Cooperative search for dynamic targets by multiple UAVs with communication data losses

Area of Science:

• Robotics and Automation
• Multi-Agent Systems
• Navigation and Control

Background:

• Cooperative searching by multiple unmanned aerial vehicles (UAVs) faces challenges due to inconsistent environmental information.
• Packet losses and detection discrepancies degrade the shared environmental data among UAVs.

Purpose of the Study:

• To develop a robust cooperative searching strategy for dynamic targets using multiple UAVs.
• To address information inconsistencies and optimize real-time path planning for enhanced search efficiency.

Main Methods:

• An improved Least Square Method (LSM) compensates for lost information by incorporating target location models.
• The Weighted Averaging Method (WAM) merges information from multiple sources, assigning weights based on data uncertainty.
• A Modified Genetic Algorithm (MGA) coupled with rolling optimization generates dynamic search paths for UAVs.

Main Results:

• The proposed methods effectively unify environmental information among UAVs, enhancing data fitting precision.
• Real-time path generation enables UAVs to dynamically search for targets and re-enter search areas.
• Simulation results demonstrate the superiority of the cooperative searching strategy compared to existing methods.

Conclusions:

• The developed strategy significantly improves cooperative searching performance for dynamic targets by multiple UAVs.
• Information fusion and adaptive path planning are critical for successful autonomous search missions.
• This research offers a validated approach for complex multi-UAV search and tracking operations.