Digital transformation has become a key driver for revitalizing rural cultural tourism, yet the lack of unified data management, real-time monitoring, and predictive analysis remains a major bottleneck in rural tourism governance. To address these challenges, this study proposes a Digital Twin (DT) system tailored for rural cultural tourism scenarios. The system is designed based on a five-layer architecture integrating physical entities, multi-source data simulation, prediction models, virtual mapping, and application services. In the absence of real deployment conditions, we establish a fully simulated experimental environment that incorporates three key models—visitor flow prediction, Markov-based tourist behavior simulation, and linear programming–based resource optimization. Using simulated spatiotemporal data, the system evaluates performance across prediction accuracy, behavioral rationality, and optimization effectiveness. The results demonstrate that the proposed DT system effectively supports rural tourism management, enabling scenario reproduction, demand forecasting, and service allocation optimization. This research contributes a reproducible DT modeling and simulation framework for rural tourism and provides methodological insights for advancing digital village construction.