The elevated concentration of airborne particulate matter on underground metro platforms, far higher than in aboveground environments, has drawn increasing public attention. This study builds on long-term particulate matter measurements from three underground metro platforms in Stockholm, introducing a platform-specified multi-factor model that improves on the previous single-station model. By incorporating platform-specific data, this new model enhances applicability across multiple stations and analyses the effects of train frequency, train type, passenger flow, and urban background PM on PM1, PM2.5, and PM10 concentration changes. Comparative analysis shows the new model has better flexibility, accuracy, and stability. The results reveal significant interactions between variables, with platform-specific differences in how these factors influence PM concentration rates. Model estimations indicate that as new train types replace older ones, PM concentrations will vary across platforms but are expected to approach the recommended levels set by Stockholm's transport authorities.