Aiming at the problem of reliable evaluation and real-time dynamic control of compaction effect of roller compacted concrete(RCC)dam on site, a new evaluation method for compaction degree of RCC thermal layer based on Rayleigh wave velocity detection is proposed. Relying on Chebyshev type I high-pass filtering combined with wavelet threshold signal denoising method, a regression prediction model for compactness of grid surface unit sample points is established, and the entire construction area visualization evaluation of compaction quality is performed by Kriging space interpolation method. Combined with an engineering field application example, the results show that the joint denoising method can effectively suppress the interference noise at the construction site, and the calculated wave velocity value is evenly distributed. In terms of goodness of fit and prediction accuracy, the multivariate nonlinear regression model based on Rayleigh wave velocity and moisture content performs better than other comparison models. The Kriging space interpolation method is used to perform regional optimization assignments to achieve real-time dynamic control of the quality defects for the entire construction area.