College of Water Conservancy and Civil Engineering, South China Agricultural University, Guangzhou 510642, China.
Sensors (Basel). 2021 Oct 30;21(21):7229. doi: 10.3390/s21217229.
Accurate estimation of cable tension is crucial for the structural health monitoring of cable-supported structures. Identifying the cable's force from its vibration data is probably the most widely adopted method of cable tension estimation. According to string theory, the accuracy of estimated cable tension is highly related to identified modal parameters including natural frequencies and frequency order. To alleviate the factors that impact the accuracy of modal parameters when using the peak-picking method in wireless sensor networks, a fully automated and robust identifying method is proposed in this paper. This novel method was implemented on the Xnode wireless sensor system and validated with the data obtained from Jindo Bridge. The experiment results indicate that, through this method, the wireless sensor is able to distinguish the cognizable power spectrum, extract the peaks, eliminate false frequencies and determine frequency orders automatically to estimate cable tension force without any manual intervention or preprocessing. Meanwhile, the results of natural frequencies, corresponding orders and cable tension force obtained from the Xnode system show excellent agreement with the results obtained using the Matlab program method. This demonstrates the effectiveness and reliability of the Xnode estimation system. Furthermore, this method is also appropriate for other high-performance wireless sensor network systems to realize self-identification of cable in long-term monitoring.
准确估计缆索张力对于索支撑结构的结构健康监测至关重要。从其振动数据中识别缆索的力可能是最广泛采用的缆索张力估计方法。根据弦理论，估计的缆索张力的准确性与所识别的模态参数（包括固有频率和频率顺序）高度相关。为了减轻在无线传感器网络中使用峰值拾取方法时影响模态参数准确性的因素，本文提出了一种完全自动化和鲁棒的识别方法。该新方法在 Xnode 无线传感器系统上实现，并使用锦州大桥获得的数据进行了验证。实验结果表明，通过该方法，无线传感器能够区分可识别的功率谱，自动提取峰值，消除虚假频率并确定频率顺序，无需任何手动干预或预处理即可估计缆索张力。同时，Xnode 系统获得的固有频率、相应顺序和缆索张力与使用 Matlab 程序方法获得的结果非常吻合。这证明了 Xnode 估计系统的有效性和可靠性。此外，该方法还适用于其他高性能无线传感器网络系统，以实现长期监测中的电缆自识别。
