Hot-rolled steel mechanical property machine learning method guided by physical metallurgy

The invention discloses a hot-rolled steel mechanical property machine learning method guided by physical metallurgy, and belongs to the crossing field of steel plate production and data statistical modeling, and the method comprises the steps: collecting hot-rolled steel plate data, and carrying out the preprocessing of the data; calculating physical metallurgical parameters based on steel plate data and a physical metallurgical theory; and optimizing parameters of the mechanical property calculation model by adopting a particle swarm optimization algorithm, and constructing a mechanical property prediction model. According to the hot-rolled steel mechanical property machine learning method guided by physical metallurgy, a high-quality data set is established, an optimal algorithm is selected for modeling to predict the mechanical property, and steel plate production is guided. 本发明公开了一种物理冶金指导的热轧钢材力学性能机器学习方法，属于钢板生产和数据统计建模的交叉领域，包括：采集热轧钢板数据并对数据进行预处理；基于钢板数据和物理冶金学理论计算物理冶金参数；采用粒子群优化算法优化力学性能计算模型参数，构建力学性能预测模型。本发明采用上