Multi-UAV Cooperative Decision-Making Under Noisy and Uncertain Environments

Le résumé fourni par la source

In the process of multi-UAV cooperative decisionmaking, the information evolution model plays a key role in the communication and fusion of information among multiple agents. However, traditional information fusion models, such as the DeGroot model and the Hegselmann-Krause bounded confidence (HK) model in opinion dynamics, have certain limitations in handling complex noisy environments, adjusting weights dynamically, and processing linguistic information. To address these challenges, this paper proposes an information fusion model that integrates the social network mechanism, the hyperbolic tangent function, and the Monte Carlo method. In each round of information iteration, the model uses the social network structure to simulate the communication relationships among agents. Specifically, the social network is represented as a binary adjacency matrix. Then, a distance measure between linguistic information is defined. Based on this distance, the hyperbolic tangent function is used to determine the weights of agents during the evolution process dynamically. To fully evaluate the model’s performance under noise, Monte Carlo simulations are used for statistical analysis. A case of multi-UAV cooperative reconnaissance is presented to validate the effectiveness of the model. In addition, detailed comparisons with several existing models are conducted. The experimental results show that the proposed model demonstrates clear advantages in terms of anti-interference capability, stability, and adaptability to various linguistic information environments.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.