A Localized Road Roughness Classification Algorithm Based on ISO 8608 Using Continuous Wavelet Transform

Abstract

In vehicle dynamics, road roughness is a primary dynamic excitation source that induces the vertical dynamic behavior of vehicles. The existing International Organization for Standardization (ISO) 8608 standard classifies road profiles into specific classes by deriving the overall Power Spectral Density (PSD) through the Fourier transform. However, this approach has the limitation of being unable to reflect the localized characteristics of road surfaces that change with location, as spatial variations in roughness are averaged across the entire section. To overcome this limitation, this study proposes a location-based localized PSD calculation technique using the Continuous Wavelet Transform (CWT). After applying the wavelet transform to 1D road elevation data, the degree of roughness for each location is calculated through log-domain integral averaging. The resulting localized degree-of-roughness coefficient is then mapped to continuous and discrete class indices. The proposed technique was demonstrated to capture spatial variations in localized roughness through both numerically synthesized scenarios and real-world measured road profiles.