부산대학교 도서관

Material corrosion is a serious threat to power apparatus, equipments, buildings and personal safety, and brings huge economic losses and a large amount of resources and energy consumption to all countries in the world. Corrosion data is difficult to be effectively treated by traditional data analysis methods due to its large quantity, large structural differences and uneven quality. The emergence of data mining technology plays a crucial role in analyzing material corrosion data, building corrosion models, studying corrosion laws and predicting the long-term development trends of corrosion. Firstly, the basic concept of data mining is introduced, and several commonly used techniques and applications of data mining are summarized. On this basis, combined with the characteristics of material corrosion data, the application and progress of data mining in building of corrosion model, analysis of the relationship between environmental factors and corrosion rate, as well as prediction of corrosion rate are discussed. The application of various methods in corrosion data mining is summarized. Finally, the limitations of data mining in corrosion data analysis are pointed out, and prospects are given.