This paper proposes a novel fault-tolerant control (FTC) reconfiguration strategy for autonomous vehicles using Model Predictive Control (MPC) based on a Linear Parameter Varying (LPV) model to address steering faults. The proposed approach compensates for the lack of redundant steering actuators by using force differences between the left and right sides of the vehicle to generate corrective yaw moments. By integrating both lateral and longitudinal dynamics, the MPC optimally allocates actuator efforts based on fault severity and desired speed. Simulation results validate the effectiveness of the proposed strategy in maintaining vehicle stability and performance under various fault scenarios, including complete steering failure, thereby ensuring safe autonomous operation.

This paper proposes a method for simultaneous estimation of both the driver torque and the sideslip angle within the context of human-machine shared driving control for autonomous ground vehicles. To this end, the driver torque is considered as an unknown input (UI) and the sideslip angle is an unmeasured state of the vehicle dynamics system. For simultaneous estimation purpose, a decoupling-based technique is leveraged to design an unknown input observer (UIO). The UIO design goal is to decouple the effect of the unknown driver torque while minimizing the influence of the modeling uncertainties, considered as unknown exogenous disturbances, from the lateral tires forces and the steering system. Linear parameter-varying (LPV) framework is used to deal with the time-varying nature of the vehicle longitudinal speed. Based on Lyapunov stability theory, we derive sufficient conditions, expressed in terms of linear matrix inequality (LMI) constraints, for LPV unknown input observer design. The simultaneous vehicle estimation is reformulated as a convex optimization problem, where the modeling uncertainty influence can be minimized via the$\ell _{\infty} -$gain performance. Hardware-in-the-loop (HiL) tests are performed with the SHERPA dynamic simulator and a human driver to show the effectiveness of the proposed UIO-based estimation method, especially within the cooperative driving control framework. Note to Practitioners—We present a method to jointly estimate the driver torque and the sideslip angle in the context of human-machine shared driving. To this end, we consider the driver torque as an unknown input and treat the sideslip angle as an unmeasured state of the vehicle dynamics system. The core of our method lies in the application of a decoupling-based technique to design an unknown input observer. The primary objective of this UIO is to effectively decouple the influence of the unknown driver torque while mitigating the impact of modeling uncertainties, considered as unknown exogenous disturbances, on the lateral tire forces and the steering system. Using an LPV framework has allowed the time-varying nature of the vehicle longitudinal velocity to be effectively addressed. Via Lyapunov stability theory, we have established sufficient conditions, expressed in terms of LMI constraints, for the design of the LPV unknown input observer. The proposed simultaneous vehicle estimation method has been reformulated as a convex optimization problem, allowing to minimize the influence of modeling uncertainties. To show the effectiveness of the proposed UIO-based estimation method, we have conducted extensive HiL tests using the SHERPA dynamic simulator with a human driver. The real-time experiments demonstrate the effectiveness of the proposed method, especially with respect to related estimation results in the literature, within the cooperative driving control framework. The proposed LPV estimation method contributes to the advancement of the field of autonomous ground vehicles by providing practitioners with a robust tool for joint estimation of essential variables critical for effective vehicle control and safety in the context of human-machine cooperative driving.

This study presents an adaptive fault-tolerant approach for the attitude control of a quadrotor unmanned aerial vehicle (UAV) using the backstepping technique and fuzzy logic estimation. The proposed controller addresses sensor and actuator faults with a comprehensive model that includes multiplicative and additive faults. To handle the inherent nonlinearities and disturbances of the system, a modified backstepping controller is employed, combined with an adaptive Fuzzy Logic System (FLS) for estimating unknown nonlinear functions. The adaptive fuzzy fault-tolerant control strategy is formulated to ensure practical fixed-time stability of the closed-loop system even when faults appear. The effectiveness of the proposed approach is validated through simulations, demonstrating its robustness against disturbances and its capability to maintain stable UAV attitude control despite sensor and actuator faults.

