Coupled Task Allocation and Path Planning for Energy-Optimal Multi-AUG Multi-Target Exploration
Hao Hu, Tonghao Wang, Xingguang Peng · IEEE Transactions on Vehicular Technology · 2025
Autonomous underwater gliders (AUGs), as low-energy mobile observation platforms, offer significant potential for multi-target exploration tasks. However, existing studies typically use Euclidean distance to decouple task allocation from path planning, ignoring the effects of ocean currents and seabed topography on task performance. To address this gap, we develop an asymmetric energy consumption matrix that incorporates pitch angle and diving depth as decision variables, enabling energy-optimal path planning between targets. Subsequently, we propose a Discrete Artificial Bee Colony algorithm with Enhanced Search Strategies (DABC-ESS) for optimal task allocation. DABC-ESS integrates a greedy insertion initialization method with three innovative search strategies: enhanced neighborhood search, enhanced insertion, and enhanced destruction-reconstruction. The effectiveness of these strategies is validated through ablation studies, while comparative experiments against five state-of-the-art algorithms demonstrate the superior performance of DABC-ESS. Furthermore, our method achieves greater energy efficiency than the latest multi-AUG task allocation approach, highlighting its potential for complex underwater tasks.