| 摘要 |
PURPOSE:To improve traveling stability by memorizing the detection data of a road surface condition in time series, judging requirement of emergency avoidance from combined value obtained as its learning result and feed-back controlling rear wheels in accordance with a detected yaw rate when emergency avoidance is required. CONSTITUTION:A condition of a road surface to travel on is detected by a parameter detection means A5 as a parameter, and it is memorized in a memory means A6 as a time series data at a specified time interval. A combination degree learning means A7 finds combined value by way of combining a combination degree previously learned about this memory data and sends it to an emergency avoidance judging means A8. The emergency avoidance judging means A8 judges whether it is an emergency avoidance condition or not in accordance with this combined value, and when emergency avoidance is required, it outputs a correction command to a gain correction means A9. Consequently, yaw rate feed back gain for the rear wheel steering mechanism A3 of a feed back control means A4 in accordance with a detected yaw rate of a yaw rate detection means A2 is corrected. |
