An improved path-tracking controller with mid-angle adaptive calibration for combine harvester

The assisted driving system of combine harvester with the mid-angle error carries the path-tracking deviation and limits the harvesting efficiency. In the manual calibration method, the accuracy and convenience of mid-angle calibration are limited by the unreliable driving skills. This paper analyzes the trajectory of harvester and builds the error model of mid-angle. An improved path-tracking controller with mid-angle adaptive calibration is proposed. The new controller calculates the average lateral errors of harvester through an interval window adjusted by fuzzy controller. The mid-angle is calibrated and the path-tracking deviation is restrained by the increment PID model. To improve the calibration accuracy and convergence speed, the PID parameters are optimized by particle swarm optimization (PSO). However, the inertia weight of adaptive inertia weight PSO decreases too fast in path-tracking system, a high-order system. The improved PSO, proposed in this paper, combines the iteration number and feedback function as the inertia weight strategy and adapts inertia weight. The experiments demonstrate that the new controller can restrain the path-tracking deviation from 17.02 cm to -0.73 cm and improve the accuracy of harvesting.