An Adaptive Prediction Target Search Algorithm for Multi-AUVs in an Unknown 3D Environment.

Area of Science:

• Robotics
• Oceanography
• Artificial Intelligence

Background:

• Autonomous underwater vehicles (AUVs) are crucial for exploring unknown 3D underwater environments.
• Effective target search in these environments is challenged by sensor limitations and dynamic conditions.
• Existing algorithms often struggle with unknown environments, sensor noise, and AUV failures.

Purpose of the Study:

• To propose a multi-AUV collaborative target search algorithm using adaptive prediction for unknown 3D underwater environments.
• To enhance the efficiency and adaptability of AUVs in complex underwater search missions.
• To address challenges posed by sensor noise, underwater communication limits, and AUV failures.

Main Methods:

• Utilizing forward-looking sonar data to assess the current view and environmental state.
• Implementing an adaptive prediction algorithm for target location estimation and search area prioritization.
• Employing an evaluation function and task assignment strategy for efficient sub-region searching.
• Considering underwater communication constraints and AUV failure scenarios.

Main Results:

• The proposed algorithm effectively searches for both static and dynamic targets in unknown environments.
• It demonstrates robustness against sonar sensor noise, reducing positioning errors.
• The algorithm successfully performs target search tasks despite random AUV failures.
• Simulation results confirm high efficiency and adaptability under various conditions.

Conclusions:

• The multi-AUV collaborative target search algorithm based on adaptive prediction is highly efficient and adaptable.
• It provides a reliable solution for target search operations in challenging, unknown underwater environments.
• The algorithm's ability to handle AUV failures and communication limits makes it suitable for real-world applications.