Element system reliability optimization method based on Bayesian network

Abstract

The invention discloses an element system reliability optimization method based on a Bayesian network, and the method comprises the following steps: 1, building a conditional probability table based on the Bayesian network; 2, performing reasoning algorithm calculation on the conditional probability table by adopting a hybrid compression coding algorithm to obtain a compression conditional probability table; 3, introducing a standby redundant structure to obtain a delay conversion hybrid model, and simulating a complex scene; 4, according to a PSO algorithm and the compression condition probability table, performing optimization processing on the delay conversion hybrid model in the complex scene to obtain distribution information; and 5, evaluating the reliability of the optimized elementsystem according to the distribution information. According to the method, the problem that the Bayesian network model cannot be normally constructed due to too many components in a complex system and too large conditional probability table when the Bayesian network model is constructed is solved, and based on a delay conversion backup technology, a standby element is adopted to replace a failureelement for optimization and reliability analysis, so that the reliability analysis of the complex system is realized, and the optimization efficiency is improved.

Description

technical field [0001] The invention relates to the technical field of reliability analysis, in particular to a reliability optimization method of a component system based on a Bayesian network. Background technique [0002] In the process of product design and production, reliability analysis can ensure the rationalization of product design and improve durability and durability. Since the development of reliability analysis, it has extended into multiple branches and continues to shine in different fields. [0003] While helping people achieve their expected goals, the development of its own analysis methods is also in continuous progress. Analyzing the components in working state to deduce the system reliability has become another focus of current reliability research. This technique is called hybrid redundant design technique. That is, multiple spare components are used to ensure the smooth operation of the system, but these advanced technologies also cause a large num...

AI Technical Summary

Problems solved by technology

This method aims to solve the problem that there are too many components in a complex system, and the conditional probability table is too large to be constructed normally when building a Bayesian network model. Based on the delayed conversion backup technology, the spare components are used to replace the failed components for optimization and reliability analysis. Reliability analysis of complex systems to improve optimization efficiency

Images