公开(公告)号：US20230289607A1

公开(公告)日：2023-09-14

申请号：US18162618

申请日：2023-01-31

Applicant: TAIYUAN UNIVERSITY OF TECHNOLOGY

Inventor： Zhixiong ZHANG ,  Yongwei FU ,  Tao WANG ,  Xiaoyan XIONG ,  Zhongkai REN ,  Yuanming LIU ,  Tianxiang WANG ,  Xi LIAO

IPC: G06N3/084 ,  G06N3/063

CPC classification number: G06N3/084 ,  G06N3/063

Abstract: Disclosed is a neural network model optimization method based on an annealing process for a stainless steel ultra-thin strip, which belongs to the technical field of data analysis. Aiming at the defects of a size effect caused by a difference of the stainless steel ultra-thin strip from a macroscopic size, a poor adaptability of a common stainless steel mechanical property prediction method and the like, and a good capability of a neural network for processing a complex nonlinear problem, the method comprises modeling various factors influencing annealing of the stainless steel ultra-thin strip by utilizing an artificial neural network, thus predicting and controlling a mechanical property and a microscopic structure after annealing.