Gas chromatograph fault diagnosis method based on Bayesian network

Abstract

The invention provides a gas chromatograph fault diagnosis method based on a Bayesian network. The gas chromatograph fault diagnosis method comprises the following steps: (1) grouping maintenance records of gas chromatographs in years according to working time of the gas chromatographs; (2) analyzing faults and symptoms of the faults of each group, and determining a Bayesian network structure of each group; (3) determining an overall Bayesian network structure; (4) grading the symptoms according to the occurrence possibility and the degree of severity; (5) counting each group of fault symptom frequency, and determining each group of Bayesian network parameters; (6) determining overall Bayesian network parameters; and (7) when a fault occurs, calculating the posterior probability of the fault, and positioning the fault. Maintenance records are grouped, the influence of working time is considered, and faults can be positioned more accurately; the Bayesian network has strong reasoning capability, can quickly calculate the posterior probability of each fault when the fault occurs, and locates the fault according to the maximum posterior probability criterion.

Description

technical field [0001] The invention belongs to the field of fault diagnosis, in particular to a fault diagnosis method for a gas chromatograph based on a Bayesian network. Background technique [0002] Gas chromatograph is an instrument that uses chromatography to separate, detect, and qualitatively and quantitatively analyze multi-component mixtures. It has the advantages of high sensitivity, high efficiency, and less samples, and has good applications in the fields of medicine, chemistry, and environmental monitoring. Once the instrument fails, the difference between the test result and the real value is too large, which will cause serious consequences for subsequent analysis, control, and decision-making. Therefore, it is necessary to quickly and accurately diagnose the fault of the gas chromatograph. Bayesian network is a probabilistic graphical model. The method applied to fault diagnosis is similar to the quick medical reference (QMR) of medical diagnosis: first, the ...

AI Technical Summary

Problems solved by technology

Once the instrument fails, the difference between the test result and the real value is too large, which will cause serious consequences for subsequent analysis, control, and decision-making. Therefore, it is necessary to quickly and accurately diagnose the fault of the gas chromatograph

Images