Abstract: An adaptive GPC algorithm for ship course-keeping is proposed. This method is applied to the ship course-keeping control system. The computer simulation results show the effectiveness of this method. Key words : Generalized predictive control; Adaptive control; Course-keeping 1 Introduction Ship steering automatic steering has always been one of the important research topics in ship motion control. It is closely related to navigation safety, energy conservation, and labor-saving operation. Experts and scholars engaged in this field have been making unremitting efforts in research. Adaptive generalized predictive control evolved from adaptive control by incorporating predictive control principles. Its main characteristics are: based on an identified process parameter model, it possesses a self-tuning control mechanism to correct model parameters online; it uses a controlled autoregressive integral moving average (CARIMA) model to describe the process, exhibiting inherent integral action, which not only naturally eliminates drift but also lays a reliable foundation for designing robust controllers; and it introduces multi-step prediction and rolling optimization strategies based on minimum variance control, overcoming the drawback of minimum variance control's high sensitivity to hysteresis, thus improving the system's stability, robustness, and applicability. For details, please click: Ship Heading Maintenance Based on Adaptive Generalized Predictive Control