Well-testing stage analysis is an important way for oilfield operation state decision-making and reservoir management. However, due to the variability and nonlinearity of the downhole data caused by the complex exploration activities and the differences of petroleum type and geological conditions, the classical methods are ineffective in feature extraction, learning network construction, and classifier optimization. In this work, we propose a new well-testing stage classification method based on a deep vector learning model (DVLM). The novelty of this study lies in the combination of multi-feature extraction, deep learning and feature vector mapping. The proposed method can overcome the problems of poor feature representation ability and poor classification model generalization ability in the existing machine learning methods, which mainly caused by the non-optimized training network structure and the unreasonable classifier design,. Firstly, the initial features are obtained by four classical methods. Then a five layers deep belief network embedded with the mutual information coefficient method is implemented for further feature extraction and purification. Finally, the optimized learning vector quantization classifier outputs the predicted tags. For model training and testing 572 field samples total of 4004 data streams are used. By considering the classification errors and accuracy metrics, the neurons number of deep learning network and the classifier are tuned, and an optimal and stable framework is obtained. Comparative experimental results with several classical integration models show that the proposed model achieves the highest classification accuracy of 98.065% as well as the least of features (nine). The results demonstrate that the proposed model has excellent performance in improving the classification accuracy and completing the feature compression. Moreover, the proposed model has very important practical significance for guiding the automatic analysis and processing oil and gas data.
In the spectral analysis of a complex solution, the model migration ability is a key challenge. The influence of changes in external factors (measurement environmental changes) is the main reason for the poor model migration ability. If this influence is not well suppressed, it may affect the model migration ability and reduce the robustness of the model. Therefore, effectively suppressing this influence can enhance the model migration ability and improve accuracy. In this paper, we employ the spatial and time domains in addition to the wavelength domain and propose a high-spatial-resolution hyperspectral method that transforms the information of the external factor changes in the spectrum into DC (direct-current) components in the spatial and time domains. Consequently, the information of the external factor changes is separated from the effective information of the tested substance. An algorithm is proposed to extract parameter spectra that are only related to the information of the tested substance. Then, the traditional transmission spectra are replaced by the extracted parameter spectra in the modeling. We design an experiment to spectrally analyze hemoglobin (HB), and two kinds of measurement containers are used to simulate a type of external factor. To verify the effectiveness of the proposed method, we determine whether the model can be migrated to an environment with a different measurement container. The experimental results show that the prediction accuracy for HB and the model migration ability are satisfactory. The results indicate that the high-spatial-resolution hyperspectral method can effectively suppress the influence of external factor changes on spectral analysis and improve the model migration ability.