目的 解决目前物流系统中智能包装RFID标签冲突的问题。方法 在研究已有防碰撞算法的性能和缺点基础上，针对大量标签场景设计一种分组动态帧时隙的混合查询树（GDFSA-HQT）算法。该算法在每一轮识别之后估计还没有识别的标签数量，如果还没有识别的标签数量小于或等于354个，则先采用动态帧时隙ALOHA（DFSA）算法对标签进行识别，再采用混合树查询算法（HQT）进行标签识别；若未被识别标签数大于354个，则先对标签进行分组处理，再分别采用DFSA和HQT进行标签识别。结果 仿真实验表明，GDFSA-HQT算法的吞吐率能够保持在0.82左右。结论 GDFSA-HQT算法解决了标签碰撞问题，在大量智能包装的物流系统中具有良好的应用前景。

The work aims to solve the intelligent packaging RFID tags conflict in the logistics system at current stage. On the basis of studying the performance and shortcomings of existing anti-collision algorithms, a grouped dynamic frame slotted ALOHA-Hybrid Query Tree (GDFSA-HQT) algorithm was designed for a large number of tags scenarios. The GDFSA-HQT algorithm estimated the number of unrecognized tags after each round of tags identification. If the number of unrecognized tags was less than or equal to 354, the dynamic frame slotted ALOHA (DFSA) algorithm was used to identify the tags firstly, and then the hybrid tree query algorithm (HQT) was used for tags identification. If the number of unrecognized tags was greater than 354, the tags should be first grouped, and then DFSA and HQT would be used for tags identification. The simulation experiment showed that GDFSA-HQT algorithm throughput rate could be maintained at around 0.82. The GDFSA-HQT algorithm solves the tag collision problem and has a good application prospect in a large number of intelligent packaging logistics systems.