Ultrasonic guided-wave testing is one of the most widely used technology for Structural Health Monitoring (SHM) of rail tracks. Currently, cable is the main tool of signal transmission for guided wave-based track monitoring systems. The installation of cables can significantly increase the system cost and restrict the flexibility of system deployment. In recent years, the NB-IoT technology has been gradually appied to the field of SHM, it offers long-range wireless communication among a large-scale sensor networks at the cost of minimum construction and maintenance. One primary obstacle hindering the integration of NB-IoT and guided wave-based track monitoring system is that the limited channel bandwidth of NB-IoT leads to significant transmission delay when transmitting the ultrasonic guided-wave signal sampled at Nyquist rate. In this paper, a Compressed Sensing (CS) framework for NB-IoT based rail-track monitoring system is proposed. The proposed CS framework utilizes the sparsity feature of the ultrasonic lamb-wave signal to enable sub-Nyquist sampling and maintain the feature of the measured signal at a low compression rate. To validate the proposed CS framework, the propagation time of lamb-wave is selected as the performance metrics. The experimental results show that compared with the traditional sampling method, the propagation time of lamb wave in rail track can be accurately extracted when the sampling rate is set to 100kHz, therefore, the channel bandwidth of NB-IoT can meet the delay-free data transmission of a single ultrasonic sensor.
Frequency-sweep examination for wave mode identification in multimodal ultrasonic guided wave signal