Model predictive control of a multi-rotor with a slung load for avoiding obstacles

Abstract

This paper addresses the control problem of transportation of a slung load by a multi-rotor. Control of the slung load system cannot be achieved with usual controllers for multi-rotors because motion of the load is highly coupled with the multi-rotor. Although many controllers have been developed for the system, there still exist several problems. For instance, in trajectory-following control, many previous works make trajectory beforehand, and perform trajectory-following control afterward. Since the time spent to generate trajectories is too long for real-time control applications, alternative efficient algorithms are required. With this in mind, we adopt a fast nonlinear Model Predictive Control (MPC) algorithm. This control framework solves a nonlinear optimal control problem with a finite time horizon using Sequential Linear Quadratic (SLQ) solver, in MPC. The originally proposed work optimizes the implementation time and successfully conducts real-time waypoint flight of general multi-rotor. Here, we apply the framework to the slung load system. Also, an obstacle-avoidance algorithm suitable for the slung load system is developed for cluttered environments. Numerical simulations are conducted to validate the developed method.