基于高斯过程回归的无人艇轨迹跟踪控制OA

When performing trajectory tracking,unmanned surface vessels often encounter difficulties in accurately fol-lowing reference trajectories due to disturbances caused by wind,waves,and currents.To address this challenge,a tra-jectory tracking control method based on Gaussian process regression is proposed for unmanned surface vessels within the framework of model predictive control.The Snow Goose optimization algorithm is utilized to optimize the hyperpa-rameters of the kernel function in the Gaussian process regression model.Afterward,the model is trained offline using disturbance-related data and optimized hyperparameters.The trained Gaussian process regression model replaces the hy-drodynamic model of the unmanned surface vessel in the model predictive control process to carry out trajectory track-ing.Experimental results show that this Gaussian process regression-based trajectory tracking control method achieves improved performance in the presence of environmental disturbances.Specifically,compared to the hydrodynamic model,tracking error along the x-axis is reduced by 30%to 60%,while tracking error along the y-axis is reduced by 30%to 50%.These findings demonstrate that the proposed trajectory tracking control method based on Gaussian pro-cess regression offers enhanced resistance to environmental disturbances for unmanned surface vessels.