Abstract: ObjectiveAiming at the nonlinear ship course control system with nonlinear terms, and considering the influence of ship speed variation on course-keeping performance, this paper proposes an adaptive course-keeping control algorithm based on speed characteristic feedback. method A 3-degree-of-freedom nonlinear mathematical model of ship motion is selected as the ship design model. A radial basis function (RBF) neural network is applied to online approximate the nonlinear terms in the course model, with compensation implemented in the control algorithm. Combined with engineering practice, the relationship between the speed parameters and ship maneuvering indices during ship navigation is considered. The control system calculates the speed feedback signals, and according to the speed variation, a speed correction rule is designed to modify the control system, so as to improve the robustness of the control system. Results Through Lyapunov stability theory, it is proven that the designed control algorithm based on speed characteristic feedback satisfies the theory of semi-global uniformly ultimately bounded (SGUUB) stability. The results of simulation experiments show that the course-keeping control algorithm based on speed characteristic feedback designed in this paper can achieve fast and stable course keeping under different speed conditions, while reducing the system energy consumption by 24.07%. Conclusion Therefore, the proposed control algorithm has certain practical application value and bears engineering significance for improving the performance of ship course-keeping control.