Abstract
ITS (intelligent transport systems) has been expected to bring us dramatic decrease of traffic accidents. A vehicle in the system is regarded to be possible not only to detect its own motion but also to get information of road and traffic environments. For a example, an information of relative position of other vehicles to the subjective vehicle in traffic flow are possible to be obtained by using such as an interactive communication system between vehicles and an embedded laser radar system. Depending upon such information, more intelligent driver-assist-system for active safety of driver-vehicle-system will be possible.
From the above point of view, a possibility of the active drive-assist-system during evasive lane change in a traffic flow based upon recognition and understanding of behavior of vicinity vehicles around the subjective vehicle is discussed in this study. It is supposed in this study that a single sine wave lateral acceleration response will give us a single lane change with some width and once the lane change width, the obstacle distance in front of the subjective vehicle and longitudinal deceleration during the lane change are given, the single sine wave lateral acceleration needed to finish the evasive lane change to avoid the obstacle is possible to be calculated. It is further supposed in this study that the yaw moment produced by longitudinal forces of tires is used for the lane change assist control. The reason why the longitudinal forces are to be used in the control is that an evasive lane change is always together with braking and the existing brake system is possible to be an actuator for the control without any additional hardware embedded. We call this type of control DYC (direct yaw moment control). It is more suitable for the DYC to control vehicle yaw rate rather than to control lateral acceleration, therefore, the single sine wave lateral acceleration needed to accomplish the evasive lane change is approximated by a single sine wave yaw rate. Then it is possible to analytically predict the required yaw moment to accomplish the lane change with the single sine wave response of yaw rate by a model following method based on an on-board-tire-model as well as feed-backed information of the vehicle motion.
It is easily found from a simple analysis of two degree of freedom vehicle plane model that there exists some time lag in the lateral acceleration response to yaw rate one subjected to the yaw moment input. Therefore, though product of yaw rate and vehicle speed is equal to lateral acceleration at steady state, a phase advance in the sine wave response of yaw rate is needed to eventually get the lateral acceleration which can accomplish the evasive lane change required. The effects of this phase advance method in which the lateral acceleration is approximated by the yaw rate are validated with the computer simulation. The results show how a single sine wave lateral acceleration response is obtained controlling the vehicle yaw rate response by the predicted yaw moment in DYC.
Then if the driver requires the active lane change assist, the predicted yaw moment is used as a driver assist steering during the obstacle avoidance. In order to prove the effect of this active evasive lane change assist, a computer simulation of the driver-vehicle-system behavior with the assist system is carried out. It is shown that the assist system gives us high performance of the obstacle avoidance of the driver-vehicle-system. The effects are compared with the effects of the recently proposed chassis controls without any environmental information. The results show that the chassis control is eminently contributive to stabilize the vehicle motion during the lane change with braking, however, it is not necessarily contributive to quick avoidance of the obstacle. On the other hand, the assist steering system proposed is effective on compensating vehicle responsiveness for the handling delay of the human driver in a sudden evasive lane change.
Then it is conclusively mentioned that this type of new concept of active driver assist and control for the active safety of driver-vehicle-system in the ITS will be one of promising systems for decrease of still increasing traffic accidents.