A vision-based approach for quadrotor line following with sliding mode controller

This article presents the design and implementation of a nonlinear control system considering the dynamics of a quadrotor that is employed in a line following mission. The initial estimated roll and yaw angles obtained through a vision-based line detection algorithm are fused with the data obtained from IMU and model-based prediction of the quadrotor’s position, altitude, and pitch angles. The rotational and altitude subsystems is controlled using nonsingular terminal sliding mode control (NTSMC) approach, while the position subsystem is controlled using sliding mode control (SMC). Both conventional SMC and NTSMC methods are designed and tested in a simulation environment before applying to real robot. The proposed control system demonstrates better performance in terms of robustness, tracking precision, and convergence rate compared to regular SMC. Experimental results show the effectiveness of hybrid control procedure in enabling the quadrotor to successfully follow a line in a challenging environment.