针对大规模分布式传感器网络提出一种拓扑三级分簇结构优化算法.通过引入传感器休眠模式,并考虑到分簇数目较多的情况,对多个簇头节点采用生成最小刚性图的方法进行拓扑优化,以实现传感器网络整体能量均衡,使传感器网络具有较好的连通性和鲁棒性.仿真实验表明,与已有相关算法相比,采用所提出的算法可使网络延缓出现节点死亡现象,有利于实现网络负载均衡,并且网络中节点整体存活时间较长,从而延长网络的生命周期. Aiming at large-scale distributed sensor network, a topology-based three-level clustering optimization algorithm is proposed. By introducing the sensor hibernation mode and taking into account the large number of clusters, the method of generating the minimum rigid graph Topology optimization to achieve the overall energy balance of the sensor network, the sensor network has good connectivity and robustness.The simulation results show that compared with the existing algorithms, the proposed algorithm can delay the network node death phenomenon , Which helps to achieve network load balancing, and the nodes in the network as a whole have a longer life span, thereby extending the life cycle of the network.