Implementation of an Adaptive Autopilot Unit

Abstract

This paper presents details in the implementation of an adaptive autopilot unit used for ship course keeping and course changing. The control strategy is based on the scheme of self-tuning regulator (STR) in which the generalized minimum variance (GMV) control law is combined with an extended least square (ELS) parameter identification technique. A time-varying extended state Kaiman filter (EKF) is designed to estimate the state of the system under disturbances induced by the sea environment and loading conditions. ELS and EKF constitute an adaptive estimator by which both state vector and parameter vector can be estimated properly. A dual microcomputer system using warm standby redundant design is implemented to guarantee the reliability in the control and safety in the steering. The construction principles of hardware and software are discussed. A sophisticated simulation facility is extensively used in the design stage of the autopilot unit. Several sea trials have been performed at “DAQING 232”, a 15000 tdw tanker, the results of which show that the unit reduces the propulsion losses of the ship motion in addition to keeping the course satisfactorily.