Architecture-independent quaternion-based attitude planning and control allocation for multirotors

This paper presents an architecture-independent framework for multirotor simulation based on a dynamic model for non-reversible rotors, a quaternion-based attitude planning, and null-space-based control allocation algorithm. This approach can be of use when investigating different multirotor architectures' characteristics and control algorithms for them, avoiding spending effort on attitude planning and control allocation for each different case. The framework is compared to current solutions by pointing out the advantages of the approach used here. Simulation results of three different architectures are presented aiming to explore the algorithm's robustness, without change, to arbitrary and non-symmetric rotor positions; under, fully or over-actuated architectures; and differently-sized rotors.