Font Size: a A A

Solid Mass Flowrate Me Asurement Method By The Double-elbow Method

Posted on:2019-03-26Degree:MasterType:Thesis
Country:ChinaCandidate:X F HuangFull Text:PDF
GTID:2382330563490622Subject:Control engineering
Abstract/Summary:PDF Full Text Request
The solid phase mass flow rate measurement of gas solid two phase flow is influenced by many factors.There exist randomness and contingency.So the measurement of solid mass flow has the problems of low accuracy,large error and unreliable result.The principle of double elbow method has been deeply studied.Considering the influence factors of solid volume in gas-solid two-phase flow,a new double elbow method mathematical model for measuring solid mass flow rate has been established: Double Elbow—Volume Ratio Method Mathematical Model(DEVRMMM).Combined with the experimental data,the DEVRMMM was analyzed by least square method,and the results were compared with the double elbow method measurement model.The analysis results showed that the relative error of solid mass flow measurement based on DEVRMMM was less than 15%,the accuracy of the results was higher than that in the original mathematical model.Based on BP neural network,the method of measuring the mass flow rate of solid phase with DEVRMMM was analyzed and studied.A 4-7-3-1 type 4 level BP neural network was established.The result showed that the error of solid mass flow measurement was less than 7%.A 3 level neural network model with 3 inputs and 1 outputs was established by using Elman neural network.The solid mass flow rate was less than 3%.The solid phase mass flow measurement based on different algorithms was compared and analyzed.Results showed: The calculation process of the least square method was simple,but the error was large.The method of neural network calculation improved the accuracy of measurement.The training speed of BP neural network was faster and the precision of Elman neural network was higher.
Keywords/Search Tags:double elbow method, gas-solid two-phase flow, solid mass flow, mathematical model, artificial neural network
PDF Full Text Request
Related items