Font Size: a A A

Electromagnetic Flowmeter Reliability Research Based On Small Sample Theory And Monte Carlo

Posted on:2016-04-14Degree:MasterType:Thesis
Country:ChinaCandidate:W Z JiangFull Text:PDF
GTID:2272330467988423Subject:Safety science and engineering
Abstract/Summary:PDF Full Text Request
Electromagnetic flowmeter is a widely used in petroleum, chemical, natural gas,environmental protection, medicine, health, food, trade settlement, variousmachinery and equipment in areas such as measurement of conductive liquid flowmeter, its reliability is restricting industrial development of flow sensors commonand key issues. Our research and development of electromagnetic flowmeters startedrelatively late, more satisfied with imitation, the level of reliability and someinternational brands there is a gap, though gradually narrowing these gaps, but hasnot been widely accepted and recognized. Electromagnetic flowmeter reliabilitytechnology research, especially reliability design, growth and reliability assessmenttechnology research is an urgent matter.With the rapid development of intelligent, digital instrumentation, the updatespeed of electromagnetic flowmeters also will accelerate. For the assessment of thereliability of the electromagnetic flowmeter using the traditional method has beenunable to meet the accuracy requirements of the assessment, and the traditionalmethod requires large amounts of failure data. The electromagnetic flowmeter isupdating constantly, causing the electromagnetic flowmeter failure data showing“small sample”“different overall” characteristics. So electromagnetic flowmetersmall sample reliability assessment has important theoretical and practicalsignificance.Reliability analysis flow sensor and process control systems sub-topic “Flowsensor reliability assessment and growth” and the Natural Science Foundation ofHeilongjiang Province,“The construction of complex system reliability analysismodel based on Bayesian networks and information” combining the national “863”key project, in order to seek reliability assessment techniques and methods ofelectromagnetic flowmeters, engaging in a small sample reliability assessmentresearch, the main contents are as follows: 1. Using electromagnetic flowmeter after-sales maintenance service data of acertain type of an enterprise, the reliability of the traditional statistical analysismethod is used for fault interval time distribution model fitting, parameter estimation,hypothesis testing, the life distribution model for describing the electromagneticflowmeter; Using the data from statistical pie chart, fault frequency statisticalanalysis and hazard degree analysis, find out weak links and defects affect thereliability of the electromagnetic flowmeter.2. Engineering used three kinds of prior distribution determination methods aresummarized by studying small sample reliability assessment techniques and methods,detailed analysis of the theory, basic step and the key issues of Bayes theory inreliability assessment. The Monte Carlo simulation method of electromagneticflowmeter fault data of two parameter exponential distribution is studied. Thismethod provides the possibility for verifying the accuracy of small sample reliabilityassessment methods.3. Failure rate of random variables with electromagnetic flowmeters kid likereliability assessment model is established based on traditional reliability assessmentresults and introduced the Bayesian theory to assess the reliability of electromagneticflowmeters. Feasibility of small sample reliability assessment methods applied to theelectromagnetic flowmeter is studied by using electromagnetic flowmeter failure datareliability index calculation.4. Aiming at the problem of accuracy assessment of Bayesian method of smallsample reliability assessment, the research of exponential distribution under differentparameters evaluated precision double prior distribution data were generated usingthe Monte Carlo simulation technology is put forward.
Keywords/Search Tags:electromagnetic flowmeter, small sample reliability assessment, Bayesian, Monte Carlo, two-parameter exponential distribution
PDF Full Text Request
Related items