明确水力压裂形成的裂缝参数对于指导压裂改造参数的设计和预测单井产量具有重要的意义。目前，已有的基于生产数据的裂缝参数反演方法难以获得各压裂段的裂缝参数。因此，借助集合卡尔曼滤波（EnKF）算法，建立了基于产气剖面测试的致密气藏各段裂缝参数自动反演方法。为兼顾模拟精度与反演效率，利用Matlab油藏数值模拟工具箱（MRST）建立了基于嵌入式离散裂缝的气藏生产数值模拟模型。模拟计算各段裂缝产量，并采用EnKF逐步更新各段裂缝半长与渗透率，实现基于产气剖面的致密气藏裂缝参数自动反演。最后，通过人造案例验证了该方法的可靠性，并将其用于矿场实际水平井的裂缝半长和渗透率的反演。研究结果表明：①裂缝方位和裂缝间距固定时，增加裂缝缝长和裂缝渗透率均能提高致密气产量，但二者对裂缝产量的影响程度随时间变化，裂缝渗透率对前3个月的产气量影响较大，而裂缝半长对生产中后期的产气量影响较大；②EnKF这种按时间顺序推进的数据拟合方法可以考虑不同时期裂缝半长和渗透率对产量的影响，在人造产气剖面拟合案例上，裂缝半长反演的相对误差不超过6.30%，裂缝渗透率反演的相对误差不超过0.88%；③针对致密气藏实际水平井的产气剖面，EnKF可同时反演出各压裂段的裂缝半长和裂缝渗透率，反演的裂缝半长与微地震监测半缝长相对误差低于8%。该方法可为致密气藏压裂裂缝诊断提供指导和参考。

Accurately determining hydraulic fracture parameters is crucial for guiding the design of fracturing treatments and predicting single-well production. Currently, existing production-data-based fracture parameter inversion methods struggle to obtain fracture parameters for individual fracturing stages. To address this, leveraging the ensemble Kalman filter (EnKF) algorithm, an automatic inversion method for fracture parameters of each stage in tight gas reservoirs based on gas production profile testing was developed. To balance simulation accuracy and inversion efficiency, a reservoir production numerical simulation model based on embedded discrete fractures was established using the MATLAB reservoir simulation toolbox (MRST). Subsequently, the production of each fracture stage was simulated, and the EnKF was employed to iteratively update the fracture half-length and permeability for each stage, achieving automatic inversion of fracture parameters in tight gas reservoirs based on the gas production profile. Finally, the reliability of this method was validated through a designed case study, and it was applied to invert the fracture half-length and permeability of a field horizontal well. The research results indicated that: (1) when fracture orientation and spacing were fixed, increasing both fracture length and permeability enhanced tight gas production, but their impact on fracture production varied over time. Fracture permeability significantly influenced gas production in the first three months, while fracture half-length had a greater effect on production during the middle and late stages. (2) EnKF, as a sequential data assimilation method, captured the influence of fracture half-length and permeability on production at different stages. In the designed production profile fitting case, the relative errors of inverted fracture half-length and permeability were below 6.30% and 0.88%, respectively. (3) Based on the gas production profile of an actual horizontal well in a tight gas reservoir, EnKF could simultaneously invert the fracture half-length and permeability for each stage, with the relative error of the inverted fracture half-length below 8% compared to microseismic monitoring results. This method provides valuable guidance and reference for diagnosing hydraulic fractures in tight gas reservoirs.