Research on Evaluation Technology of Pavement Roughness Based on Smartphone

Based on the principle of membership function in fuzzy theory, this research established the prediction model between smartphone acceleration value and pavement roughness level under different detection speeds after processing the sensitive data by detrending term, filtering discrete outliers, and data compensation. Finally, the prediction error of the model was analyzed. The research results show that the membership function can resolve the ambiguity between the pavement roughness level and the acceleration value, with high prediction accuracy, and strong generalization ability, which can accurately reflect the relationship between the acceleration value and the roughness level.