Modularized complicated equipment Bayesian network failure prediction method

Abstract

The invention discloses a modularized complicated equipment Bayesian network failure prediction method, which is used for solving the technical problem of poor practicability of the existing equipment failure prediction method. The modularized complicated equipment Bayesian network failure prediction method has the technical scheme that firstly, complicated equipment is decomposed and modularized according to actual requirements and physical boundaries among components; then, a FPBN (Failure Prediction Bayesian Network) model of each module is built for each module of the complicated equipment by different modeling methods, and the model of each module is further modified so that each module is modularized for forming a FPBNM (Failure Prediction Bayesian Network Module); then, each PBNM is integrated for building an FPBN integrated model of the whole complicated equipment system; and finally, the actual operation state of the equipment is predicted on the basis of the built fault prediction integrated model through using detection information as a drive by using a probability theory formula. The modularized complicated equipment Bayesian network failure prediction method has the advantages that the method is used for the maintenance and the guarantee of the complicated equipment on the basis of the modularization and the Bayesian network; the time state of the complicated equipment can be fast and accurately predicted; the maintenance efficiency is effectively improved; the maintenance cost is reduced; and the practicability is high.

Description

technical field [0001] The invention relates to an equipment failure prediction method, in particular to a modular complex equipment Bayesian network failure prediction method. Background technique [0002] Document "Cai Zhiqiang, Sun Shudong, Si Shubin, et al. FMECA-based Bayesian Network Modeling for Fault Prediction of Complex Equipment [J]. Systems Engineering Theory and Practice, 2013, 33(1): 187-193." A Bayesian network model modeling method for failure prediction based on failure mode, effects and criticality analysis (FMECA) knowledge. Based on the analysis of the fault information contained in the existing FMECA knowledge, this method proposes a fault prediction Bayesian network (FPBN) network structure transformation method and a calculation method of FPBN probability parameters based on FMECA units, and establishes a corresponding FPBN unit model. Then, the FPBN unit models corresponding to each component of the complex equipment are connected to construct the F...

AI Technical Summary

Problems solved by technology

[0005] In order to overcome the shortcomings of poor practicability of existing equipment failure prediction methods, the present invention provides a Bayesian network failure prediction method for modular complex equipment

Images