Unlock instant, AI-driven research and patent intelligence for your innovation.

An Active Modeling and Prediction Method for the Flow Rate of a Reciprocating Mixed Transport Pump

A prediction method, a technology of mixed infusion pump, applied in neural learning methods, biological neural network models, instruments, etc., to achieve the effect of easy implementation

Active Publication Date: 2021-10-26
ZHEJIANG UNIV OF TECH
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, a literature search found that the variation of GPR prediction uncertainty is used in the modeling method of actively obtaining training data, and has never been used in the study of flow rate characteristics of reciprocating mixed pumps

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • An Active Modeling and Prediction Method for the Flow Rate of a Reciprocating Mixed Transport Pump
  • An Active Modeling and Prediction Method for the Flow Rate of a Reciprocating Mixed Transport Pump
  • An Active Modeling and Prediction Method for the Flow Rate of a Reciprocating Mixed Transport Pump

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0055] Such as figure 2 As shown, an active modeling and prediction method for the flow rate of a reciprocating mixed pump of the present invention comprises the following steps:

[0056] (1) Determine the input and output variables of the model, and collect a small number of sample sets of typical working conditions;

[0057] In view of the fact that the air content rate is the most important factor affecting the flow rate, and considering the design parameters of the three-cylinder double-acting reciprocating mixed-transport pump prototype used in the experiment, without loss of generality, the inlet pressure P in this example s =0.4MPa, outlet pressure P d = 3.0MPa, air content rate β = 0.1, 0.3, 0.5, 0.7, 0.8, 0.9, crank angle θ = 0° ~ the end of the discharge process; in view of the flow rate of a conventional reciprocating pump can be determined by the flow rate of a pump chamber through theory Calculated by the formula, without loss of generality, the flow rate Q of ...

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

No PUM Login to View More

Abstract

The invention discloses an active modeling and prediction method for the flow rate of a reciprocating mixed pump, comprising the following steps: 1) determining the input and output variables of the model, collecting a small number of sample sets S of M typical working conditions, and obtaining M samples Subset; 2) Based on the change of the prediction uncertainty information provided by the local GPR, each of the above sample subsets is individually trained; (3) Based on M training sample sets that have completed the learning process, establish a reciprocating mixed input The global GPR model of the pump flow rate, and predict each sample point under the new working condition, and obtain the flow rate curve under the new working condition. The present invention realizes the modeling and prediction of the flow rate of the reciprocating mixed-transport pump under the mixed-transport working condition by adopting the above technology. Compared with the dependence of the CFD modeling process on the designer's experience level, and the traditional method of obtaining useful dynamic information from a large amount of experimental data by the data-driven empirical model, this method provides a method suitable for a small number of modeling samples, and is easy to implement in engineering Methods.

Description

technical field [0001] The invention relates to a modeling and prediction method for important parameters in the design stage of a reciprocating mixed transportation pump, in particular to an active modeling and prediction method for the flow rate of a reciprocating mixed transportation pump suitable for complex frequency-variable oil and gas mixed transportation conditions. Background technique [0002] The reciprocating mixed pump has both the functions of a pump and a compressor, and can transport gas-liquid mixed media at the same time, effectively increasing the output of oil and natural gas in the process of oil extraction. Under the condition of mixed transportation, the pump is in the conveying state of oil, gas, water and other multi-phase mixtures with varying ratios. The frequency-varying impact load of heterogeneous flow generated in the pump will induce the opening and closing hysteresis of the inlet and outlet valves. As attached to the manual figure 1 As show...

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
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06F30/17G06F30/27G06F30/28G06N3/08G06F119/14
CPCG06F30/17
Inventor 刘毅邓鸿英张生昌
Owner ZHEJIANG UNIV OF TECH