BP算法在辽宁观音阁水库枯季地表入库径流实时预报中的应用

Application of BP Network in Real-time Forecast of Drought Period Surface Runoff in Guanyinge Reservoir of Liaoning

根据太子河流域降水和产汇流特点 ,利用流域试点区观音阁水库枯季入库径流量、上游汛期降水量、汛末入库径流量和枯季降水量等实测系列资料 ,对所建立的BP神经网络模型进行训练、检验和对比分析 ,结果表明 ,所建立的枯季径流BP预报模型是合理可行的 ,并且具备较高的精度 。

According to the characteristics of rain and runoff of Taizihe River Basin in Liaoning, a BP network model was set up to forecast the seasonal variation of flow of Guanyinge Reservoir. The calculation results showed that the BP network model was suitable and its degree of accuracy was satisfactory.