Multivariate information fusion method and application of two-phase flow based on complex network and deep learning
A complex network and deep learning technology, applied in the field of dual-modal multi-information fusion, can solve the problem that the classification effect cannot be very accurate, and achieve the effect of high sensitivity
Active Publication Date: 2018-10-30
TIANJIN UNIV
View PDF2 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, most of the training models in the past were shallow, and the classification effect could not be very accurate.
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0043] The multi-information fusion method and application of the two-phase flow based on the complex network and deep learning of the present invention will be described in detail below in combination with the embodiments and the accompanying drawings.
[0044] The multi-information fusion method and application of two-phase flow based on complex network and deep learning of the present invention proposes a method for building a network based on a complex network of correlation coefficients and using a deep learning model, namely a convolutional neural network, to achieve dual-mode multi-information fusion. By establishing a complex network of weighted correlation coefficients and a complex network of unweighted correlation coefficients, the multivariate information measured by the dual-mode sensor, that is, the multi-channel data, is fused to extract the node weighted aggregation coefficient, node weighted degree, node aggregation coefficient, node degree, node Complex networ...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
A two-phase flow multi-information fusion method based on complex network and deep learning and its application: construct a complex network for the multi-channel time series obtained by cyclically excited dual-mode sensors; for each multi-channel time series obtained by cyclically excited dual-modal sensors Establish weighted correlation coefficient complex network and unweighted correlation coefficient complex network respectively in time series, obtain network indicators, and combine the network indicators into a feature vector; deep learning model training and phase containment measurement, obtain a dual-modal A deep learning model for multivariate information fusion. The vertical oil-water two-phase flow experiment was carried out by using a circularly excited dual-mode sensor composed of a circularly excited multi-electrode conductivity sensor and a circularly excited multi-electrode capacitive sensor. The ratio of the oil phase and the water phase was fixed, and the flow rate of the oil phase and the water phase was changed experiment. The invention can realize the measurement of the two-phase fluid phase holdup.
Description
technical field [0001] The invention relates to a dual-mode multiple information fusion method. In particular, it involves a two-phase flow multi-information fusion method and its application based on complex networks and deep learning. Background technique [0002] Oil-water two-phase flow exists widely in oil exploration and transportation industries. In the oil-water two-phase flow system, the distribution of each phase changes with time and space, forming different flow patterns, which are called flow patterns. The flow pattern of two-phase flow is complex and changeable, and it is difficult to accurately capture local flow information, which makes the measurement of two-phase flow flow parameters such as phase holdup present many difficulties. This has caused many impacts on oil extraction and process transformation. At present, the research on flow pattern mainly adopts observation method, wavelet feature analysis and fuzzy C clustering, fuzzy logic and genetic algo...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06N3/08G06K9/62
CPCG06N3/08G06N3/084G06F18/25
Inventor 高忠科杨宇轩党伟东董长松蔡清
Owner TIANJIN UNIV