中文摘要：

在污染源下游布设水质在线监测断面并对监测数据进行分析,通过求解河流污染源反问题可以实时反演识别上游污染源信息,从而为污染源的控制和取水口的水质安全保障及预警提供基础信息。针对一维河流,在非恒定水流条件下,提出了以有限差分法作为正问题求解器、Nelder-Mead单纯形算法为优化模型求解器的污染源识别算法（FDM-SIM法）;在此基础上,进行了污染源和参数的联合反演尝试。结合两个算例对FDM-SIM法进行了验证,计算结果表明：该方法不仅能够对污染源进行识别,还可实现污染源与水质参数的联合反演。

英文摘要：

Pollutant sources can be identified from analysis of online monitoring data collected in a river reach at its downstream by solving inverse problems of the sources, which provides basic information for control of pollutant sources and protection of headwaters area. For unsteady flow in 1D channel, a new FDM-SIM method combining finite difference method with simplex algorithm is presented herein. This method was applied to determination of pollution sources and their parameters, and validated by two numerical cases. The results indicate that the method is very useful in identification of pollution sources and in coupled inversion calculation for coefficients of the sources and water quality.