In the numerical calculation of the ILQ optimal regulator using the Kalman filter, the influence of magnitude of a measurement error and the rate of thruster power change on the position keeping performance of the FPSO are considered.

　

 When a measurement error is small, due to using the Kalman filter, it is possible to perform position keeping control of the FPSO based on the estimated value with high accuracy. Therefore, although the horizontal displacement of the FPSO increases little compared with that with no measurement error, there is almost no change of other terms.

　

 According to aggravation of the estimation accuracy by the increase in measurement error, the fluctuation of the thruster power (τB,τS) and the angular velocity r are also increase. In the event of τB　and τS, since the FPSO tends to move from the original trajectory by the command of the control system, which refer to the estimation value with low-accuracy, the thruster power is fluctuated in order to prevent such a phenomenon. By the fluctuation of the bow and stern thrusters, the value of the angular velocity r is changed greatly. Also, the FPSO has a large overshoot angle. Due to the large overshot angle, it is impossible to achieve the control objective of the FPSO rapidly. When a standard deviation becomes bigger than 2.2, it is impossible to achieve the control objective.

　

 In the condition of δTh = 1% , it is impossible to achieve the control objective. Since the thruster output is small, it is impossible to output thrust based on the estimated value acquired from the Kalman filter. Particularly, from the time history of the heading angle, it is found that the FPSO tends to rotate in the opposite direction of a target angle.

　

 When δTh becomes larger, the horizontal displacement of the FPSO decreases greatly. However, there is a tendency, for the fluctuation of he thruster power to increase.

　

[1] J. Takashina "Ship Manoeuvring Motion due to Tugboats and Its Mathematical Model", Journal of the Society of Naval Architects of Japan, Vol.160, pp. 93-102, 1986

[2] T. Fujiwara, M. Ueno and T. Nimura "Estimation of Wind Forces and Moments acting on Ships", Journal of the Society of Naval Architects of Japan, Vol.183, pp. 77-90, 1998

[3] J.N. Newman "The Drift Force and Moment of Ships in Waves", Journal of Ship Research, Vol.1, pp. 51-60, 1967

[4] T. Fujii and T. Shimomura "Generalization of ILQ Method for the Design of Optimal Servo Systems", Transaction of the Institute of Systems, Control and Information Engineers, Vol.1, No.6, 1988

[5] S. Arimoto "Linear System Theory", Sangyo Publishing, 1974

　

 Y. Ahn received his Doctor of Engineering degree in marine system engineering from Kyushu University, Japan in 2003. His current research interests are linear quadratic optimal control and its application to ships affected by external forces and interaction forces.

　

 K. Kijima received his Doctor of Engineering degree from Kyushu University, Japan in 1974. Presently employed as Professor of the Department of Marine Systems Engineering, Kyushu University. His main research interests are prediction of ship manoeuvrability and automatic ship navigation system.

　

 Y. Furukawa received his Doctor of Engineering degree from Kyushu University, Japan in 1993. Presently employed as Associate Professor of the Department of Marine Systems Engineering, Kyushu University. His current research interests are automatic ship navigation system and prediction of hydrodynamic forces acting on ship hull.