Apparent Characteristics Analysis and Development Trend Prediction of Dagangshan High Arch Dam
To evaluate the operation state of the high arch dam of Dagangshan Hydropower Station,based on the field monitoring data(stress,temperature,and displacement),the correlation of the apparent characteristic data was deeply analyzed,and three deep learning models(ESN,ELM,BI-LSTM)combined with the optimization algorithm SSA(Salp Swarm Algorithm)were used to make a prediction comparison.The results show that when the water level changes,the stress response of the No.14 dam section is the most severe,and the displacement and temperature changes are within the normal range.The optimal prediction models for positive vertical displacement and stress are ELM and BI-LSTM.The root means square errors were 0.985 and 0.061,respectively.The study can provide reference for prediction of dam auto-mation monitoring data.
high arch damapparent characteristicdeep learningpredictionoptimization algorithm
万方数据
维普
