Study on the Automatic Maneuvering System of a Ship for the Disturbance
Grant-in-Aid for Scientific Research (B)
|Allocation Type||Single-year Grants|
|Research Institution||KYUSHU UNIVERSITY|
OGAWARA Yoichi Kyushu University, Faculty of Eng., Prof., 工学部, 教授 (20214033)
IWAMOTO Seiji Kyushu University, Faculty of Eng., Asst., 工学部, 助手 (80091338)
SHINKAI Akiji Kyushu University, Faculty of Eng., Asst.Prof., 工学部, 助教授 (10112301)
KIJIMA Katuro Kyushu University, Faculty of Eng., Prof., 工学部, 教授 (90038042)
|Project Period (FY)
1995 – 1996
Completed(Fiscal Year 1996)
|Budget Amount *help
¥3,900,000 (Direct Cost : ¥3,900,000)
Fiscal Year 1996 : ¥1,100,000 (Direct Cost : ¥1,100,000)
Fiscal Year 1995 : ¥2,800,000 (Direct Cost : ¥2,800,000)
|Keywords||Ship / Maneuvering Motion / Control System / Follow-up Control / Learning Control / Disturbance Compensation / Feed-forward Control / Inverse Transfer Function / 制御 / 外乱 / 風 / 逆システム / 学習|
1. Study on the mathematical model for the influence of the wind disturbance to the maneuvering motion
The mathematical model was obtained which was added the influence of the wind disturbance to the MMG model.
2. Study on the mathematical model for the simple estimation of the wind disturbance in the compensation system
It could be made sure by the result of the model experiment done before ago that the pressure and the moment to the ship from the wind could be estimated precisely and easily with the Isherwood estimation expression.
3. Study on the performance of the wind disturbance compensation system
(1) It was confirmed that the influence of the disturbance could be compensated with the linear comensation system using the inverse transfer function due to the learning function.
(2) It became clear that the compensation of the distrbance influence could be performed quickly due to the feed-forward control system.
4. Study on the compensation system to the plural disturbances (ex.wind and tidal current)
(1) It was confirmed that the influence of the plural disturbances could be compensated excellently without interaction.
(2) The method to decide the many coefficients of the learning equations suitable could be obtained, and it became possible that the learning could be carried out quickly even for the case of plural control variables and plural distrurbances.
Research Output (11results)