基于BP神经网络的泉州市山美水库降雨径流模拟研究

RAINFALL-RUNOFF SIMULATION OF SHANMEI RESERVOIR IN QUANZHOU CITY ON BP NEURAL NETWORKS

考虑到降雨-径流的非线性关系的复杂性,同时保证较高的计算时间效率,本研究采用率定-验证同步进行的BP神经网络模型,以降水量、时间、气温、风速、辐射量等5个因子作为输入层变量,以实测流量为训练数据,对泉州市山美水库来水量进行月时间尺度的模拟.并与SWAT分布式水文模型的模拟结果进行比较,以此评价BP神经网络模型模拟降雨径流过程的精确度与可行性.研究结果表明,BP模型模拟结果的误差均在允许范围内且精确度较高,适用于山美水库来水量预测研究.

Considering the complex nonlinearity of rainfall-runoff relationship and computation speed, a back-propagation （BP） artificial neural networks （ANN） model was proposed to simulate inflow of the Shanmei Reservoir in Quanzhou City. The input layer data included rainfall, time, air temperature, wind speed and solar radiation. The in situ observed inflow data were used as training data. To explore the feasibility of the BP-ANN model, a comparison was made with the SWAT hydrological model. It was found that the BP-ANN model could make better simulation with satisfactory efficiency, and the BP-ANN model could be used to forecast inflow of the Shanmei Reservoir.