states. Simulations on different scenarios are provided to demonstrate the performance of the proposed guidance strategy. [ABSTRACT FROM AUTHOR]

We have recently witnessed a number of new software-defined paradigms of VANET in what is referred to as software-defined vehicle networks (SDVN). In order to evaluate the performance of these new proposals and architectures, analytical and simulation models are needed. In this paper, we propose an analytical model based on ON-OFF queueing networks under exponential and general service time distributions. The model can be used to evaluate the performance of SDVNs and takes into account the effect of mobility such as, hand overs, node turning ON/OFF, node going temporary out of coverage, and intermittent connections. This mobility effect was modelled as a queueing station with exponentially random ON-OFF service times, where traffic arrives according to a Poisson random process during the exponentially random ON period and the service time is exponentially distributed. However, during the OFF period the service time is exponentially distributed but with lower rates. We studied the ON-OFF queueing behaviour extensively for both finite-capacity and infinite-capacity queues. Three hypothetical SDVN scenarios were considered, taking into account the effect of mobility and the large number of connected nodes. Results were cross-validated with those obtained by a simulation model. These tools will be valuable for researchers interested in getting quantitative answers for their SDVN architectures. [ABSTRACT FROM AUTHOR]

With the rapid development of intelligent transportation, automated traffic surveillance is considered as an important component. In the field of traffic surveillance, it is particularly important to achieve robust and real-time tracking of vehicles in complex scenes. In this paper, a robust real-time vehicle tracking model named NoisyOTNet is proposed, which formulates tracking as reinforcement learning with parameter space noise. In this formulation, the exploration ability of the model is enhanced to improve the robustness of tracking. Specifically, we develop a new implementation for noisy network based on deep deterministic policy gradients (DDPGs) with parameter noise, which can better cope with the tracking task and directly predict the tracking result. To improve the tracking accuracy in complex conditions, e.g. fast motion and large deformation, this paper presents an adaptive update strategy that can exploit the vehicle spatial-temporal information based on Upper Confidence Bound (UCB) algorithm by exploiting. Moreover, as for the recovery of the lost target, a relocation algorithm based on incremental learning is developed. The results of extensive experiments demonstrate that the proposed NoisyOTNet can effectively track vehicles in complex scenes and achieve competitive performance compared to the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Robust and adaptive vehicle state estimation and tracking algorithms have become a very important part within the autonomous driving field. The family of interacting multiple model (IMM) filters has shown to provide very effective and accurate state estimation in systems whose behavior patterns change significantly over time. The main reason for the improved performance of IMM filters compared to single model approaches is the mode mixing, which constantly aligns the different models. This paper proposes an innovative way for the mode mixing, when the state-vectors of the models are of different size. The proposed mixing approach consists of two parts: firstly mixing the common states and secondly weighting between original and mixed states based on the model probabilities. Results from artificial simulations and real world measurement setups are shown to demonstrate the validity of the approach. Compared to previously suggested solutions, the proposed approach is more general and the overall complexity of the mode mixing step is reduced, which is the main contribution of the presented paper. [ABSTRACT FROM AUTHOR]

Electric Vehicle (EV) demand modeling constitutes the cornerstone of studies aiming to facilitate the integration of EVs into the power system. The different characteristics of the EV demand (departure time, arrival time, and electric demand), as well as the correlation between thereof, render EV demand modeling a complex task. The Majority of previous methods, which were developed based on the Monte Carlo simulation, are unable to observe and preserve the correlation between EV demand characteristics; because, in these methods, the EV demand characteristics are generated separately in an unsupervised manner. This study proposes a novel semi-supervised EV demand modeling approach by mapping the different EV demand characteristics into a three-dimensional (3D) space as a 3D image. To effectively realize the 3D EV demand modeling, we have employed Generative Adversarial Networks (GANs) with a 3D convolutional structure to develop EV-GANS network—a GANs structure tailored to the needs of EV demand modeling in environments hosting high demand diversity such as EV charging stations. Numerical results confirmed the effectiveness of the proposed EV-GANS in estimating the trend of the actual EV demand on the test day with a small error margin compared to the existing benchmark generation-based methods (Monte Carlo and Copula). [ABSTRACT FROM AUTHOR]

