Dujiangyan irrigation system optimization (DISO): A novel metaheuristic algorithm for dam safety monitoring

Abstract

This work utilizes a novel metaheuristic algorithm named Dujiangyan Irrigation System Optimization (DISO) to resolve the dam deformation prediction problem. The Dujiangyan irrigation system (DIS) served as the model for the DISO algorithm, which has good metaheuristic algorithmic properties to handle various issues. This study utilizes the Fish Mouth project, the Baopingkou project, and the DIS's water separation and sand discharge principle. With the help of a mathematical model and implementation, DISO can perform the optimization process in various search spaces. Two comparative experiments with various characteristics are performed to evaluate the effectiveness of the DISO algorithm. 29 CEC 2017 benchmark test functions, 4 high-constrained test functions, and 2 typical engineering design challenges are chosen in the first step. 14 advanced heuristic techniques are chosen from the literature, and the statistical outputs of these techniques are examined and compared. Three nonparametric statistical tests are performed to confirm the proposed algorithm's effectiveness. The results demonstrate that the DISO algorithm outperforms all the evaluated algorithms. In the following stage, the proposed algorithm is applied to the dam deformation prediction problem, and a KICA-DISO-SVM spatial–temporal monitoring model is established for dams. The findings of the studies indicate that the proposed method can effectively monitor the influence of dam displacement on dam operation. The proposed model overcomes the inability of the standard single measurement point mode to capture the intrinsic relationship of the spatial deformation field accurately. The suggested DISO technique offers encouraging outcomes in resolving complex optimization problems.