541 / 2024-07-11 10:40:06

Monthly runoff simulation based on BP-ANN and correlation analysis

runoff, simulation, BP-ANN, correlation analysis

Subtheme 2: Integrated Water Resources Management > Topic 1: Water Resources Management system, policy and regulation

Accurate inflow forecasting plays a very important role in the safe operation and optimal operation of reservoirs. In this paper, correlation analysis is used to select predictors, and a three-layer BP-ANN model is constructed. The model is applied to the mid-long term runoff forecast of Guan Yin Yan hydropower station dam site. The following conclusions are shown: (1) the results of correlation analysis show that Q_{i, j} has a strong correlation with Q_{i, j-1}, Q_{i, j-2}, Q_{i-1, j} and Q_i-2, _j, which can be used as the main predictor of the model. (2) the trained model is applied to the monthly runoff simulation of Guan Yin Yan Reservoir dam site. The simulation results show that the model has a good simulation effect during the training period and the test period, and the error meets the requirements of runoff simulation. It can be popularized and applied in monthly runoff simulation.(3) the NSE and R² in the test period were 0.86 and 0.93, respectively, which were greater than 0.82 and 0.91 in the training period, reflecting the strong generalization ability of the model.