A novel longitudinal, lateral, and vertical integrated stability control system is proposed for distributed electric vehicle in this paper. The integrated stability control system is based on model predictive control and takes advantage of its multi-objective optimization feature to takes manipulability, stability and comfort requirements into consideration at the same time, and realize the optimal longitudinal $/$ lateral $/$ vertical forces distribution with constraints overall considered. The advantage of the proposed integrated control system is that it can solve the control objective conflict of the cooperative motion control subsystems and the interference of the actuators compared to distributed control system. Finally, the simulation and comparison analysis are given in several different driving conditions under Carsim and Matlab $/$ Simulink environment. Results show that an improvement of over 80 $\%$ in the pitch angle and roll angle is realized compared to the simple control strategy. [ABSTRACT FROM AUTHOR]

As the capabilities for vehicular communications have become widespread, periodic beaconing is becoming fundamental to tracking neighbors. Specifically, a vehicle periodically broadcasts its kinematic data and receivers estimate the sender’s evolving position. To ensure safety, tracking neighbors via beaconing requires position errors and transmission delay to be small. To satisfy stringent requirements, previous proposals have employed multiple RF devices based on the unrealistic assumption that vehicles have the same type of RF devices. In reality, vehicles possess different RF devices (heterogeneous vehicle conditions). Satisfying the requirements under heterogeneous conditions is challenging because network connectivity is low and multi-hop transmissions to improve the connectivity aggravate network congestion. To address this challenge, we propose a novel scheme using a model-based trajectory prediction and multi-hop transmissions adaptively. To maintain accuracy of the model, each vehicle creates a model for predicting its own trajectory and distributes the model. For reliable multihop transmissions, our scheme employs periodic scan for translator and disconnected neighbors (PSTN) and probabilistic relay (PR). To our knowledge, this is the first to consider heterogeneous vehicle conditions for tracking neighbors via beaconing. Evaluation confirms that our scheme tracks neighbors more accurately than previous work. [ABSTRACT FROM AUTHOR]

This paper presents a novel estimation and identification approach for lateral vehicle dynamics. The algorithm leverages on a Linear Fraction Transform (LFT) reformulation of vehicle and tyre models, allowing for a simple and computationally efficient inclusion of complex and nonlinear dynamic models, like, for example, two-wheels, four-wheels or single-track as vehicle model, and Pacejka, brush or Fiala as tyre model. As a result, this technique can be easily adopted in the development of an online identification system, able to run on a standard embedded device, implementing a flexible identification procedure that can handle different driving conditions, up to the limits of handling, different vehicle modelling approaches, and different input measurements. Experimental results demonstrate the effectiveness of the proposal, either in a persistent excitation and in a non-persistent excitation scenario. [ABSTRACT FROM AUTHOR]

Message exchange among vehicles plays an important role in ensuring road safety. Emergency message dissemination is usually carried out by broadcasting. However, high vehicle density and mobility lead to challenges in message dissemination such as broadcasting storm and low probability of packet reception. This paper proposes a federated learning based blockchain-assisted message dissemination solution. Similar to the incentive-based Proof-of-Work consensus in blockchain, vehicles compete to become a relay node (miner) by processing the proposed Proof-of-Federated-Learning (PoFL) consensus which is embedded in the smart contract of blockchain. Both theoretical and practical analysis of the proposed solution are provided. Specifically, the proposed blockchain based federated learning results in more vehicles uploading their models in a given time, which can potentially lead to a more accurate model in less time as compared to the same solution without using blockchain. It also outperforms other blockchain approaches in reducing 65.2% of time delay in consensus, improving at least 8.2% message delivery rate and preserving privacy of neighbor vehicle more efficiently. The economic model to incentivize vehicles participating in federated learning and message dissemination is further analyzed using Stackelberg game. The analysis of asymptotic complexity proves PoFL as the most scalable solution compared to other consensus algorithms in vehicular networks. [ABSTRACT FROM AUTHOR]

