中文摘要：

地下水污染源位置与强度的确定有助于提高地下水污染治理和修复的效果。本文在场地信息不确定的前提下，基于卡尔曼滤波技术与单纯形法提出一种地下水污染源识别的新方法。该方法利用卡尔曼滤波技术连续作用采样点来估计复合污染羽与误差协方差矩阵，进而利用误差协方差矩阵进行新采样点的选择；模糊集被用来进行污染羽的表示，通过复合污染羽与单个污染羽的形态对比以确定污染源位置；在进行污染源位置反演时，嵌入单纯形法以进行污染源强度的同步反演。算例研究表明，该方法在判断污染源位置时考虑了污染羽的整体形态，从而降低了场地局部信息不确定对于污染源识别结果的影响，并能够通过合理的采样点的选择，正确地识别出污染源位置与强度，反演结果具有较高的可靠性。

英文摘要：

Identification of the location and intensity of groundwater pollution source is contributive to the effect of pollution remediation. In this study, a new approach to identify the groundwater pollution source is proposed based on the Kalman filtering and simplex method with the uncertainty of fields. The general pollution plume and covariance matrix of error are predicted through continuous sampling with Kalman filtering. Afterwards, the sampling points are selected in combination with the covariance matrix, to reduce the uncertainty as far as possible. The pollution plume is represented by fuzzy set, and the pollution location is identified by the comparison of general and single plume. The simplex method is embedded in the inversion of source location to reverse source intensity. The case study shows that this approach give sufficient consideration to the overall shape of pollutants in order to reduce the influence of the uncertainty of local information on the recognition results, and the approach is an effective way to identify the location and intensity of pollution source through the reasonable sampling points, providing the inversion with higher reliability.