Research on the Prediction Method of Tire/Pavement Noise Based on Asphalt Pavement Texture
Tire/Pavement Interaction Noise(TPIN)has become an important aspect of urban environmental noise pollution,and pavement texture is a significant influencing factor of TPIN.Three-dimensional laser scanning technology was utilized to acquire pavement point cloud data,create a three-dimensional model of pavement texture,and analyze the pavement texture by computing overall,macro,and micro texture indices using the gray level covariance matrix(GLCM),fractal theory(FD),and power spectral density theory(PSD).Subsequently,a linear regression and random forest prediction model of TPIN was developed based on the texture index.The results show that:1)The texture metrics calculated by the three theories can effectively deconstruct the texture and have good correlation with TPIN;2)The correlation between TPIN and micro-texture is stronger when the vehicle speed is at 60 km/h,and the critical wavelength of the pavement is related to the tire wavelength of the noise test;3)The prediction accuracy of random forest is higher than that of a one-factor prediction model,and the optimal prediction accuracy is R2=0.988 5,MSE=0.000 4.The prediction effect is better.Therefore,pavement texture can be used as a method to effectively evaluate TPIN.
NETL
NSTL
万方数据
