Research on Tracking Control of Unmanned Lawn Mower Based on Absolute and Visual Relative Positioning

As a new product to replace traditional manual lawn mowers, unmanned lawn mowers have been widely concerned by universities and enterprises. However, there are many problems when the traditional tracking control algorithms based on absolute positioning are applied to the unmanned lawn mower, including the influence of the ground and the outdoor environment. Consequently, this paper proposes a tracking control method for unmanned lawn mowers based on absolute and visual relative positioning. First, the threshold segmentation and boundary extraction of the image are carried out by using the excess-green feature. Based on the designed control quantities, this paper proposes the control strategies corresponding to different states. In addition, we propose the local reference track layer based on visual relative positioning and optimize the objective function of model predictive control. Finally, the feasibility of the improved MPC algorithm is verified by simulation and experiment. We believe this work should receive research attention to provide more references to tracking control of unmanned lawn mowers.