Entropy-Based Health State Evaluation of Unmanned Cluster Systems

Unmanned cluster system refers to an overall system in which a number of unmanned systems cooperate to accomplish complex tasks in a certain time and space according to the division of tasks. In recent years, the technology of unmanned cluster system has been gradually intelligent and the application has been gradually open, and its health state is dynamic and changeable, and the health state assessment of unmanned cluster system faces new problems. To address this problem, this study analyzes the current health state elements of unmanned cluster systems, and divides them into three levels: topology, traffic, and task. And based on the characteristics of unmanned cluster system, the system was modeled as a multi-layer complex network. The health state indicators were constructed through the theory of reliability entropy and group entropy combined with the physical characteristics of clusters. The health state assessment of unmanned cluster system was carried out by using the entropy indexes, and finally, the method was verified by simulation cases, which were analyzed by Wiener process simulation and multi-intelligence body simulation to better guide the health state assessment of unmanned cluster system in practical applications.