Abstract:Due to the existence of uncertainty disturbance, the accuracy of the established kinematics model of the autonomous rice seeding machine was not of highly accuracy, which would lead to problems such as slow convergence time and poor path tracking effect. Firstly, in order to improve the accuracy of the kinematics model and reduce the convergence time, the fast terminal sliding mode control algorithm was proposed based on the unmanned rice seeding machine. To realize the accurate observation of the unknown disturbance, the corresponding nonlinear disturbance observer was established according to the kinematics model, and the disturbance observed would be compensated to the controller. Next, the stability of the control system was proved according to Lyapunov criterion. Then, the kinematics simulation model of the agricultural machinery was built by Matlab/Simulink. The simulation results showed that the uncertain disturbance can be accurately observed by the disturbance observer. The control algorithm had better performance than sliding mode control algorithm without disturbance observer, which can effectively reduce the convergence time and almost had no chattering phenomenon. Finally, an experiment in a paddy field was carried out. When machine worked efficiently at the speed of 1.2m/s, the absolute average lateral deviation was 0.0247m, and the mean square error was 0.0311m. The turning speed was also fast and there was no overshot phenomenon. The results showed that the path tracking effect of the algorithm proposed was of high performance and can satisfy the actual operation requirements well.