Meihua Tai

Modeling and control of steering system of heavy vehicles for automated highway systems

This work presents modeling, analysis, and controller design of the steering subsystem of heavy vehicles as a subsystem of vehicle lateral control system for the automated highway systems. A physical model of the steering subsystem is derived where the hydraulic power assist unit is modeled as a family of static nonlinear boost curves. Based on open-loop frequency tests and analysis of the physical model structure and its dynamical characteristics, a nominal second order linear model of the steering subsystem is obtained. Then, a linear robust loop-shaping controller is designed to provide a good tracking performance of the closed-loop dynamics of the steering subsystem for varying gain cross over frequencies which is a result of the nonlinear characteristics of the hydraulic power assist. The controller has been successfully incorporated as an inner-loop controller into the nested lateral control architecture for autonomous driving and its efficacy has been demonstrated experimentally.

Robust longitudinal velocity tracking of vehicles using traction and brake control

Vehicle longitudinal control is an essential part of advanced vehicle control systems. The paper is concerned with robust longitudinal velocity tracking of vehicles using traction control and brake control. First, a longitudinal vehicle model is introduced both for traction and braking modes. Then, for each mode, a robust velocity tracking controller is designed based on backstepping. At each step of constructing a candidate Lyapunov function, a scaling parameter is introduced for each added term to take into account badly scaled system states. The concepts of effective traction mass and effective braking mass are introduced in the vehicle longitudinal model such that the derived model is applicable to any types of traction and braking configuration of the vehicles. The designed controller is proved to be effective by simulation.

Automated lane guidance of commercial vehicles

This paper is concerned with automated lane guidance of heavy vehicles in the context of automated highway systems. Emphasis is on tractor-semitrailer combinations. Vehicle models are formulated in both the vehicle coordinate frame and the road coordinate frame. The model in the latter frame is appropriate for dynamical analysis and controller design since the lateral error is measured relative to the road. The vehicle speed and the look-ahead distance, i.e., the location of lateral error sensor relative to the tractors center of gravity, significantly affect the open loop dynamics from the front wheel steering angle to the lateral error. This aspect is studied by linear analysis. Two classes of lateral controllers are presented: one class is designed based on linear H/sub /spl infin// loop shaping methods and the other is based on nonlinear robust control and adaptive control. Simulation results are presented to verify the linear and nonlinear designs.

Modeling and robust control of power steering system of heavy vehicles for AHS

This paper is concerned with the modeling and control of steering system as a sub-system of lateral control architecture of heavy vehicles for the automated highway systems (AHS). A steering system retrofitted with an actuator is considered. The input and output of the steering system are the reference steering angle command to the actuator and the actual steering angle of the front wheel respectively. Open loop experimental data is fitted to a second order linear model. A linear loop-shaping controller has been designed and experimentally verified. It has also been successfully used as an inner-loop controller of the vehicle lateral control system in the open and closed loop experiments of the heavy vehicle system.

Robust Lateral Control of Heavy Duty Vehicles for Automated Highway Systems

Abstract This paper is concerned with the automated lane following of heavy-duty vehicles in the context of automated highway systems. The problem is formulated in the road reference coordinate frame for a tractor-semitrailer system. Two robust control schemes are studied: one is based on the sliding mode control (SMC) and the other is adaptive robust control (ARC) which is an enhanced version of SMC with parameter adaptation. ARC achieves smaller tracking errors and better performances under both modeling and parametric uncertainties. The concept of virtual look ahead distance is discussed and it is shown that larger look ahead distances render smoother system dynamics while sacrificing trajectory tracking accuracy.

Nonlinear robust loop shaping controller design for automated lane guidance of heavy vehicles

A nonlinear robust controller design method is proposed for the lateral control of heavy vehicles for automated highway systems. In the controller design, the actuator dynamics is explicitly taken into account. A first order pre-filter is introduced to reduce chattering which is inherent in the nonlinear robust controller and to prevent the saturation in the front wheel steering angle and steering system slew rate. A constructive backstepping design technique is utilized in the design of a robust controller for the augmented system. Simulation results are shown.

ROBUST VELOCITY TRACKING CONTROL OF GENERAL TYPES OF VEHICLES WITH ARBITRARY TRACTION AND BRAKING CONFIGURATIONS

A longitudinal control model for general types of vehicle systems with arbitrary numbers of units is formulated. By introducing the concept of effective angular velocity, the model considers arbitrary configurations of an active set of traction wheels and an active set of braking wheels. It is assumed that inner-loop controllers are already designed for both traction actuators and braking actuators. The local closed-loop dynamics are approximated by linear first order systems with model uncertainty. Due to the discontinuity in the tyre longitudinal slip model, the vehicle longitudinal control model admits two different operating modes: traction mode and braking mode. For each mode, a robust backstepping controller is designed and the robustness is shown by simulation.

FEEDFORWARD COMPENSATION FOR LATERAL CONTROL OF HEAVY VEHICLES FOR AUTOMATED HIGHWAY SYSTEM (AHS)

Abstract In this paper, we propose a feedforward compensation method to enhance the performance of the lateral guidance system of heavy vehicles on automated highway systems (AHS). The feedforward controller blends well with the linear robust feedback controller. The proposed feedforward compensation is motivated by the analogy between a vehicle lateral control system on curved roads and a mechanical system with Coulomb friction. The sliding mode controller for vehicle lateral guidance is shown to inherently involve a feedforward compensation term. Performance of a tractor-semitrailer combination vehicle under linear robust feedback control is used as the base line to show by simulations the enhancement of performance by fixed gain and adaptive feedforward controllers. The tracking performance is further improved by sliding mode control. In view that the sliding mode controller is more complex than the linear controller, the combination of the linear feedback controller and the feedforward controller is at a mid-point between ease of implementation and control performance.