Power grid equipment in southern coastal regions faces severe corrosion due to the high salt content, humidity, and intense ultraviolet radiation characteristic of the maritime climate. This study focuses on developing a digital corrosion map for grid materials in typical southern regions, utilizing machine learning and data analytics to model the relationships between environmental factors and corrosion rates. Dimensionality reduction techniques, including grey relational analysis, maximum correlation minimum redundancy (mRMR), and principal component analysis (PCA), were employed to identify critical environmental factors from 14 candidates. Using K-nearest neighbors (KNN) and support vector regression (SVR) algorithms, we built semi-supervised regression models to predict corrosion. ArcGIS and kriging interpolation were then applied to visualize the spatial distribution of corrosion levels. Results indicate that UV radiation, temperature, and humidity are crucial factors affecting coastal equipment corrosion, with UV having a particularly strong influence on galvanized steel degradation. This research provides data-driven insights to support corrosion monitoring and intelligent protection for power grid equipment in coastal areas.