An improved site characterization method based on interval type-2 fuzzy C-means clustering of CPTu data

Autor:	Jian-Fei Lu, Jie Yin, Leslie Opoku, Ping-ping Zuo, Yang Yang, Yong-hong Miao
Rok vydání:	2021
Předmět:	business.industry Pattern recognition Interval (mathematics) Fuzzy logic Stratigraphy Outlier Cluster (physics) General Earth and Planetary Sciences Soil horizon Artificial intelligence business MATLAB Cluster analysis computer General Environmental Science Mathematics computer.programming_language
Zdroj:	Arabian Journal of Geosciences. 14
ISSN:	1866-7538 1866-7511
DOI:	10.1007/s12517-021-07555-9
Popis:	This paper presents an improved method to identify the soil strata using the interval type-2 fuzzy C-means (IT2FCM) clustering algorithm representing the uncertainty in CPTu data. Representative parameters (i.e., cone resistance qt and pore pressure ratio Bq) chosen from CPTu data obtained from two engineering sites were normalized and thereafter processed by the self-written IT2FCM subprogram in MATLAB. The cluster analysis results indicated that the IT2FCM method can identify similar groups of data in the soil profile, detect different layer boundaries and transitions, and delineate the lenses and outliers within a sub-layer. Compared with the exiting cluster method, the method proposed in this study uses fewer clustering numbers, and the cluster results can discover more drastic changes within the soil strata. Combined with the validity index, the cluster results are more reasonable in identifying the soil stratigraphy.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::98f9498a1a4244606e9f2bae9cc2a59e https://doi.org/10.1007/s12517-021-07555-9 Zobrazit plný text záznamu Full text from SpringerLink