广义回归神经网络在检测钢淬火冷却曲线数据处理上的应用

Application of GRNN to testing steel quenching cooling curve data processing

广义回归神经网络是建立在数理统计基础上的径向基函数网络,具有很强的非线性映射能力和学习速度,特别适合于各种检测数据的处理和预测。淬火介质的冷却效果和钢铁材料的淬火冷却行为的测定一般通过冷却曲线测定仪来获得一组时间、温度相关的数据,这些数据量长短不一,存储整理不方便且不利于进一步的对比分析应用。通过将原有杂乱无章的数据处理成按固定时间间隔的标准化数据,并将此类标准化数据汇集成数据库便于后续的分析、对比和应用。

Generalized regression neural network(GRNN)is a radial basis function network based on mathematical statistics.It has strong nonlinear mapping ability and learning speed,and is especially suitable for processing and forecasting various detection data.The measurement of cooling effect of quench medium and quenching cooling behavior of iron and steel materials is generally obtained by means of cooling curve tester to obtain a set of time and temperature related data.The data volume vary in length,which is inconvenient for storage and sorting and is not conducive to further comparative analysis and application.By processing the original chaotic data into standardized data at a fixed time interval,this kind of standardized data is collected into a database for subsequent analysis,comparison and application.