基于LSTM的三峡水库短期上游水位预测方法研究

Study on short-term upstream water level prediction method of Three Gorges Reservoir based on LSTM

为精确预测三峡水库上游水位短期变化过程,保障三峡工程效益发挥和稳定运行,分析了三峡水库上游水位预测面临的技术难题,指出了基于水量平衡原理的传统水位预测方法的弊端,提出了基于长短时记忆(LSTM)的三峡水库短期上游水位预测方法。模拟计算结果表明:该方法考虑了三峡动库容、调峰及入库流量计算密度不够等方面的影响,在复杂情况下,水位过程预测精度较高,实现了三峡水库短期上游水位连续24 h的逐时准确预测。研究成果可为水库精细化调度提供可靠的理论支撑。

In order to accurately predict the short-term upstream water level change process of the Three Gorges Reservoir and ensure the benefits and safe operation of the Three Gorges project, this paper analyzes the technical problems faced by the upstream water level prediction of the Three Gorges Reservoir and the defects of the traditional water level prediction method based on the principle of water balance and puts forward the short-term upstream water level prediction method of the Three Gorges Reservoir based on LSTM.The simulation results show that this method can consider the influence of the dynamic storage capacity of the Three Gorges Reservoir, peak shaving, and insufficient calculation density of inflow flow And the prediction accuracy of the water level process under complex conditions is high, and the accurate hourly prediction of water level in the upper reaches of the Three Gorges is realized The research results can provide reliable theoretical support for the refined operation decision of the reservoir.