|
Abstract-This paper is concerned with the design of an adaptive
ship steering autopilot. Specifically, an Internal Model
Control(IMC) design method is coupled with a recursive least-
squares method in forming the adaptive autopilot. The IMC design
method is very attractive from the viewpoint of its relative
easiness in satisfying the nominal internal stability
conditions. A second order Nomoto model is taken as the ship
model in the controller design. The autopilot obtained with the
IMC design method is equivaa 2nd order PID type controller.
Moreover, the proposed IMC design method offers explicit
formulas relating the autopilot gain coefficients to the ship
model parameters and a design parameter that determines the
speed of response of the system. A recursive least-squares
method is employed in estimating the ship model parameters under
various operating conditions, and then the autopilot gain
coefficients can be determined accordingly. This allows the
autopilot to adapt to changing operating conditions. To avoid
instability phenomenon caused by rudder saturation and rudder
rate limitation, a reference conditioning technique is employed
to bring the rudder limit and rudder rate limit into the
adaptation design consideration. Specifically, the reference
signal is modified such that the rudder input command is always
consistent with the actual rudder input supplied by the
actuator. That is, the controller uses the full power of the
steering machine whenever necessary, without pushing the
steering machine to the nonlinear saturation and slew rate
limitation boundaries indefinitely. This so-called anti-windup
consideration is really necessary, since the wind-up problem is
much serious in adaptive control than in standard controller
with constant gain. Computer simulations indicate that the
proposed adaptive autopilot achieves satisfactory course-
changing maneuvers while sailing through waters of different
depth.
|