Path planning is a critical part for improving the driving safety and driver comfort of autonomous vehicles (AVs), especially in complex maneuvering conditions. In addition, different drivers have different preferences for AVs, thus, how to provide personalized trajectories for different drivers is a vital issue for AVs. The collision-free path planning problem in conditions with large road curvatures is investigated in this paper, with the consideration of environmental safety constraints, drivers’ comfort, vehicle actuator constraints, etc. Firstly, a Driver-Vehicle-Road (DVR) system is established based on the combination of the kinematic vehicle model and the two-point visual preview driver model, such that the driver's individual handling characteristics can be considered in the controller. The kinematic vehicle model is modified to have the similar understeering characteristics with those of the nonlinear full car models, and then the proposed DVR system can satisfy different groups of drivers and cars. Secondly, for environmental constraints, a new artificial potential field (APF) method is proposed, which can form a banana-shaped 3-D dangerous imaginary mountain and a lane boundary cliff suitable for arbitrary curvature roads to generate a collision-free evasive path. Finally, the Linear-Time-Varying (LTV) model predictive control (MPC) method is adopted to design the path planner. The CarSim-Simulink joint simulation illustrates that with the proposed planner, the host vehicle is capable of avoiding obstacles with a safer and more comfortable maneuver on large curvature roads. And the proposed path planner can provide individually safe trajectories for different drivers with good maneuverability. [ABSTRACT FROM AUTHOR]

This article suggests a model-based controller for an all-wheel drive (AWD) vehicle using a novel vehicle dynamics model. Recently, an active type of AWD system that automatically controls the clutch engagement force of the transfer case has become popular. However, its performance has been limited by its rule-based controller, which necessitates the development of a model-based controller. Although the bicycle model has been most widely adopted for upper level controller of vehicle control systems, it is not appropriate for AWD controllers because the model does not have a direct relationship between the states and control inputs. In this sense, this study adopted a tire force-based full car model for the AWD controller. The clutch engagement force limit, which varies depending on the current engagement state, was calculated systematically by considering changeable dynamic behavior that is inherent in AWD vehicle and was used as a valid value for input limit in linear quadratic regulator (LQR) control algorithm. For the real-time application of the designed controller, only data obtained from the controller area network (CAN) of production vehicles were used. Then, the performance of the proposed controller was validated through vehicle experiments that included scenarios of both longitudinal and lateral movements. [ABSTRACT FROM AUTHOR]

Controller Area Network is the most frequently used in-vehicle communication system in the automotive industry today. The communication inside the Controller Area Network (CAN) bus is typically encoded using proprietary formats in order to prevent easy access to the information exchanged on the bus. However, it is still possible to decode this information through reverse engineering, performed either manually or via automated tools. Existing automated CAN bus reverse engineering methods are still time-consuming and require some manual effort, i.e., to inject diagnostic messages in order to trigger specific responses. In this paper, we propose CANMatch – a fully automated CAN bus reverse engineering framework that does not require any manual effort and significantly decreases the execution time by exploiting the reuse of CAN frames across different vehicle models. We evaluate the proposed solution on a dataset of CAN logs, or traces, related to 479 vehicles from 29 different automotive manufacturers, demonstrating its improved performance with respect to the state of the art. [ABSTRACT FROM AUTHOR]

In this work, a novel single-level adaptive trajectory planner and tracking controller is developed for off-road autonomous vehicles operating on deformable terrains. Trajectory planning and tracking algorithms often rely on a simplified vehicle model to predict future vehicle states based upon control inputs, hence requiring accurate modeling and parameterization. On off-road deformable terrains this is a challenging task due to unknown terrain parameters and the complex interactions at tire-terrain interfaces, which pose issues in continuous differentiability, operating conditions, and computational time. To address these difficulties, in this paper, a neural network deformable terrain terramechanics model and its implementation within a terrain adaptive model predictive control algorithm is presented to improve vehicle safety and performance through more accurate prediction of the plant response. It is shown in simulations that the neural network is able to predict the lateral tire forces accurately and efficiently compared to the Soil Contact Model as a state-of-the-art model and is able to yield accurate bicycle model predictions. It is demonstrated that the implementation of the neural network within model predictive control can outperform both a baseline Pacejka-based and a rapidly exploring random tree controller by improving performance and allowing for more severe maneuvers to be completed that otherwise lead to failure when terrain deformations are not explicitly taken into account. The improved performance achieved through estimating terrain parameters online in an adaptive controller is highlighted against the nonadaptive realization. Finally, it is shown the algorithm is conducive to real-time implementation. [ABSTRACT FROM AUTHOR]

