| The number of high-parameter thermal equipment operated in power plants is increasing,which puts forward higher requirements on the performance of heat-resistant steel.The current conventional and nondestructive testing techniques have certain limitations on the time and operation of the failure assessment of metal,hence it is a significant to develop the in-situ and rapid running state evaluation technique for equipment.During the service process of the hightemperature pressure-bearing equipment,the microstructure of the steel is constantly aging,and the macro performance is gradually decreasing.The macroscopic and microscopic damage of the steel makes obvious different in the sample matrix,and the laser-induced breakdown spectroscopy(LIBS)was affected by matrix effect.Hence,LIBS technique was applied to the rapid in-situ evaluate the steel aging estimation in this work.The laser spectroscopy tomography of typical heat-resistant steel service in industry was first performed.The spectral line intensities of the oxide layer on the surface and the matrix along the radial direction were analyzed.Then the correlation analysis of the spectrum of the presence or absence of oxide layer interference in industrial pipe was carried out,it was found that the change trend of the line intensities along the radial direction of the pipeline can realize the identification of the boundary between the oxide layer and the steel matrix,and obtain the representative spectrum of the matrix.In addition,the influence of the surface oxide layer on the plasma characteristics of the steel matrix was discussed.Then different pre-processing methods(multiplicative scattering correction,standard normal variate transformation,and wavelet transform)were used to correct the interference caused by the oxide layer to the spectral characteristics of the steel matrix.The influence of spectral correction on the K-fold cross-validation support vector machine recursive feature elimination(K-SVM-RFE)feature selection process was analyzed,and the steel aging estimation model was established based on the optimal feature subset.The results showed that multiplicative scattering correction(MSC)can effectively correct the interference of the steel surface oxide layer.The steel aging estimation model based on the optimal feature subset has a prediction accuracy of 84.72% on the external test set,and the model performs similarly on samples with oxide layer interference and corresponding oxide layer interference eliminated.Then a highly integrated portable LIBS measuring device principle prototype was developed.The optimal feature subset selected based on K-SVM-RFE was modified by MSC spectrum and the aging estimation model was established.Subsequently,the model was packaged to develop a real-time measurement software for steel characteristics of the prototype,and the performance of the portable device prototype was tested and verified in the laboratory,which provided an effective basis and method for the portable LIBS equipment to realize the steel aging estimation in service.Finally,the result of the whole research work was summarized and the directions for the further studies were suggested. |
