Trajectory Tracking Control of Autonomous Lawn Mower Based on ANSMC

Abstract

For weeding in an open area, an automatic lawnmower is used. However, the positioning accuracy of the GPS used by the automatic lawn mower currently only reaches the meter level. Inertial positioning accumulates errors and causes control path errors, thus making the grass cannot be completely removed. Thus, this study is conducted to develop a lawn mower system with remote monitoring and automatic tracking control. The system uses a Dual Wheels electric transporter to set up the lawn mower pole while controlled by the rotation of the two wheels so that the lawn mower moves forward, turns, and moves backward. The positioning adopts simple RTK-GPS, including the base station and mobile station. It provides centimeter-level positioning accuracy. The movement trajectory planned with the computer is sent to the automatic lawn mower through the Lora wireless communication interface. The lawn mower follows the trajectory tracking control. The tracking control method of the lawn mower is based on Adaptive Neural Sliding Mode Control (ANSMC). The results show the maximum error of the tracking control of the automatic lawn mower is 10 cm. The computer monitors the condition of the lawn mower in real time and stores the tracking history to be checked. In the study, the prototype of the lawn mower is suitable for applications in open fields.