K.A. Unyelioglu

Design and stability analysis of a lane following controller

This paper considers the design and stability analysis of a steering controller. The objective of the controller is to steer a ground vehicle along a reference line located in the middle of the lane to be followed. We define an arbitrary look-ahead point located on the local longitudinal axis of the vehicle. The distance between the look-ahead point and the reference line is called the look-ahead offset. During perfect lane tracking, the ratio of the steer angle to the look-ahead offset is independent of the curve radius under reasonable approximations. That ratio is computed in terms of the vehicle speed and various vehicle parameters. Then, a constant controller is designed to achieve that ratio at steady state. The controller is a continuous function of the vehicle speed. The only information processed by the controller is the look-ahead offset, which can be measured using a radar-based or a vision-based sensor. Using Routh-Hurwitz analysis, we analytically prove that the closed-loop system is stable. Given any range of longitudinal speeds, there exists a sufficiently large look-ahead distance ensuring the closed-loop stability for all speeds in that speed range. For a particular set of parameter variations, it is also shown that choosing the look-ahead distance large enough guarantees the robustness of closed-loop stability, Various simulation results demonstrating the performance of the controller are included.

An analytical study of vehicle steering control

This paper considers the design and stability analysis of a steering controller. The objective of the controller is to steer a ground vehicle along a reference line located in the middle of the lane to be followed. The authors define an arbitrary look-ahead point located on the local longitudinal axis of the vehicle. The distance between the look-ahead point and the reference line is called the look-ahead offset. During perfect lane tracking, the ratio of the steer angle to the look-ahead offset is independent of the curve radius under reasonable approximations. That ratio is computed in terms of the vehicle speed and various vehicle parameters. Then, a constant controller is designed to achieve that ratio at steady-state. The controller is updated as a function of the vehicle speed. The only information processed by the controller is the look-ahead offset, which can be measured using a radar-based or a vision-based sensor. Using Routh-Hurwitz analysis, the authors analytically prove that the closed-loop system is stable. Given any range of longitudinal speeds, there exists a sufficiently large look-ahead distance ensuring the closed-loop stability for all speeds in that speed range.

On optimal design of a lane change controller

This paper is concerned with automated lane change maneuvers. We first consider the time optimal open loop lane change controller design with nonlinear constraints on the design parameters. Then, we combine the open loop design with a modified closed loop lane following controller so that the closed loop system accomplishes the maneuver as in the open loop case. The performance of the overall controller is demonstrated by simulation examples.

H ∞ sensitivity minimization using decentralized feedback: 2-input 2-output systems

H∞ norm minimization of the sensitivity and complementary sensitivity functions of 2-input 2-output linear time-invariant finite-dimensional systems using decentralized feedback is considered. It is shown that a set of fixed zeros and a set of fixed poles play a central role in the related minimization problems, and lead to various performance limitations. A comparison between decentralized and centralized controllers in these minimization problems also yields a rigorous characterization of a class of 2-input 2-output plants for which centralized controllers bring no extra benefits compared to decentralized controllers in some special performance objectives.

Nonstandard control inputs in the integrated design of vehicles

We consider a vehicle equipped with rear steering, variable drive/brake torque proportioning, and normal force control. Using extensive nonlinear and linear models, and using the notion of reachable sets, an open-loop analysis is conducted to investigate the possible benefits of these nonstandard control inputs in enhanced maneuverability and handling.

Design of a Lateral Controller for Cooperative Vehicle Systems

This paper considers the lateral control problem for an automated vehicle system. Different theoretical and experimental results obtained through experiments at Ohio Statue University are described. The control system is based on a microwave radar which also provides distance and speed measurements for the longitudinal control purposes. Emphasis is on the lateral controller design procedure.

Advanced Automatic Lateral Control Schemes for Vehicles on Highways

In this paper, the authors examine open-loop and closed-loop lane change maneuvers for autonomous vehicles. Focus is on determining the steering signal A previously defined lane following controller is modified to achieve the steering signal using a feedback controller. Simulation examples illustrate the effectiveness of the developed methodology for automobiles and articulated vehicles.

Fixed zeros of decentralized control systems

Considers the notion of decentralized fixed zeros for linear, time-invariant, finite-dimensional systems. For an N-channel plant that is free of unstable decentralized fixed modes, an unstable decentralized fixed zero of channel i (1/spl les/i/spl les/N) is defined as an element of the closed right half-plane, which remains as a blocking zero of that channel under the application of every set of N-1 controllers around the other channels, which make the resulting single-channel system stabilizable and detectable. The paper gives a complete characterization of unstable decentralized fixed zeros in terms of system-invariant zeros.

A Decomposition Method for the Design of Active Suspension Controllers

Abstract A method is developed to decompose the vertical dynamics of a full car model with active suspension into noninteractiug roll, bounce and pitch, and actuator pressure dynamics. Based on this decomposition, a generic representation of a compensator is given which achieves pole placement, regulation, and partial eigenvector assignment for ride heights and normal force distribution control.

OPTIMAL DESIGN OF A NORMAL FORCE CONTROLLER

Abstract The ride heights (roll, pitch, and heave) and the normal force distribution control problem is considered for an automobile using an active suspension system. A nonlinear vehicle model is linearized at a straight driving operating point and an H optimal controller is designed for the linear model using a computer package. The resulting controller is applied to the original nonlinear model and the closed-loop system is simulated for different operating conditions.