In vehicle trajectory prediction, the difficulty in modeling the interaction relationship between vehicles lies in constructing the interaction structure between the vehicles in the traffic scene. Majority of existing models only focus on the interaction between the historical trajectory of the vehicle and the surrounding vehicles in the spatial domain, and do not pay attention to the interaction between the vehicle and the non-Euclidean correlation structure (graph structure) that exists in the environment. In order to overcome the deficiencies in the existing models, this paper proposes the Environment-Attention Network model (EA-Net) to obtain the full interactive information between the vehicle and its driving environment. In the proposed model, a new type of parallel structure consisting of Graph Attention network (GAT) and Convolutional social pooling containing Squeeze-and-Extraction mechanism (SE-CS), is constructed as the environmental feature extraction module and embedded in LSTM encoder-decoder. This structure solves the limitation of the dimension and structure influence when constructing the interaction model between the vehicle and the surrounding environment, making the extracted feature information comprehensive and effective. The prediction accuracy of the model with the RMSE loss function is tested on two public datasets– NGSIM and highD, and compared with several state-of-the-art trajectory prediction algorithm models. The results show that the prediction accuracy of the proposed Environment-Attention Network in the two datasets is more than 20% higher than that of the single structure model, which indicates that the proposed model proposed has superior performance and better adaptability to different traffic environments compared with the existing models. [ABSTRACT FROM AUTHOR]

Energy storage devices such as batteries and supercapacitors are arousing a growing interest in many areas of industrial applications since they allow improving the dynamic response and efficiency. In this sense, the availability of electric models of batteries and supercapacitors is necessary for the correct sizing, control, and energy management of power systems. In order to identify the parameters of the equivalent electric model, the identification current profile is often a pulse of charge or discharge. Subsequently, these models are validated with a current verification profile. However, the verification profile compared to the identification profile is more dynamic in amplitude and frequency, producing validation errors due to the identification signal's limited bandwidth. This paper proposes a novel Pulse – Pseudo Random Binary Sequence (Pulse – PRBS) identification current profile, with similar spectral characteristics to the verification profile, in order to improve the modeling accuracy. The parameters of a Lithium-ion battery and a supercapacitor are obtained with this new identification current profile and the Optimization Toolbox of Matlab, using the electrical models in Simulink or Simscape. Finally, the models are validated with the Urban Driving Cycle ECE-15 and the Hybrid Pulse Power Characterization (HPPC) for electric vehicle applications. [ABSTRACT FROM AUTHOR]

This article is motivated by an interesting recent result on antilock braking system (ABS) design using two-time scale (TTS)-based estimation. In this article, the basic method of TTS is extended to estimate not only the disturbances but also their derivatives, improving the accuracy of estimation by an order of magnitude. Further, the problem of ABS is extended to an eight-DOF vehicle model to include the effects of load transfer, steering dynamics, and actuator dynamics. Since analysis showed that the use of basic TTS to ABS design results in a poor transient in the estimation error dynamics, this article proposes a new formulation of the TTS method specifically for ABS. Two new ABS-specific sliding surfaces are proposed to compensate the actuator dynamics where the second sliding surface is proposed as an improvement over the first one. The stability, and convergence of estimation, and tracking errors are analyzed. The proposed methods are validated on MATLAB/Simulink, and CarSim platforms for a D-class sedan car. The results are also validated on an experimental ABS platform in the laboratory. [ABSTRACT FROM AUTHOR]

While bringing convenience to people, the growing number of vehicles already on roads causes inevitable traffic congestion. Traffic congestion can happen for both observable reasons, such as accidents, or occur naturally due to human driving style in high vehicles density scenarios. In the latter case, traffic congestion are referred to as phantom jams. To alleviate the traffic instability caused by phantom jams, bilateral control models have been proposed with the development of intelligent transportation system (ITS) in ideal situations. However, in road scenarios, some factors that severely affect the performance of existing models cannot be ignored, such as uncertainties of vehicle state measurements caused by on-board sensors, time delay caused by intervehicle communications and control systems of vehicles. In this paper, a novel predictable bilateral control model (PBCM) with local wireless communication, which consists of best estimation and state prediction, is proposed to accurately control the host vehicle in traffic flow to alleviate traffic instability. Compared with other models, theoretical analysis and simulation results show that more accurate control decisions could be obtained with the PBCM, which could distribute all vehicles more evenly, and our model could maintain higher velocities when disturbances occurred. The experimental results also reveal that our proposed model could more effectively suppress traffic instability (i.e., the improvement in control decision $a$ is more than 10%) and could restore traffic balance 20% faster than other models on average. [ABSTRACT FROM AUTHOR]

