Metacognitive Decision Making Framework for Multi-UAV Target Search Without Communication

TL;DR

This paper introduces a metacognitive decision-making framework for decentralized multi-UAV target search that enhances sensor switching and search strategies without communication, improving detection and confirmation efficiency.

Contribution

The paper proposes a novel metacognitive framework inspired by human cognition, enabling UAVs to self-regulate multi-sensor search strategies in unknown environments without communication.

Findings

MDM outperforms existing stochastic search algorithms in target detection and confirmation.

The framework effectively manages sensor switching based on sensing attributes and search levels.

Simulation results demonstrate improved efficiency in unknown environments.

Abstract

This paper presents a new Metacognitive Decision Making (MDM) framework inspired by human-like metacognitive principles. The MDM framework is incorporated in unmanned aerial vehicles (UAVs) deployed for decentralized stochastic search without communication for detecting stationary targets (fixed/sudden pop-up) and dynamic targets. The UAVs are equipped with multiple sensors (varying sensing capability) and search for targets in a largely unknown area. The MDM framework consists of a metacognitive component and a self-cognitive component. The metacognitive component helps to self-regulate the search with multiple sensors addressing the issues of "which-sensor-to-use", "when-to-switch-sensor", and "how-to-search". Each sensor possesses inverse characteristics for the sensing attributes like sensing range and accuracy. Based on the information gathered by multiple sensors carried by each UAV, the self-cognitive component regulates different levels of stochastic search and switching levels for effective searching. The lower levels of search aim to localize the search space for the possible presence of a target (detection) with different sensors. The highest level of a search exploits the search space for target confirmation using the sensor with the highest accuracy among all sensors. The performance of the MDM framework with two sensors having low accuracy with wide range sensor for detection and increased accuracy with low range sensor for confirmation is evaluated through Monte-Carlo simulations and compared with six multi-UAV stochastic search algorithms (three self-cognitive searches and three self and social-cognitive based search). The results indicate that the MDM framework is efficient in detecting and confirming targets in an unknown environment.