基于自回归模型参数超声无损评价20钢的时效组织

Ultrasonic Nondestructive Evaluation of Steel Aging Microstructure Based on Autoregressive Model Parameters

摘要: 基于时序分析理论,提取并优化选择表征20钢不同时效组织形态超声回波信号的自回归(AR)模型参数,用于超声无损评价.结果表明:利用AR模型参数无损评价20钢的时效组织形态是可行的,对五种组织的识别率可达84％;此外,对Cr5Mo钢高温时效组织形态和30Mn2SiV结构钢的热处理组织形态的识别率分别达到80%和88%.

Abstract: Based on the time series theory,the autoregressive (AR) model parameters which had been extracted and optimized were applied to the ultrasonic nondestructive evaluation of 20 steel aging microstructure at high temperature.The results showd that it was feasible to evaluate 20 steel aging microstructure at high temperature using auto-regression model parameters,and the recognizing ratio was 84% for five sorts of aging microstructure at high temperature.In addition,the recognizing ratios for aging microstructure of Cr5Mo at high temperature and heat treatment microstructure of 30Mn2SiV were 80% and 88%,respectively.