Levy flight-improved grey wolf optimizer algorithm-based support vector regression model for dam deformation prediction

Considering the strong non-linear time-varying behavior of dam deformation, a novel prediction model, called Levy flight-based grey wolf optimizer optimized support vector regression (LGWO-SVR), is proposed to forecast the displacements of hydropower dams. In the proposed model, the support vector regression is used to create the prediction model, whereas the Levy flight-based grey wolf optimizer algorithm is employed to search the penalty and kernel parameters for SVR. In this work, a multiple-arch dam was selected as a case study. To validate the proposed model, the predicted results of the model are compared with those derived from Grid Search algorithm, Particle Swarm Optimization, Grey Wolf Optimizer algorithm, and Genetic algorithm. The results indicate that the LGWO-SVR model performs well in the accuracy, stability, and rate of prediction. Therefore, LGWO-SVR model is suitable for dam engineering application.

Automated summary

A novel prediction model, LGWO-SVR, is proposed for forecasting dam displacements, which combines support vector regression and Levy flight-based grey wolf optimizer. The model is validated with multiple-arch dam as a case study and compared with other algorithms. Results show that the LGWO-SVR model has high accuracy, stability, and prediction rate, making it suitable for dam engineering applications.