Prediction of long-term maximum settlement deformation of concrete face rockfill dams using hybrid support vector regression optimized with HHO algorithm

• Published in: Journal of civil structural health monitoring Vol. 13; no. 2-3; pp. 371 - 386
• Main Authors: Li, Yanlong, Yin, Qiaogang, Zhang, Ye, Qiu, Wen
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.03.2023; Springer Nature B.V
• Subjects: Aging; Algorithms; Artificial intelligence; Civil Engineering; Concrete; Control; Dams; Deformation; Deformation effects; Design optimization; Dynamical Systems; Engineering; Measurement Science and Instrumentation; Neural networks; Optimization; Optimization algorithms; Original Paper; Parameter sensitivity; Prediction models; Rockfill dams; Sensitivity analysis; Support vector machines; Vibration; Water levels; Long-term settlement prediction; Monitoring model; HHO; SVR; Dam health monitoring
• ISSN: 2190-5452; 2190-5479
• DOI: 10.1007/s13349-022-00641-w

Excessive settlement deformation is one of the primary defects of concrete face rockfill dams (CFRDs). Intelligent monitoring and prediction of settlement deformation is a crucial aspect of dam health monitoring (DHM); however, the design of prediction models is a subsisting challenge. This paper presents an innovative algorithm combining the Harris hawk optimization (HHO) algorithm and support vector regression (SVR) for DHM modeling and sensitivity analysis of the parameters for optimizing the HHO algorithm. An analysis of measured data of different subarea monitoring points of the Jishixia CFRD proved that the monitoring model developed can effectively simulate the settlement deformation of the dam. Finally, this model was used to predict the long-term maximum settlement of Jishixia Dam. The prediction results showed the long-term maximum settlement value of the dam to be 503 mm, accounting for only 0.488% of the maximum dam height, which is comparatively low. The internal settlement deformation of the dam tended to stabilize with time, indicating that its long-term operation was safe and reliable. This model provides a new method for predicting and analyzing the long-term maximum settlement of dams, and can also provide a reference for deformation prediction modeling of other structures.