신경망 함수를 이용한 자동차강의 변형저항 개발 및 압연하중 예측

Development of Flow Stress equation of High strength steel for automobile using Neural Network and Precision Roll Force Model

The flow stress value was calculated by comparing predicted and measured roll force. Using basic on-line roll force model and logged mill data the flow stress equation of high strength steel for automobile was derived. The flow stress equation consists of the flow stress equation of carbon steel and flow stress factor calculated by neural network with input parameters not only carbon contents, strip temperature, strain, and strain rate, but also compositions such as Mn, p, Ti, Nb, and Mo. Using the flow stress equation and basic roll force model, precision roll force model of high strength steel for automobile was derived. Using test set of logged mill data the flow stress equation was verified.