International audience

This paper proposes a fault tolerant highway insertion management approach for Automative Driving (AD) system. The principle is as follows: the AD system in the ego vehicle (vehicle in the main lane) proposes an insertion decision to the merging vehicle based on the comparison between: i) the interdistance between the ego-vehicle itself and the merging vehicle and ii) a decision threshold. This decision threshold is computed by the ego-vehicle from the speed of the ego-vehicle, the speed of the merging vehicle communicated by the merging vehicle to the ego-vehicle and other parameters (such as safety margin, vehicles lengths,…). Since the insertion decision threshold depends on the speed measurements, adding a diagnostic layer to the merging vehicle speed estimation is crucial to obtain a safe and accurate insertion decision. With the merging vehicle's sensors measurements (GPS, wheel encoder, and speed sensor) sent to the ego-vehicle, the ego-vehicle can compute three estimates of the merging vehicle's speed: 1/ one is directly measured by a speed sensor, 2/ the other two are computed using Extended Kalman Filters (EKF) and data from respectively the GPS and the wheels encoders. Depending on the estimate of the current dynamic driving maneuver, the prediction model used in the EKFs switchs from a simple constant acceleration (CA) evolution model to a maneuvering model integrating the jerk estimation. Finally, based on the three redundant velocity estimates, residuals sensitive to sensor fault are generated. From the signature matrix, faulty measurements (if any) can be isolated and excluded of the estimation process.

The work described in this paper proposes a new dynamic conflict attenuation strategy in driving shared control for intelligent vehicles lane keeping systems (LKS). This strategy takes into account the activity and availability of the driver as well as the external risk and conflict between the driver and the control system in order to manage and adapt the level of assistance in real time. The design of an adaptive shared controller is based on a dynamic multi-objective cost function that changes according to the level of assistance. Based on Lyapunov stability arguments, the global asymptotical stability of the closed-loop control system with the adaptive cost function and the variation in vehicle speed is proven and an LMI optimization is used to formulate the control design. The simulation results, conducted with the SHERPA dynamic car simulator under real-world driving situations, for different scenarios show the importance of adapting the controller in real time in order to decrease the conflict between the driver and the lane keeping system and to ensure the safety of the vehicle as well as to increase the confidence and acceptability of the driver.

Haptic shared control has proven to be an effective method to assist a driver in controlling a vehicle. This method is now being considered for use in developing strategies for smooth transitions between manual and autonomous driving modes. This article has two objectives. First, it proposes to adapt an existing haptic shared control strategy to achieve transitions between manual and autonomous modes and to evaluate this approach with real drivers on a driving simulator. Second, it proposes to evaluate four different transition profiles in an obstacle-avoidance context. The first profile is a gradual transition from the autonomous mode to shared control mode, followed by another transition from the shared control mode to autonomous mode once the obstacle is passed. The second is a gradual transition from autonomous mode to manual mode. The third is a binary transition from autonomous mode to manual mode. Finally, in the fourth condition, the driver overrides the autonomous mode. These transition profiles were evaluated in curves and straight lines on a driving simulator. The results first validated the use of the haptic shared control strategy to execute transitions between manual and autonomous modes. The distribution of the torques delivered by the automation system and the driver corresponded to the progression of the expected sharing level. Second, the gradual transitions showed advantages over binary transitions and the override of the autonomous mode, both in terms of steering performance and subjective evaluation.

International audience

Understanding and predicting the behavior of the powered wheelchair users play an important role in developing the driver assistance systems for the intelligent wheelchair. This article presents a model of wheelchair users, from the point of view of control engineering, based on the interaction between the human hand-joystick and the lumped-parameter model of the human hand muscle system. The interaction between hand and joystick is represented by a robot arm with a four-bar closed-chain mechanism. This mechanism is operated by four musculotendon units based on Hill's muscle model. This configuration allows simulating whole motions of the joystick as well as the evolutions of muscle activities during maneuvering the powered wheelchair. The advantage of this model is the integration of the biomechanical parameters of the muscular system of the human hand, which is very useful for simulating the defects related to the degree of physical impairment of wheelchair users. To validate the proposed model, we compare the simulation outputs of the user model with the previously published experimental results in the framework of the Wheelchair Skills Test (v. 4.1). The simulation results show that the proposed model can reflect the difficulties at the biomechanical level of users in driving the wheelchair.

