Sichuan Knowledgeable Intelligent Sciences International Scientific Technical and Economic Research 2959-1309 4 2 2026 05 29 Comparison and Analysis of Multiple Entropy Feature Tests for Abnormal Signals in Fixed-Point Deformation Monitoring 206 231 10.71451/ISTAER2622 eng Sirui Liu Institute of Seismology, China Earthquake Administration 0009-0002-0524-4947 Cong Pang Institute of Seismology, China Earthquake Administration, Wuhan, Hubei, China and Wuhan Gravitation and Solid Earth Tides, National Observation and Research Station, Wuhan, Hubei, China 0009-0001-8579-7527 Xin Wang Shenyang Earthquake Monitoring Center, Liaoning Earthquake Administration, Shenyang, Liaoning, China 0009-0002-8056-182X Meiping Song Datong Earthquake Monitoring Center, Shanxi Earthquake Administration, Datong, Shanxi, China 0009-0000-9529-9778 2026 03 26 2026 04 28 Fixed-point deformation monitoring is essential for geological hazard early warning, and identifying abnormal signals remains a key challenge. To evaluate the effectiveness of five entropy methods for this purpose, a quantitative comparison was conducted. Abnormal deformation data were preprocessed to extract sample entropy (SE), fuzzy entropy (FE), distribution entropy (DE), permutation entropy (PE), and adaptive weighted multi-scale fusion entropy (AWM-FE). Their ability to distinguish abnormal signals was compared using the T-test, followed by the Kruskal‑Wallis test and post hoc multiple comparisons. Simulation experiments showed that AWM-FE exhibited stable, reliable performance and was well-suited for complex field environments with multi-scale analysis needs. Real deformation data analysis revealed that SE had an average T-test p-value of 0.0016, indicating significant distinction across six category pairs, while DE achieved an F-value of 74.7205 in ANOVA, reflecting the largest overall inter-group variation. This study provides a reference for feature selection in identifying abnormal signals in fixed-point deformation observations.