Application of global optimisation algorithms to multi-rotor systems

Abstract

The presented study proposes a novel method for optimising rotors’ position in a generic, non-overlapping multi-rotor system. The approach uses global optimisation algorithms with an objective function based on a recently developed low-order aerodynamic model for rotor-on-rotor interference “RORI”. Two classical optimisation algorithms, Particle Swarm Optimisation and Genetic Algorithm, were used to ensure an algorithm-independent solution. The analysis was conducted on Twin-rotor, Tri-rotor and Tetra-rotor systems in a two-dimensional, unbounded optimisation space at a constant operating point. The results indicated that the best available minimum of the power ratio existed when rotors were reassembled into an inverse V-shaped formation pattern. The performance assessment of optimisation algorithms revealed that the Particle Swarm Optimisation outperformed the Genetic Algorithm in terms of mean execution time and solution repetitiveness for a given optimisation problem. The comparison against the Free-Vortex Wake model showed an agreement within 5% across all verification cases, indicating the potential of the proposed approach.