This paper proposes a novel approach for designing an assistance system via haptic joystick force feedback on an electric wheelchair using model-based shared control approach. Assistance system supports wheelchair users through haptic force at joystick which can make the user-wheelchair interaction becomes more intuitive. Firstly Tagaki-Sugeno fuzzy model is used to build an augmented model of user-wheelchair system to deal with non-linear nature of system. Unknown input observers are developed to estimate wheelchair position errors which are considered as user intention based on joystick motions. Fuzzy logic optimal controller is synthesized by Linear Matrix Inequalities method to provide assistance haptic forces feedback to user via joystick. Simulation and experimental results show that this assistance system can predict the desired motion of user and reduce user hand force thanks to haptic force at joystick.

This paper presents a methodology for the systematic synthesis of haptic shared control (HSC) of a car. This HSC design is based on a two-part architecture. The first part is a trajectory generator that provides a reference trajectory to the second part, which is a static output feedback. In this paper, the haptic shared control is used as an lane keeping assist system (LKA); hence, the reference trajectory is chosen to fulfill this function. The main contribution of this article is related to the combination of the H2/H∞feedback synthesis. This, involves an H2criterion quantifying the sharing level and quality as an objective function, and H2/H∞constraints for lane-keeping performance, driver comfort and robustness. The control design relies on a driver cybernetic model, which decreases conflicts between the assistance and the driver. A systematic method to tune criterion and constraints is described, enabling the attainment of desired lane-following and shared-control performance. The proposed methodology facilitates the design of lateral assistance, ensuring stability and guaranteed performance regardless of the prescribed level of sharing between the actions of the driver and the automaton. The shared control between human driver and automation for the lane keeping task over the Satory test track with the sharing level adaptation is then shown for the validation of the proposed architecture. This work introduces perspectives on smooth transitioning between manual and autonomous driving modes.

This paper provides a new solution for path following control of autonomous ground vehicles. H2control problem is considered to attenuate the effect of the road curvature disturbance. To this end, we formulate a standard model from the road-vehicle dynamics, the a priori knowledge on the road curvature, and the path following specifications. This standard model is then represented in a Takagi-Sugeno fuzzy form to deal with the time-varying nature of the vehicle speed. Based on a static output feedback scheme, the proposed method allows avoiding expensive vehicle sensors while keeping the simplest control structure for real-time implementation. The concept of V-stability is exploited using Lyapunov stability arguments to improve the transient behaviors of the closed-loop vehicle system. In particular, the physical upper and lower bounds of the vehicle acceleration are explicitly considered in the design procedure via a parameter-dependent Lyapunov function to reduce drastically the design conservatism. The proposed H2design conditions are expressed in terms of linear matrix inequalities (LMIs) with a single line search parameter. The effectiveness of the new path following control method is clearly demonstrated with both theoretical illustrations and hardware experiments under realworld driving situations.

This paper describes a real-time multi-objective predictive control strategy for the wheelchair ergometer platform, which allows simulating the manual wheelchair (MWC) propulsion in virtual reality (VR). A nonlinear least-square method is used to build the predictive wheelchair ergometer model, where the trust-region-reflective algorithm is adopted. A detailed study of the wheelchair dynamics, and its performance on the rollers, and the frictions are presented. The main contribution is the implementation of a wheelchair ergometer model into a proposed controller in order to provide realistic navigation within the VR world by accurately detecting and tracking the rotation of the driving wheels. In particular, the proposed strategy has two objectives: 1) to generate force feedback (haptics) during the push phase (hand-to-rim contact); 2) assistance control to make the wheelchair movement more realistic in the immersive virtual environment during the recovery phase (hand-to-rim no contact). Various numerical examples emphasize the flexibility of the approach and experimental results confirm that the proposed controller successfully tracks the speed reference. It is also shown that the real-time tracking ability of the explicit MPC (eMPC) controller is superior to that of the PI controller.

