Compressed Sensing Application In Lamb Wave Structure Damage Monitoring

Structural health monitoring technology can be real-time online monitoring of large structures, and play a vital role in ensuring structural safety. The distributed sensor network structure is generally used in the Lamb wave structural health monitoring system, large number of sensor nodes and high frequency sampling rate leading to the system energy consumption during signal transmission. Therefore, how to extract the important information effectively from the massive data has become a new hot spot of the structural health monitoring. Due to the problem of high cost of signal transmission and low efficiency of traditional reconstruction algorithm, a method of compressed sensing based on Lamb wave structural health monitoring is proposed in this thesis.The research work of this thesis was illustrated as follow:(1) The research background and contents of structural health monitoring are given. Focusing on the problems encountered in the field of structural health monitoring, such as massive collection of data, storage and transmission pressure, a classification system of Lamb wave signal processing method based on compressive sensing is established.(2) The propagation characteristics of Lamb wave are investigated theoretically, excitation process and the application of compressive sensing based on Lamb wave are discussed. And take further research on sparse decomposition, observation matrix and the limitation of the principle.(3) Firstly, through the analysis of traditional reconstruction algorithm shortcomings based on Lamb wave, then put forward the improved compressed sensing reconstruction algorithm. Under the ensurance of reconstruction accuracy, this algorithm can improve the signal compression ratio, reduce the error rate and save the cost of signal transmission. Finally using the tomography method to extract the characteristic parameters of damage.(4) An integrated experimental system is designed for monitoring and evaluation of three different state of damages in the aluminum plate based on compressive sensing. Compare the evaluation results with the traditional sampling methods and then analyze the advantages and the accuracy of the results.The experimental results show that the proposed method can reduce the network loss and redundant information burden within the margin of error. Furthermore, it provides technical method for damage online monitoring and evaluation with a good application foreground.