In view of the shortcomings of the real-time, accuracy and adaptability of the traditional power network equipment state monitoring method, this paper proposes a state monitoring algorithm optimization scheme based on the Internet of Things technology. By constructing the distributed sensor network and multi-source data acquisition model, and correcting the monitoring deviation caused by environmental interference, improve the accuracy of data collection; designing the dynamic safe transmission threshold evaluation mechanism to realize real-time interception and warning of abnormal data; at the same time, integrate the Internet of Things architecture and intelligent analysis algorithm to classify and dynamically update the operation status of the equipment. By comparing the topological performance of star, tree and network, the results show that the network structure is the best in bit error rate (0%–0.15%), signal to noise ratio (90–95 dB) and transmission delay (<65 ms), increasing by 84% and 64.1% compared with star and tree structure respectively. Through multi-dimensional data fusion and dynamic threshold optimization, the research has broken through the limitations of traditional static model, provided high precision, high security and robustness solution for equipment condition monitoring in complex power grid scenarios, and verified the practical value of Internet of Things technology in intelligent operation and maintenance of power system.