The precondition of well testing interpretation is to determine the appropriate well testing model. In numerous attempts in the past, automatic classification and identification of well testing plots have been limited to fully connected neural networks (FCNN). Compared with FCNN, the convolutional neural network (CNN) has a better performance in the domain of image recognition. Utilizing the newly proposed CNN, we develop a new automatic identification approach to evaluate the type of well testing curves. The field data in tight reservoirs such as the Ordos Basin exhibit various well test models. With those models, the corresponding well test curves are chosen as training samples. One-hot encoding, Xavier normal initialization, regularization technique, and Adam algorithm are combined to optimize the established model. The evaluation results show that the CNN has a better result when the ReLU function is used. For the learning rate and dropout rate, the optimized values respectively are 0.005 and 0.4. Meanwhile, when the number of training samples was greater than 2000, the performance of the established CNN tended to be stable. Compared with the FCNN of similar structure, the CNN is more suitable for classification of well testing plots. What is more, the practical application shows that the CNN can successfully classify 21 of the 25 cases.

Unstable water injection can effectively improve the recovery ratio of reservoirs with strong heterogeneity. However, the oil displacement mechanism and the determination method of unstable water injection parameters still need to be clarified, especially for complex fracture reservoirs, which greatly restrict the popularization and development of unstable water injection technology. This paper studies unstable water injection technology in highly heterogeneous reservoirs from the core and reservoir scales, utilizing many displacement experiments and numerical simulations. The differences in oil displacement efficiency, remaining oil distribution, pressure field, and streamlines between continuous water injection and unstable water injection are compared and evaluated. Five flow stages of unstable injection and production are precisely divided, and the microscopic and macroscopic displacement mechanism is clarified. A numerical model of two injection wells and one production is established to determine the best time to implement unstable water injection technology. Based on the principle of pressure superposition, the expression of pressure field distribution between injection and production well in each period of unstable water injection is analytically solved. This formula has provided a new development parameter optimization method aiming at the maximum pressure fluctuation range and optimized the development technology parameters in the water injection process. The results show that precise control injection and production parameters can expand water swept volume, effectively improve the degree of reserve utilization, and improve the recovery of complex reservoirs by 3–7%, which provides a reliable basis for the practical implementation of unstable water injection technology.