【作者】：瞿云华1.2 常锦才2 张玉柱*.2 (1.东北大学，东北 沈阳110000；2.华北理工大学，河北 唐山 063000)

【内容提要】：本文利用BP神经网络对热轧螺纹钢的各种微量元素与力学性质的线性或非线性表达式，探究各个微量元素含量比例为何值时，热轧带肋钢筋性能达到优化，再将其数值指导生产实践，不断反复验证理论与生产实践，最终得到微量元素与钢筋性能之间的生产实践与理论计算的差异仅为6%左右。

【关键词】：BP神经网络；螺纹钢性能；螺纹钢组成元素；数值计算与模拟

本文受河北省钢铁联合基金项目：《钢铁物流网络区间规划设计及物流园区SLP布置的研究》课题资助，项目号：E2016209304。

Research on Dependence of Rebar Performance and Element Content Based on BP Neural Network Model

Quyunhua1.2Changjincai2Zhangyuzhu*.2 (1. Northeastern University 2. North China University of Technology, Hebei Tangshan 063000)

In this paper, using the BP neural network, the linear or non-linear expression which describe the relation of various trace element contents and mechanical properties of hot-rolled rebar were applied to find the optimal value of element contents when the performance of hot rolled ribbed steel is the best. Then the numerical values were applied to guide the production practice, and the theory and production practice were repeatedly. Ultimately, the difference between the production practice and the theoretical calculation between the content of trace elements and the properties of steel bars is only about 6%. 

key words： BP neural network; properties of rebar; composition of rebar; numerical calculation and simulation

注：本文发表于《铸造技术》201812