Research on the Marine Corrosion Model of Ferrous Metal Based on PLS and Grey Relational Analysis

The marine corrosion of metal materials is a complex chemical process which is affected by multiple factors that are nonlinear interconnected. This paper aims at integrating PLS and Grey theory in the model building of corrosion time， rate，maximum corrosion depth and average corrosion depth based on small sample data measured in a certain marine environment. The result shows that the time of data，temperature，salinity，dissolved oxygen，ph value and humidity are strongly correlated with the maximum corrosion depth in the full immersion zone. The fitting effects of PLS modeling based on grey relational analysis are better than the PLS model and the GM (1, 1) model in both precision and stability.