本文提供一种新的方法,即历史拟合的方法来反求地层参数和处理试井数据。把待定的地层参数作为控制变量来看待,当计算机模型上自动调节所得的地层参数值能使理论曲线与实测曲线达到最佳匹配时,此时的控制变量(如渗透率、窜流系数、弹性系数等)的数值就认为是地层的实际参数值。本文提供了求解试井问题的基本公式和算法,并结合算例对单一介质和双重介质系统运用微扰动和自动拟合方法作出了具体结果。这些初步结果表明,把试井问题作为一个可控系统的参数辨识问题来对待,可以运用已成熟的自控理论和最优化方法,较准确而有效地进行试井资料解释。 This article provides a new approach, history matching, to reverse formation parameters and process well test data. The formation parameters to be determined as a control variable to look at, when the computer model automatically adjusts the resulting formation parameters can make the theoretical curve and the measured curve to achieve the best match, the control variables (such as permeability, channeling coefficient, elastic Coefficient, etc.) is considered as the actual parameter value of the formation. In this paper, the basic formulas and algorithms for solving well testing problems are provided. Concrete examples are given to illustrate the application of single perturbation and auto-fitting methods to single media and dual media systems. These preliminary results show that the well test problem can be treated as a parameter identification problem of a controllable system. Well-established self-control theory and optimization methods can be used to explain the well test data more accurately and effectively.