Heterogeneous Multi-UAV Task Allocation Based on MTOC Algorithm

This paper proposes a Modified Tornado Optimizer with Coriolis force (MTOC) algorithm for the heterogeneous multi-UAV task allocation problem. We constructed a multi-tuple parametric model of the UAVs, targets, and constraints to portray the mission. Additionally, we designed a multi-objective evaluation function that combines maximizing task benefits, minimizing threat costs, and minimizing navigation costs to evaluate the superiority and inferiority of the obtained task allocation scheme. We enhanced the optimization-seeking capability of the Tornado Optimizer with Coriolis force (TOC) algorithm through two key improvements the application of Tent chaotic mapping to optimize population initialization distribution, and the integration of multi-stage simulated annealing to raise solution quality and stability. We tested the proposed algorithm in two scenarios: one with 6 heterogeneous UAVs and 12 random targets, and another with 12 heterogeneous UAVs and 40 random targets. Compared with TOC, Grey Wolf Optimizer (GWO), and Genetic Algorithm (GA), the proposed algorithm achieved improvements of 19.23%, 26.02%, and 22.53% in overall evaluation metrics for task allocation schemes, respectively. MTOC not only improved the convergence speed but also achieved more accurate solutions, achieving an evaluation metric variance as low as 0.0002. This represents a 98.71% reduction compared to the TOC algorithm.