A manual wheelchair (MWC) is an indispensable support for a person with reduced mobility. The researches on improving performance of MWC, therefore, should be validated by a reliable simulation platform, which allows simulating the MWC operation in virtual reality (VR), before being tested on real conditions. This research aims to provide a fully-constituted dynamical model of MWC which will be used in simulator platform to simulate the real movements of wheelchair and to generate the force feedback (haptics) during the push phase (hand-to-rim contact). With focus on the performance of wheelchair at the low-speed region, this model comprised the tire-road contact forces and the influence of the front free caster wheels. Validation experiments simulating several typical daily maneuvers of the manual wheelchair were conducted. Compared to previous research, the new model presented in this paper overcomes the problem of overestimating the the yaw rate.

International audience

This paper is concerned with a new control method which can share the control authority between a human driver and a lane keeping assist system. Based on the concept of multiple controller switching, this shared control method is composed of two levels: operational and tactical. At the operational level, two local optimal-based controllers are designed to satisfy their own predefined control goals. At the tactical level, a supervisor is designed to orchestrate a smooth control authority transition between two local controllers. The closed-loop properties of the human-in-the-loop vehicle system are guaranteed via Lyapunov stability arguments. In particular, the design of both local controllers is recast as a convex optimization problem, easily solved with numerical solvers. The effectiveness of the proposed shared control method is experimentally validated with a human driver and a dynamic driving simulator.

This paper describes the design of a new haptic shared steering control framework for automated driving systems. In this framework, the shared control problem is formulated as a constrained optimization problem which is solved online by a model predictive controller. Without driver's intervention, the system assumes automatic lane-keeping control. When the driver takes over control, by adapting the weight on the stage cost and implementing dynamic constraints, the framework ensures seamless control transfer from the system to the driver while conveying potential hazards through haptic feedback. Simulation results are presented to demonstrate the ability of this framework to handle control allocation and hazard warning.

Driver's arm viscoelastic properties during steering vehicle maneuver are very important information to understand how the driver regulates his torque and they could be used to design a new driver assistance system. This correspondence paper investigates the driver's arm viscoelastic properties during steering vehicle maneuver. First, a time-varying method for estimating driver's arm viscoelastic properties during driving the vehicle is proposed. Then, this method is implemented experimentally in order to examine the driver's arm viscoelastic properties during driving the vehicle. The estimation method does not require a specific perturbation torque to be applied during driving and could be applied online. The method is based on a time-varying model of the human driver's arm coupled with the electric power-assisted steering system model. Experiments were carried out with five healthy subjects who drove on a hardware-in-the-loop simulator without assistance and with assistance.

This article presents an automated driving system that ensures cooperation with the driver. The system architecture is structured in hierarchical levels to allow suitable interaction with the driver on multiple driving levels. A multi-level cooperative interaction concept is developed to continuously share control and dynamically manage interferences and decision authority between the driver and the system according to the situation. The system extends the lane keeping function with an active lane change assistance function. The necessary components for the multi-level cooperation concept are presented and experimental results show a good and intuitive interaction for active lane change assistance.

This paper describes an adaptive vehicle longitudinal trajectory prediction method for automated highway driving applications. A major strength of this method is that it can cope with highly dynamic situations in which the constant acceleration (CA) assumption cannot guarantee long term prediction accuracy. In this method, a quintic polynomial is used to model the longitudinal dynamics of a vehicle that is maneuvering. The decision to switch to it from the CA model is formulated as a maneuver detection problem. A maneuver is detected through monitoring measurement innovations of a Kalman filter that tracks target longitudinal states. The longitudinal jerk, as a dynamic characteristic of a maneuver is also estimated from measurement innovations. Finally the estimated jerk and context information are incorporated into the quintic polynomial model. The overall approach was tested on recorded human driving data from a simulator in a dynamic highway merging scenario. The results show the proposed method has higher prediction accuracy than the CA based method in such a dynamic scenario.

