Tom E. Pilutti

Identification of driver state for lane-keeping tasks

Identification of driver state is a desirable element of many proposed vehicle active safety systems (e.g., collision detection and avoidance, automated highway, and road departure warning systems). In the paper, driver state assessment is considered in the context of a road departure warning and intervention system. A system identification approach, using vehicle lateral position as the input and steering wheel position as the output, is used to develop a model and to update its parameters during driving. Preliminary driving simulator results indicate that changes in the bandwidth and/or parameters of such a model may be useful indicators of driver fatigue. The approach is then applied to data from 12 2-h highway driving runs conducted in a full-vehicle driving simulator. The identified model parameters (/spl zeta/ /spl omega//sub n/, and DC gain) do not exhibit the trends expected as lane keeping performance deteriorates, despite having acceptably white residuals. As an alternative, model residuals are compared in a process monitoring approach using a model fit to an early portion of the 2-h driver run. Model residuals show the expected trends and have potential in serving as the basis for a driver state monitor.

An optimal-control-based framework for trajectory planning, threat assessment, and semi-autonomous control of passenger vehicles in hazard avoidance scenarios

This paper formulates the vehicle navigation task as a constrained optimal control problem with constraints bounding a traversable region of the environment. A model predictive controller iteratively plans an optimal vehicle trajectory through the constrained corridor and uses this trajectory to establish the minimum threat posed to the vehicle given its current state and driver inputs. Based on this threat assessment, the level of controller intervention required to prevent departure from the traversable corridor is calculated and driver/controller inputs are scaled accordingly. Simulated and experimental results are presented to demonstrate multiple threat metrics and configurable intervention laws.

Vehicle steering intervention through differential braking

This paper examines the usefulness of a brake steer system (BSS) which uses differential brake forces for steering intervention in the context of intelligent vehicle highway systems. The resulting moment on the vehicle affects yaw rate and lateral position, thereby providing a limited steering function. The steering function achieved through BSS can then be used to control lateral position in an unintended road departure system. Models for the vehicle and the brake system are presented. A state feedback regulator and PID controller are developed to explore BSS feasibility and capability. Computer simulation results are included.

Vehicle Steering Intervention Through Differential Braking

This paper examines the usefulness of a brake steer system (BSS) which uses differential brake forces for steering intervention in the context of Intelligent Transportation Systems. The resulting moment on the vehicle affects yaw rate and lateral position, thereby providing a limited steering function. The steering function achieved through BSS can then be used to control lateral position in an unintended road departure system. Control design models for the vehicle and the brake system are presented. A state feedback regulator and PID controller are developed to explore BSS feasibility and capability. Computer simulation results, using a nonlinear seven degree-of-freedom vehicle model are included, and show the feasibility and limitations of BSS.

On-line identification of driver state for lane-keeping tasks

Online identification of driver state is a desirable element of many proposed active safety systems (e.g. collision detection and avoidance, automated highway and road departure warning systems). Here we consider driver state assessment in the context of a road departure warning and intervention system. A system identification approach, using vehicle lateral position as the input and steering wheel position as the output, is used to develop a model and to continually update its parameters during driving. Driving simulator results indicate that changes in the bandwidth and/or parameters of such a model may be useful indicators of driver fatigue.

Sampling-Based Threat Assessment Algorithms for Intersection Collisions Involving Errant Drivers

Abstract This paper considers the decision-making problem for a vehicle crossing a road intersection in the presence of other, potentially errant, drivers. This problem is considered in a game-theoretic framework, where the errant drivers are assumed to be capable of causing intentional collisions. Our approach is to simulate the possible behaviors of errant drivers using RRT-Reach, a modified application of rapidly-exploring random trees. A novelty in RRT-Reach is the use of a dual exploration-pursuit mode, which allows for efficient approximation of the errant reachability set for some fixed time horizon. Through simulation and experimental results with a small autonomous vehicle, we demonstrate that this threat assessment algorithm can be used in real-time to minimize the risk of collision.

Design and Development of an Optimal-Control-Based Framework for Trajectory Planning, Threat Assessment, and Semi-autonomous Control of Passenger Vehicles in Hazard Avoidance Scenarios

This paper describes the design of an optimal-control-based active safety framework that performs trajectory planning, threat assessment, and semi-autonomous control of passenger vehicles in hazard avoidance scenarios. This framework allows for multiple actuation modes, diverse trajectory-planning objectives, and varying levels of autonomy. A model predictive controller iteratively plans a best-case vehicle trajectory through a navigable corridor as a constrained optimal control problem. The framework then uses this trajectory to assess the threat posed to the vehicle and intervenes in proportion to this threat. This approach minimizes controller intervention while ensuring that the vehicle does not depart from a navigable corridor of travel. Simulation and experimental results are presented here to demonstrate the framework’s ability to incorporate configurable intervention laws while sharing control with a human driver.

Semi-autonomous Avoidance of Moving Hazards for Passenger Vehicles

This paper presents a method for semi-autonomous hazard avoidance in the presence of unknown moving obstacles and unpredictable driver inputs. This method iteratively predicts the motion and anticipated intersection of the host vehicle with both static and dynamic hazards and excludes projected collision states from a traversable corridor. A model predictive controller iteratively replans a stability-optimal trajectory through the navigable region of the environment while a threat assessor and semi-autonomous control law modulate driver and controller inputs to maintain stability, preserve controllability, and ensure safe hazard avoidance. The efficacy of this approach is demonstrated through both simulated and experimental results using a semi-autonomously controlled Jaguar S-Type.Copyright

Experimental Study of an Optimal-Control- Based Framework for Trajectory Planning, Threat Assessment, and Semi-Autonomous Control of Passenger Vehicles in Hazard Avoidance Scenarios

This paper describes the design of an optimal-control-based active safety framework that performs trajectory planning, threat assessment, and semiautonomous control of passenger vehicles in hazard avoidance scenarios. The vehicle navigation problem is formulated as a constrained optimal control problem with constraints bounding a navigable region of the road surface. A model predictive controller iteratively plans an optimal vehicle trajectory through the constrained corridor. Metrics from this “best-case” scenario establish the minimum threat posed to the vehicle given its current state. Based on this threat assessment, the level of controller intervention required to prevent departure from the navigable corridor is calculated and driver/controller inputs are scaled accordingly. This approach minimizes controller intervention while ensuring that the vehicle does not depart from a navigable corridor of travel. It also allows for multiple actuation modes, diverse trajectory-planning objectives, and varying levels of autonomy. Experimental results are presented here to demonstrate the framework’s semiautonomous performance in hazard avoidance scenarios.

Dynamic trajectory planning for trailer backup

We consider the directed waypoint following problem for automated trailer backup. Waypoint following is achieved through the design of a hybrid planner, which is computationally efficient and robust to tracking errors. The planner requires feedback from a localized positioning system. Simulation results are generated using the vehicle dynamics simulator CarSim.