In this article, we propose two novel experimental design techniques for designing maximally informative experiments to estimate the parameters of nonlinear dynamical vehicle models. The two techniques include a batch design and a sequential design technique that seek to maximize the expected Shannon information gain of the parameter distribution using either an online or offline approach (respectively). We apply and compare the techniques in both simulation and real-world experiments with a wheeled vehicle. In our simulation experiments, both of our proposed designs provide superior Shannon information gains relative to an unoptimized benchmark technique. In our real-world experiments, our sequential design technique achieves superior expected Shannon information gains relative to our batch design technique and the benchmark technique. [ABSTRACT FROM AUTHOR]

Internet-of-Vehicles (IoV) is a novel technology to ensure safe and secure transportation by enabling smart vehicles to communicate and share sensitive information with each other. However, the realization of IoV in real-life depends on several factors, including the assurance of security from attackers and propagation of authentic, accurate and trusted information within the network. Further, the dissemination of compromised information must be detected and vehicle disseminating such malicious messages must be revoked from the network. To this end, trust can be integrated within the network to detect the trustworthiness of the received information. However, most of the trust models in the literature relies on evaluating node or data at the application layer. In this study, we propose a novel hybrid trust management scheme, namely, NOTRINO, which evaluates trustworthiness on the received information in two steps. First step evaluates trust on the node itself at transport layer, while second step computes trustworthiness of the data at application layer. This mechanism enables the vehicles to efficiently model and evaluate the trustworthiness on the received information. The performance and accuracy of NOTRINO is rigorously evaluated under various realistic trust evaluation criteria (including precision, recall, F-measure and trust). Furthermore, the efficiency of NOTRINO is evaluated in presence of malicious nodes and its performance is benchmarked against three hybrid trust models. Extensive simulations indicate that NOTRINO achieve over 75% trust level as compared to benchmarked trust models where trust level falls below 60% for a network with 35% malicious nodes. Similarly, 92% precision and 87% recall are achieved simultaneously with NOTRINO for the same network, comparing to benchmark trust models where precision and recall falls below 87% and 85% respectively. [ABSTRACT FROM AUTHOR]

This study discusses vehicle stability control based on explicit nonlinear model predictive control (NMPC) and investigates the influence of prediction model fidelity on controller performance. The explicit solutions are generated through an algorithm using multiparametric quadratic programming (mp-QP) approximations of the multiparametric nonlinear programming (mp-NLP) problems. Controllers with different prediction models are assessed through objective indicators in sine-with-dwell tests. The analysis considers the following prediction model features: 1) nonlinear lateral tire forces as functions of slip angles, which are essential for the operation of the stability controller at the limit of handling. Moreover, a simple nonlinear tire force model with saturation is shown to be an effective alternative to a more complex model based on a simplified version of the Magic Formula; 2) longitudinal and lateral load transfers, playing a crucial role in the accurate prediction of the lateral tire forces and their yaw moment contributions; 3) coupling between longitudinal and lateral tire forces, which has a significant influence on the front-to-rear distribution of the braking forces generated by the controller; and 4) nonlinear peak and stiffness factors of the tire model, with visible yet negligible effects on the results. [ABSTRACT FROM AUTHOR]

We describe the design and validation of a decision-making system in the guidance and control architecture for automated driving. The decision-making system determines the timing of transitions through a sequence of driving modes, such as lane following and stopping, for the vehicle to eventually arrive at the destination without colliding with obstacles, hence achieving safety and liveness. The decision-making system commands a transition to the next mode only when it is possible for an underlying motion planner to generate a feasible trajectory that reaches the target region of such next mode. Using forward and backward reachable sets based on a simplified dynamical model, the decision-making system determines the existence of a trajectory that reaches the target region, without actually computing it. Thus, the decision-making system achieves fast computation, resulting in reactivity to a varying environment and reduced computational burden. To handle the discrepancy between the dynamical model and the actual vehicle motion, we model it as a bounded disturbance set and guarantee robustness against it. We prove the safety and liveness of the decision-making system and validate it using small-scale car-like robots. [ABSTRACT FROM AUTHOR]

