Two-phase fluid flow pattern identification method based on time sequence and neural net pattern identification
A neural network and time series technology, applied in flow characteristics, measurement devices, instruments, etc., can solve the problems of low imaging accuracy, difficult measurement of small signals, poor real-time performance, etc., to improve accuracy, fast learning speed, pattern recognition and The effect of strong classification ability
Patent Information
- Authority / Receiving Office
- CN · China
- Current Assignee / Owner
- Publication Date
- 2005-09-07
- Estimated Expiration
- Not applicable · inactive patent
Abstract
Description
technical field
[0001] The present invention relates to a gas-liquid two-phase flow pattern identification method, in particular to a two-phase flow pattern identification method based on time series and neural network pattern identification, which can carry out online identification of gas-liquid two-phase horizontal pipe flow pattern recognition methods. It belongs to the technical field of multiphase flow measurement and control and data processing. Background technique
[0002] Two-phase flow flow pattern identification is the basis of two-phase flow scientific research. The determination of flow pattern provides reliable parameters for the design, operation and operation of two-phase flow systems and related equipment, as well as the calculation of pressure drop along the pipeline. The design, analysis and operation of the system are of great significance. In the past 30 years, many flow pattern recognition methods based on different models and theories have been prop...
Examples
Embodiment Construction
[0025] The specific implementation of the present invention will be further described below in conjunction with the real-time online flow pattern identification of oil-gas two-phase flow in horizontal pipes carried out on a large-scale multiphase flow experimental simulation device.
[0026] (1) Fix and change the liquid phase flow and gas phase flow respectively to obtain various flow patterns of different flow conditions, different gas and liquid flows: such as laminar flow, stirred flow, slug flow and annular flow. The differential pressure signal time series of the gas-liquid two-phase flow in the horizontal pipe is collected by a differential pressure sensor. The sampling point spacing is 10 times the pipe diameter, the sampling frequency is 200Hz, and the sampling time is 30 seconds.
[0027] (2) After the wavelet transform method is used to denoise the obtained time series signals, and through data processing, the probability density function curves of differential pres...