The characterization techniques were employed like transmission electron microscope,X-ray diffraction and microstructural characterization to investigate microstructural evolution and impact of precipitate-phase preci...The characterization techniques were employed like transmission electron microscope,X-ray diffraction and microstructural characterization to investigate microstructural evolution and impact of precipitate-phase precipitation on strength and toughness of a self-developed 32Si_(2)CrNi_(2)MoVNb steel during the quenching and tempering process.Research outputs indicated that the steel microstructure under the quenching state could be composed of martensite with a high dislocation density,a small amount of residual austenite,and many dispersed spherical MC carbides.In details,after tempering at 200℃,fine needle-shapedε-carbides would precipitate,which may improve yield strength and toughness of the steel.However,as compared to that after tempering at 200℃,the average length of needle-shapedε-carbides was found to increase to 144.1±4 from 134.1±3 nm after tempering at 340℃.As a result,the yield strength may increase to 1505±40 MPa,and the impact absorption energy(V-notch)may also decrease.Moreover,after tempering at 450℃,thoseε-carbides in the steel may transform into coarse rod-shaped cementite,and dislocation recoveries at such high tempering temperature may lead to decrease of strength and toughness of the steel.Finally,the following properties could be obtained:a yield strength of 1440±35 MPa,an ultimate tensile strength of 1864±50 MPa and an impact absorption energy of 45.9±4 J,by means of rational composition design and microstructural control.展开更多
业务流程合规性检查可以帮助企业及早发现潜在问题,保证业务流程的正常运行和安全性。提出一种基于改进BERT(Bidirectional Encoder Representations from Transformers)和轻量化卷积神经网络(CNN)的业务流程合规性检查方法。首先,根据...业务流程合规性检查可以帮助企业及早发现潜在问题,保证业务流程的正常运行和安全性。提出一种基于改进BERT(Bidirectional Encoder Representations from Transformers)和轻量化卷积神经网络(CNN)的业务流程合规性检查方法。首先,根据历史事件日志中的轨迹提取轨迹前缀,构造带拟合情况标记的数据集;其次,使用融合相对上下文关系的BERT模型完成轨迹特征向量的表示;最后,使用轻量化CNN模型构建合规性检查分类器,完成在线业务流程合规性检查,有效提高合规性检查的准确率。在5个真实事件日志数据集上进行实验,结果表明,该方法相比Word2Vec+CNN模型、Transformer模型、BERT分类模型在准确率方面有较大提升,且与传统BERT+CNN相比,所提方法的准确率最高可提升2.61%。展开更多
基金the National Natural Science Foundation of China(Key Program)(52031004).
文摘The characterization techniques were employed like transmission electron microscope,X-ray diffraction and microstructural characterization to investigate microstructural evolution and impact of precipitate-phase precipitation on strength and toughness of a self-developed 32Si_(2)CrNi_(2)MoVNb steel during the quenching and tempering process.Research outputs indicated that the steel microstructure under the quenching state could be composed of martensite with a high dislocation density,a small amount of residual austenite,and many dispersed spherical MC carbides.In details,after tempering at 200℃,fine needle-shapedε-carbides would precipitate,which may improve yield strength and toughness of the steel.However,as compared to that after tempering at 200℃,the average length of needle-shapedε-carbides was found to increase to 144.1±4 from 134.1±3 nm after tempering at 340℃.As a result,the yield strength may increase to 1505±40 MPa,and the impact absorption energy(V-notch)may also decrease.Moreover,after tempering at 450℃,thoseε-carbides in the steel may transform into coarse rod-shaped cementite,and dislocation recoveries at such high tempering temperature may lead to decrease of strength and toughness of the steel.Finally,the following properties could be obtained:a yield strength of 1440±35 MPa,an ultimate tensile strength of 1864±50 MPa and an impact absorption energy of 45.9±4 J,by means of rational composition design and microstructural control.
文摘业务流程合规性检查可以帮助企业及早发现潜在问题,保证业务流程的正常运行和安全性。提出一种基于改进BERT(Bidirectional Encoder Representations from Transformers)和轻量化卷积神经网络(CNN)的业务流程合规性检查方法。首先,根据历史事件日志中的轨迹提取轨迹前缀,构造带拟合情况标记的数据集;其次,使用融合相对上下文关系的BERT模型完成轨迹特征向量的表示;最后,使用轻量化CNN模型构建合规性检查分类器,完成在线业务流程合规性检查,有效提高合规性检查的准确率。在5个真实事件日志数据集上进行实验,结果表明,该方法相比Word2Vec+CNN模型、Transformer模型、BERT分类模型在准确率方面有较大提升,且与传统BERT+CNN相比,所提方法的准确率最高可提升2.61%。
