Abstract: The traditional quality assessment method of earth-rock dam compaction is primarily based on control parameters collected from the job site of compaction and by sampled spots testing. Compaction parameters cannot satisfy the requirement by every location of the entire working surface, and using the same parameter may lead to different compaction quality due to the influence of inhomogeneity in material property. A limited number of test samples are usually an unreliable representation of compaction quality over the entire surface layer, and a certain error occurs when compaction value (CV) alone is used in the evaluation. This paper describes a fuzzy comprehensive compaction quality evaluation method based on the Dempster-Shafer evidence theory and dynamical monitoring of compaction parameters and CV. We also develop a concept of equivalent continuous weak area using the density-based spatial clustering of applications with noise (DBSCAN) to ensure that no obvious continuous weak area occurs in the compacted layer, and apply it to quantitative evaluation of the spatial distribution of weak areas. Case analysis shows that our evaluation method can effectively avoid one-sidedness in evaluating compaction parameters and enhance the comprehensiveness of compaction quality assessment, thereby providing a new way for quality control in earth-rock dam construction.

Key words: earth-rock dam construction, compaction quality, fuzzy comprehensive evaluation, D-S evidence theory, spatial clustering analysis