Safety at high speed is the most important performance of autonomous vehicle. Wherein, yaw stability control (YSC) is a key technology to improve vehicle safety. However, the high-speed vehicles are easy to be out of control in nonlinear region, which will result in serious accidents. This study proposes a model predictive control (MPC) method based on the Takagi-Sugeno (T-S) fuzzy model to realize YSC in the nonlinear region. Firstly, a T-S fuzzy model of vehicle is established based on a two-degree-of-freedom model to describe the vehicle dynamic characteristics in the nonlinear region. Then, a robust MPC is designed and its solution is given in the linear matrix inequalities (LMIs) form. Next, a sliding mode controller (SMC) is proposed to prevent wheel lock in the YSC process. Finally, the proposed strategy is verified in simulations and experiments. The results show that the proposed strategy has similar performance in vehicle stable region with linear MPC, and shows the ability to suppress the instability of the vehicle in the nonlinear region, and its computation burden is acceptable. [ABSTRACT FROM AUTHOR]

Accurate predictions of future pedestrian trajectory could prevent a considerable number of traffic injuries and improve pedestrian safety. It involves multiple sources of information and real-time interactions, e.g., vehicle speed and ego-motion, pedestrian intention and historical locations. Existing methods directly apply a simple concatenation operation to combine multiple cues while their dynamics over time are less studied. In this paper, we propose a novel Long Short-Term Memory (LSTM), namely, to incorporate multiple sources of information from pedestrians and vehicles adaptively. Different from LSTM, our considers mutual interactions and explores intrinsic relations among multiple cues. First, we introduce extra memory cells to improve the transferability of LSTMs in modeling future variations. These extra memory cells include a speed cell to explicitly model vehicle speed dynamics, an intention cell to dynamically analyze pedestrian crossing intentions and a correlation cell to exploit correlations among temporal frames. These three individual cells uncover the future movement of vehicles, pedestrians and global scenes. Second, we propose a gated shifting operation to learn the movement of pedestrians. The intention of crossing the road or not would significantly affect pedestrian’s spatial locations. To this end, global scene dynamics and pedestrian intention information are leveraged to model the spatial shifts. Third, we integrate the speed variations to the output gate and dynamically reweight the output channels via the scaling of vehicle speed. The movement of the vehicle would alter the scale of the predicted pedestrian bounding box: as the vehicle gets closer to the pedestrian, the bounding box is enlarging. Our rescaling process captures the relative movement and updates the size of pedestrian bounding boxes accordingly. Experiments conducted on three pedestrian trajectory forecasting benchmarks show that our achieves state-of-the-art performance. [ABSTRACT FROM AUTHOR]

The complex slip characteristics between the tracks and the terrain make it difficult to build an accurate model for intelligent tracked vehicles. Both the path planning and lateral control, however, are highly dependent on the accurate tracked vehicle model. To overcome the issue, a slip model based on the instantaneous centers of rotation (ICRs) of tracks and a dual-layer adaptive unscented Kalman filter (DAUKF) are used to estimate the ICR locations in real-time without requiring prior knowledge of terrain parameters. First, the historical trajectory information is used to estimate ICR locations by the upper-layer AUKF estimator preliminarily. The ICR locations estimated in the upper-layer and the current vehicle state are then imported into the model to predict the future trajectory which can be used to estimate the offset as compensation of preliminarily estimated ICR locations by the lower-layer AUKF estimator. The proposed DAUKF is verified by simulations on MATLAB/Simulink. In order to further verify the effectiveness and practicability of the algorithm, a large number of experiments under different terrains and road conditions are implemented on the electric-drive tracked vehicle. The simulation and experimental results illustrate that the proposed DAUKF can estimate the ICR locations efficiently and accurately, which can improve the accuracy of the tracked vehicle model compared with those of extended Kalman filter (EKF), UKF, and AUKF. [ABSTRACT FROM AUTHOR]

