流域中修建的大量小型水库会改变洪水的自然特性，而小型水库资料往往缺失或信息不全，使得在洪水预报中难以考虑其调蓄作用，给洪水预报带来困难。利用遥感RS、地理信息系统GIS和支持向量回归机SVR算法，构建地形地貌参数-水库库容关系模型，推算水库资料缺失地区的小型水库库容信息；在定量概化小型水库对洪水拦蓄影响的基础上，建立了新安江－水库模型。以淮河上游北庙集流域为对象进行应用研究，结果表明：①小型水库总库容与地形地貌特征有较强的相关关系，建立的地形地貌参数-水库库容模型具有较高精度，可用于资料缺失的小型水库的库容推算；②加入水库拦蓄模块后，新安江模型的洪水预报精度得到一定提高。研究成果可供受小型水利工程影响地区的实时洪水预报参考。

A large number of small reservoirs built in the basin may change the natural characteristics of the flood， and the data of small reservoirs are often missing or incomplete， which makes it difficult to consider the regulation and storage function in flood forecasting， and brings difficulties to flood forecasting. Based on RS， GIS and SVR algorithm， the relationship model of topography and geomorphology parameters and reservoir capacity is constructed to calculate the small reservoir capacity information in the area where the reservoir data is missing. Based on the quantitative generalization of the impact of small reservoirs on flood detention， the XAJ-reservoir （XAJ-R） model is established. The results show that： ① there is a strong correlation between the total storage capacity of small reservoirs and the characteristics of topography and geomorphology， and the established model of topography and geomorphology parameters and reservoir storage capacity has a high accuracy， which can be used to calculate the storage capacity of small reservoirs with missing data. ② After adding the reservoir retention module， the flood prediction accuracy of XAJ model is improved to a certain extent. The research findings can be used as a reference for real-time flood prediction in the area affected by small-scale water conservancy projects.