Abstract
<title>Abstract</title> <p>Heterogeneous unmanned aerial vehicle (UAV) swarm cooperative scheduling constitutes a fundamental challenge in advancing UAV swarm intelligence. This problem entails the tightly coupled optimization of task allocation and sequencing, requiring simultaneous trade-offs among conflicting objectives, including mission completion time, energy consumption, and computational load balance, under multiple operational and physical constraints. To overcome critical limitations of the original Gold Mine Optimization Algorithm (GMO)- specifically, its insufficient population diversity and sluggish convergence toward the Pareto-optimal front- this paper proposes an Improved Gold Mine Optimization Algorithm with a dual-layer information-sharing framework and dynamic Pareto dominance (IGMO-DP). First, a dual-layer information-sharing mechanism is introduced that synergistically integrates local neighborhood enhancement with global elite guidance, thereby preserving population diversity while expediting convergence. Second, we design a dynamic exploration–exploitation balancing strategy based on adaptive step-size adjustment, enabling context-aware modulation of search behavior throughout the optimization process. Third, an elite archive maintenance scheme is incorporated, coupled with a dynamically updated Pareto dominance relationship to enhance both the convergence and uniformity of the final non-dominated solution set. Based on these algorithmic components, we formulate a multi-objective optimization model for heterogeneous UAV cooperative scheduling that jointly minimizes mission completion time, total energy expenditure, and inter-UAV load imbalance. The proposed IGMO-DP is rigorously evaluated across three representative complex operational scenarios. Comparative experiments demonstrate that IGMO-DP consistently surpasses the baselines, GMO, NSGA-II, MOPSO, and HMOEA in terms of convergence speed, diversity of the obtained solution set, and the quality of the Pareto front distribution. Quantitatively, the Inverted Generational Distance (IGD) metric improves by 14.2%–17.6%, while the Hypervolume (HV) metric increases by 4.9%–5.5%, thereby substantiating the algorithm’s efficacy and competitive advantage for heterogeneous UAV swarm scheduling.</p>