The inefficiency and high cost of last-mile delivery in rural areas are major challenges for modern logistics systems. To address these issues, a truck-drone collaborative delivery routing optimization problem with dynamic rendezvous points and time synchronization was investigated. An integer programming model was formulated to minimize delivery cost and waiting-time cost. A two-level interactive optimization algorithm based on a dynamic rendezvous mechanism was developed. The outer layer was implemented with an improved GA (genetic algorithm) to determine truck routes, while the inner layer was executed with an improved DBO (dung beetle optimization) algorithm to optimize drone paths. The dynamic rendezvous points between the truck and drone were determined using the golden-section method to achieve precise temporal coordination. The effectiveness and feasibility of the proposed algorithm were validated through comparisons with Gurobi solutions on Solomon benchmark instances. A case study conducted in Changqingqiao Town, Qingyang City, Gansu Province, demonstrates that the proposed dynamic coordination mode achieves up to 31.4% cost reduction compared with the static strategy and up to 38.8% compared with truck-only delivery. The advantages are most evident when customer locations are numerous and widely dispersed. Sensitivity analysis shows that drone range and customer scale significantly affect delivery cost. The economic benefits of the dynamic coordination mode become more pronounced as drone range increases and customer density grows.