Efficient Multi-Target Localization Using Dynamic UAV Clusters.

Area of Science:

• Robotics and Control Systems
• Wireless Communication and Networking
• Optimization Algorithms

Background:

• Accurate multi-target localization in complex 3D environments using Unmanned Aerial Vehicles (UAVs) is challenging due to dynamic environments and limited resources.
• Existing collaborative localization methods often struggle with mobility-aware cluster formation and handling measurement/motion uncertainties.

Purpose of the Study:

• To propose a dynamic UAV clustering model for enhanced multi-target localization accuracy in complex 3D environments.
• To develop a robust algorithm for mobility-aware cluster formation that improves collaborative localization.
• To analyze localization performance considering measurement and motion-induced uncertainties via the Cramér-Rao lower bound (CRLB).

Main Methods:

• A dynamic UAV clustering model integrating mobility-aware cluster formation for enhanced collaborative localization accuracy.
• Derivation of the Cramér-Rao lower bound (CRLB) for performance analysis under uncertainties.
• Development of the Multi-Swarm Discrete Quantum-inspired Particle Swarm Optimization with Adaptive Simulated Annealing (MDQPSO-ASA) algorithm, including a repair mechanism for constraints.

Main Results:

• The MDQPSO-ASA algorithm demonstrates superior localization accuracy compared to baseline methods.
• The proposed model shows enhanced computational efficiency and adaptability to varying UAV and target scales.
• Simulation results validate the effectiveness of the mobility-aware clustering approach.

Conclusions:

• The dynamic UAV clustering model effectively enhances multi-target localization accuracy in complex 3D environments.
• The MDQPSO-ASA algorithm provides an efficient and adaptable solution for resource-constrained collaborative localization.
• This work offers a practical approach for real-world UAV-based localization applications.