提出了一种基于流形半监督学习的移动节点定位算法.该算法利用基于流形学习的半监督方法,通过一定量的有标签样本和无标签样本,获取隐含在节点接收信号强度信息中的流形结构,直接建立节点物理位置与接收信号强度之间的映射关系.算法不需要使用现有的理论或经验信号传播模型,避免了模型不准确带来的定位误差,而且允许网络中存在大量无标签样本,降低了数据采集难度,提高了算法实用性.冶金工业现场的实际应用结果表明,相对RADAR算法,本文算法具有较高的定位精度. This paper proposes a semi-supervised learning-based mobile node localization algorithm based on manifold learning, which uses a semi-supervised method based on manifold learning to obtain the signal implicit in the received signal strength through a certain number of labeled samples and unlabeled samples , Which directly establishes the mapping relationship between the physical location of the node and the received signal strength.The algorithm does not need to use the existing theoretical or empirical signal propagation model and avoids the positioning error caused by inaccurate model and allows the network to exist A large number of unlabeled samples reduce the difficulty of data acquisition and improve the practicability of the algorithm.The practical application results of the metallurgical industry field show that the proposed algorithm has higher positioning accuracy than RADAR algorithm.