成分-结构-性能之间的关系始终是材料科学研究的主题,传统的试错法等经验或半经验的材料研究方法造成了资源、人力和时间上的极大浪费,因此需要从理论上解决材料设计、评价、预报等方面问题.人工神经网络是具有在线学习、记忆和分析推理功能能力的数学方法,它能够获得输入与输出之间的相互关系.其中BP神经网络结构简单、理论研究比较成熟.在材料研究领域中,BP神经网络已用于材料性能的研究与预测,复合材料工艺参数优化和预报,以及对金属的腐蚀研究等方面. The relationship between component-structure-property has always been the subject of material science research. The traditional or trial-and-error methods and other material-research methods have caused a great waste of resources, manpower and time, and thus need to be theoretically solved Material design, evaluation, forecasting, etc. Artificial neural network is a mathematical method with online learning, memory and analysis of reasoning capabilities, which can get the relationship between input and output.The BP neural network is simple in structure, theoretical research Maturity.In the field of materials research, BP neural network has been used in the research and prediction of material properties, the optimization and prediction of the technological parameters of composite materials, and the research on the corrosion of metals.