Abstract: To ensure the safe operation of the gate dam and to prevent accidents, the permutation entropy (PE) algorithm is used to monitor the operational status of the gate dam. A gate being the research object, and its flow induced vibration data by each measuring point of the gate dam are obtained. Then, the improved variational modal decomposition (IVMD) method is used to denoise the original signals from different measuring points under different operation conditions and extract the effective operational feature information of the structure. Followed, the feature information of each measuring point at the data level are fused by using variance contribution rate and the subsequence entropy values of the fused data are calculated by permutation entropy method. By analysis of the entropy characteristics of different vibration states, online monitoring of structural vibration state is realized. The results show that the permutation entropy method can not only comprehensively extract effective information of the structure, but also present the vibration state of the structure in a relatively intuitive form. Furthermore, the calculation process is simple and efficient, which can provide reference for the safe operation and online monitoring of gate dams.

Key words: gate dam, vibration monitoring, IVMD, permutation entropy, dynamic fusion

摘要： 为保证闸坝安全运行，防止安全事故发生，采用排列熵算法（PE）实现闸坝的运行状态监控。以某水闸为研究对象，首先，取得闸坝各测点的流激振动数据；其次，采用改进变分模态分解（IVMD）方法对各工况不同测点的原始信号进行降噪处理，提取结构的有效运行特征信息，并以方差贡献率的方式将各测点特征信息进行数据级融合；最后，通过排列熵方法计算融合后数据的子序列熵值，分析结构不同振动状态的熵值特点，实现结构振动状态的在线监控。结果表明，排列熵方法能全面地提取结构的有效信息，以相对直观的形式表现结构的振动状态，且计算过程简单、效率较高，能为闸坝的安全运行与在线监控提供参考。

关键词: 闸坝, 振动监控, IVMD, 排列熵, 动态融合