Autonomous Performance Evaluation of Unmanned Aerial Vehicle Based on Dynamic Bayesian Network

As a typical combat characteristic of Unmanned aerial vehicle (UAV), autonomy determines its combat limit. This study aims to comprehensively assess autonomous performance by identifying key influencing factors from four perspectives: perception, analysis, decision-making, and execution abilities. To achieve this, a Dynamic Bayesian Network evaluation model is proposed, effectively incorporating observation evidence from different time points to dynamically evaluate network nodes. Case simulations validate the applicability of this approach in assessing UAV's autonomous performance. The findings of this study have significant implications for advancing the theory of quantitative evaluation of autonomous performance, thereby enhancing UAV's combat capability.