Multi-Index Fusion Debris Flow Early Warning Model Based on Spatial Interpolation and Support Vector Machine

• Published in: Water (Basel) Vol. 16; no. 5; p. 724
• Main Authors: Jin, Rui, Wang, Shaoqi, Liu, Jianfei
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 01.03.2024
• Subjects: Algorithms; Analysis; Artificial intelligence; debris flows; Disasters; early warning model; Emergency communications systems; Hydrology; Landslides & mudslides; Machine learning; Mountains; multi-index fusion; principal component analysis; Rain; Rain and rainfall; Rivers; Runoff; Sediment transport; spatial interpolation; support vector machine; Support vector machines; Valleys; China; Japan
• ISSN: 2073-4441; 2073-4441
• DOI: 10.3390/w16050724

Debris flow early warning is an effective method to prevent major disasters, so a multi-index fusion debris flow early warning model based on spatial interpolation and a support vector machine is designed. Aiming at the discrete rainfall data in the study area, the collaborative Kriging spatial interpolation method based on Kriging spatial interpolation is adopted to process the rainfall data into multi-index fused surface data. The rainfall data after spatial interpolation are used as the input sample data of the support vector machine early warning model, and the optimal parameters of the support vector machine are calculated by the sea squirt algorithm, and then the debris flow early warning results are output. After experimental analysis, the model can obtain rainfall surface data. After calculation by the model, the accuracy of the early warning probability of debris flow is improved, and the early warning result is consistent with the actual result of debris flow.