Yukihiro Fujiwara

Development of obstacle avoidance system using co-operative control

Abstract This paper proposes a new driver support system which assists drivers operation in an obstacle avoidance scene. The proposed system has actuator systems which are Electric Power Steering (EPS) and Direct Yaw-moment Control (DYC). The obstacle is detected as a differential feature value by using a micro-wave radar instead of a camera. Co-operative control is constructed to manage the actuators by frequency weighting functions. By experiments using an actual vehicle, it is shown that the proposed system has effectiveness for assistance the drivers operation for an avoidance task and for improving vehicle dynamics.

Control design of steering assistance system for driver characteristics

This paper proposes a new driver support system which uses both the steering wheel angle and the steering torque of the vehicle so as to take into account the drivers operational characteristics. The proposed method is based on gain scheduled control to accomplish cooperation between the driver and the support system. By numerical experiments which simulate obstacle avoidance, it is shown that the proposed method has good tracking capability and cooperates well with the drivers operation.

Automated steering control system design for passenger vehicle in consideration of steering actuator dynamics

A control system is described for automated steering systems to prevent traffic accidents. A model of the system is built by integrating a lane model and a vehicle model based on the visual servoing approach. The system is controlled by a guidance controller and an angle controller. The former is designed on the basis of steering actuator dynamics and uncertainties of vehicle parameters, and the latter is designed on the basis of steering wheel dynamics using /spl mu/ synthesis. It is proved that the proposed control system achieves vehicle stability for steering maneuvers such as obstacle avoidance.

Control system design for obstacle avoidance system using model predictive control

This paper proposes a new driver support system which assists drivers operation in an obstacle avoidance scene. The proposed system is designed by model predictive control using a simplified model of the drivers operation. By numerical experiments which simulate for obstacle avoidance, it is shown that the proposed method has effect for operation of the avoidance task and reducing a drivers operational load.

Robust control of 4WS actuator using LMI

The actuator in a 4-wheel steering system must satisfy many control requirements; responsiveness and stability arising from improved vehicle dynamics, robustness against disturbance and noise in the operating environment, reduction of the effects of disparities in component quality, and electric power efficiency. We selected a design method based on the LMI technique in order to satisfy these requirements. This choice was due to the following three reasons: 1) A numerical solution can be obtained to design the controller because the requirements of responsiveness, stability, disturbance suppression and robustness can be managed using mixed H/sup 2//H/sup /spl infin// synthesis with regional pole placement; 2) A LMI-based control design has less limitation than Riccati-based control design. 3) External disturbance, which is dependent on the state quantities of the object of control, can be expressed as a model with polytopic uncertainty. This makes it possible for the controller design to take into account this uncertainty. Based on the above, we decided on an integrator-type optimal regulator incorporating a reference model for the structure of the controller. Control gain design results are described.