This paper is devoted to the development of a new saturated control law for constrained Takagi-Sugeno fuzzy systems. These systems are subject to both control input and state constraints and also persistent disturbances bounded in amplitude. The design procedure is formulated through linear matrix inequalities (LMIs) form which can be solved by means of convex optimization techniques. Based on the concept of robust invariant set in non-quadratic Lyapunov control framework, the proposed method provides a characterization of the closed-loop domain of attraction. Numerical example is given to demonstrate the interests of the proposed methodology.

This paper is devoted to the development of a shared lateral control strategy for a Driver Steering Assist System (DSAS) that can share the authority with the driver. Up to now, this control issue is still an open research subject in automotive industry due to the complex interactions according to different driving situations between the Human (driver) and the Machine (DSAS). In this work, such interactions are handled by introducing into the vehicle system a fictive time-varying term representing the driver activity. In this way, the actions of the DSAS are computed in function of the driver behaviors (actions and intentions). Using Takagi-Sugeno control technique in the framework of Lyapunov stability theorem, the designed controller is able to handle a large range of variation of vehicle longitudinal speed. Moreover, the proposed controller requires only measured output signals for the design procedure and implementation. The effectiveness of the proposed method is demonstrated with different driving scenarios.

This paper describes the design of a shared control framework for vehicle longitudinal control applied in Automated Driving (AD) systems. Due to the complexity of automation design, highway merging scenario is firstly selected as the use case for system's functional design. Besides the distance control functionality, the framework is capable of managing merging situation thanks to the proposed situation assessment function. In order to provide the driver a mean to override the operating automated system, shared control logics are designed for accelerator and brake pedal respectively. In case of driver's override on accelerator, hap tic feedback is rendered aiming to communicate the system's activity and situation's criticality. Preliminary simulation results are presented to demonstrate the design concepts of the framework.

This paper presents on-line method to estimate the driver's arm viscoelastic properties during vehicle steering maneuver. A model of driver's arm impedance was coupled to the Electric Power Assisted Steering System (EPAS) model. The inputs-output of the impedance model are estimated and then used in the implementation of Exponentially Weighted Recursive Least Squares (EWRLS) algorithm to estimate the model properties. The validity of this approach is demonstrated using human and Hardware-In-the-Loop (HIL) testing.

The paper deals with the design of lateral shared vehicle control taking into account the interaction between the driver and the assistance system. The shared control system is designed in such a way to ensure a good transfer of the control authority without generating negative interference. For that a driver model that allows making valid predictions on the driver behaviour is integrated in the design process of the controller. In order to avoid complex conflict situations such as during lane change maneuver, a decision making algorithm for the control authority shifting is also proposed and implemented. Experimental results provided in the paper, using interactive simulator, show the effectiveness of the approach to ensure shared lateral vehicle control.

This paper describes a concept of lane keeping assistance system based on the fuzzy Takagi-Sugeno (T-S) optimal controller. Nonlinearities due to longitudinal velocity variation in the lateral dynamics are considered to give a T-S representation and Linear Matrix Inequalities (LMI) are used to reach the appropriate T-S optimal controller that insure the global stability of the closed loop system. In conclusion the effectiveness of the proposed approach is illustrated in Matlab/Simulink and experimental tests carried out on the ”SHERPA” (Simulateur Hybride d'Etude et de Recherche de PSA pour l'Automobile) simulator.

The work described in this paper is part of a larger research program named Partage whose goal is to appropriately share lateral control between the driver and an electronic copilot or assistance. The present study is looking for a cybernetic model of a driver steering road vehicle, which makes the sensorimotor dynamics explicit. The proposed model takes into account both visual and kinesthetic perception, and includes compensatory and anticipatory processes. As such, it extends previous works. Starting from simulated car trajectories, the unknown parameters of the model are identified using the grey box identification concept. The driver-vehicle system is then analyzed with respect to the parameter sensitivity.