Report on optimisation for efficient dynamic task distribution in drone swarms using QRDPSO algorithm

Abstract

The primary aim was to develop a Quantum Robot Darwinian Particle Swarm Optimisation (QRDPSO) algorithm and assess its performance against the conventional RDPSO. Using MATLAB-based mathematical modelling, QRDPSO was evaluated for its efficiency in dynamic task distribution and inter-drone communication stability. The results demonstrate that QRDPSO finds optimal solutions 16.3% faster than RDPSO, with performance improvements as the swarm size increases. Specifically, when the number of drones was increased from 5 to 20, the number of iterations required for QRDPSO changed from 384 to 189. However, for RDPSO, the number of iterations changed from 439 to 242. Additionally, QRDPSO showed a 27.1% reduction in drone loss rates, outperforming RDPSO in terms of maintaining operational resources, especially in larger swarms. These findings have practical implications, as QRDPSO’s efficiency and stability can support extensive drone applications requiring synchronised, reliable swarm behaviour.