This paper proposes a model predictive current control (MPCC) method of six-phase permanent magnet synchronous hub motor (PMSHM) using auxiliary voltage vectors. First of all, 26 voltage vectors can be obtained from the switching states of six inverters, and 24 auxiliary voltage vectors can be generated by combining them. Second, four virtual voltage vectors can be found through the required locations, and the two vectors with the lowest cost will be selected by cost function. Then, additional current error correction module can better track the current. Finally, the proposed method is compared with the conventional MPCC method and another method with virtual voltage vectors. The innovation of this paper is that the recombination of vectors can better track and predict the current. A current error compensation module is added to reduce the error between the actual current and the predicted current. And the experimental results show that the proposed method has better performance such as smaller torque ripple and current THD both in steady and dynamic states. Finally, the strategy proposed in this paper was applied to the vehicle model through HIL test platform. The possibility of applying the strategy proposed in this paper to pure electric vehicles was verified under New European Driving Cycle (NEDC). [ABSTRACT FROM AUTHOR]

This article proposes a low-computation framework for on-board calculation of energy-optimal cruising speed, and uses the framework to investigate the critical parameters for energy-optimal cruising speed determination. The main features of the proposed framework are as follows: first, inclusion of all internal and external vehicle losses, particularly the accessory loads, which have a large impact on optimal cruising speed and have not been investigated in other works, second, use of accurate motor/inverter loss look-up-tables rather than approximated polynomial models, and, third, no requirement for connected data or knowledge of the future route, as the optimal cruising speed is calculated precisely for the current state of the vehicle and surroundings. To validate the framework, three electric vehicle models are developed in MATLAB/Simulink and tuned to match real-world data for a Chevrolet Bolt, Chevrolet Spark, and Tesla Model S. Furthermore, an algorithm is developed to determine near-optimal speed transitions when the optimal cruising speed changes. [ABSTRACT FROM AUTHOR]

Accurate physical modeling of vehicle dynamics requires extensive a priori knowledge of the studied vehicle. In contrast, data-driven modeling approaches require only a set of data that are a good account of the vehicle’s driving envelope. In this brief, we compare, for the first time, the prediction capabilities of both approaches applied to a large-scale real-world driving data set. The data set contains several cornering maneuvers, acceleration, and deceleration stages and was collected over public roads. Linear and nonlinear physical models were identified through nonlinear optimization of their unknown parameters. Closed-form subspace identification methods were used to initialize the estimate of a linear state-space model, and the initialization was then refined through nonlinear optimization. The optimized models were validated against 59 km of independent driving data. The model fits, in the longitudinal velocity, were 68.9% versus 80.2% for the nonlinear physical model and linear data-driven (second-order) model, respectively, and, in the yaw rate, 43.0% versus 63.5%. These results show that, for this vehicle, a simple linear data-driven model outperformed both linear and nonlinear physical models under real-world driving conditions. This has important implications for control design approaches in autonomous vehicles. [ABSTRACT FROM AUTHOR]

In this article, we aim to investigate the path following control problem for four-wheel-independent-drive electric vehicles with consideration of modeling errors and complex driving scenarios. In order to guarantee operational safety and robustness, a supertwisting second-order sliding mode (SOSM) control strategy is employed to suppress the heavy chattering issue existing in the traditional sliding mode control (SMC). To degrade the lumped disturbance, a nonlinear disturbance observer (NDOB) technique is proposed to estimate the disturbance and compensate the modeling error. Then, a composite control strategy, which combines the SMC algorithm and NDOB technique is proposed based on stability analysis of the corresponding closed-loop system. Finally, hardware-in-the-loop experimental results show that the supertwisting SOSM control algorithm with the NDOB technique can achieve better tracking performance and suppress the chattering phenomenon. [ABSTRACT FROM AUTHOR]

