Impact Toughness Prediction for TMCP Steels Using Knowledge-based Neural-fuzzy Modelling

摘要: As one of the most important characteristics structural steels, toughness is assessed by Charpy V-notch impact test. The absorbed energy and transition temperature defined at a given level are regarded as common criteria for assessment. This paper aims establishing generic prediction models which link materials compositions processing conditions with properties. Hybrid knowledge-based neural-fuzzy modelling techniques incorporate linguistic knowledge into data-driven have been used to develop properties thermo-mechanical control process (TMCP) steels. Two basic ways incorporation discussed improve performance obtained fuzzy models. Simulation experiments show that both numeric data information can be combined in unified framework (ITT) predicted same model.