Vehicle Re-Identification (Re-ID) is a challenging vision task mainly because the appearance of a vehicle varies dramatically under different viewpoints. Moreover, different vehicles with the same model and color commonly show similar appearance, thus are hard to be distinguished. To alleviate negative effects of viewpoint variance, we design a multi-view branch network where each branch learns a viewpoint-specific feature without parameter sharing. Being able to focus on a limited range of viewpoints, this viewpoint-specific feature performs substantially better than the general feature learned by an uniform network. To further differentiate visually similar vehicles, we strengthen the discriminative power on their subtle local differences by introducing a spatial attention model into each feature learning branch. The multi-view feature learning and spatial attention learning compose our neural network architecture, which is trained end to end with the softmax loss and triplet loss, respectively. We evaluate our methods on two large vehicle Re-ID datasets, i.e., VehicleID and VeRi-776, respectively. Extensive experiments show that our methods achieve promising performance. For example, we achieve mAP accuracy of 76.78% and 72.53% on VehicleID and VeRi-776 dataset respectively, substantially better than current state-of-the art. [ABSTRACT FROM AUTHOR]

This paper investigates the robust adaptive neural control of nonminimum phase hypersonic flight vehicle using composite learning. To overcome the nonminimum phase behavior, the output redefinition is employed and the attitude subsystem is transformed to the internal subsystem and the input–output subsystem. For the input–output subsystem, the adaptive neural control works together with the robust control to follow the reference command of pitch angle derived from the internal subsystem. Furthermore, the sliding mode control is constructed in a similar way. For the update of the neural weights, the composite learning is constructed using the prediction error. The stability of the closed-loop system is analyzed via the Lyapunov approach and the ultimately uniform boundedness of the tracking errors can be guaranteed. The effectiveness of the methodology is illustrated by the simulation results. [ABSTRACT FROM AUTHOR]

In a typical traffic scenario, autonomous vehicles are required to share the road with other road participants, e.g., human driven vehicles, pedestrians, etc. To successfully navigate the traffic, an adaptive robust level-k framework can be used by the autonomous agents to categorize the agents based on their depth of strategic thought and act accordingly. However, mismatch between the vehicle dynamics and its predictions, and improper classification of the agents can lead to undesirable behavior or collision. Robust approaches can handle the uncertainties, however, might result in a conservative behavior of the autonomous vehicle. This paper proposes an adaptive robust decision making strategy for autonomous vehicles to handle model mismatches in the prediction model while utilizing the confidence of the driver behavior to obtain less conservative actions. The effectiveness of the proposed approach is validated for a lane change maneuver in a highway driving scenario. [ABSTRACT FROM AUTHOR]

In this article, the formation control of unmanned surface vehicles (USVs) is addressed considering actuator saturation and unknown nonlinear items. The algorithm can be divided into two parts, steering the leader USV to trace along the desired path and steering the follower USV to follow the leader in the desired formation. In the proposed formation control framework, a virtual USV is first constructed so that the leader USV can be guided to the desired path. To solve the input constraint problem, an auxiliary is introduced, and the adaptive fuzzy method is used to estimate unknown nonlinear items in the USV. To maintain the desired formation, the desired velocities of follower USVs are deduced using geometry and Lyapunov stability theories; the stability of the closed-loop system is also proved. Finally, the effectiveness of the proposed approach is demonstrated by the simulation and experimental results. [ABSTRACT FROM AUTHOR]

Active safety systems are of significant importance for autonomous vehicles operating in safety-critical situations like an obstacle-avoidance manoeuvre with high vehicle speed or poor road condition. However, a conventional electronic stability control system, may not always yield desired path following and yaw stability performance in such circumstances merely through brake intervention. This paper pursues a detailed investigation on utilising model predictive control (MPC) and torque vectoring for path following and yaw stability control of over-actuated autonomous electric vehicles (AEVs) in dangerous double lane change manoeuvre. The control problem of the AEV is formulated based on MPC by utilising active front steering and torque vectoring, and constraints are imposed explicitly on yaw rate and sideslip angle to ensure yaw stability. Four MPC-based controllers are designed based on double-track vehicle models. Specifically, they include two one-level controllers, i.e. one with torque vectoring and one with equal torque allocation, and two two-level controllers, i.e. one with optimisation-based torque allocation and one with rule-based allocation. These controllers are assessed extensively, with respect to passing velocity, tracking accuracy, tyre utilisation and robustness. The effect of horizon length on the control performance and computational efficiency is also investigated. [ABSTRACT FROM AUTHOR]