System reliability assessment method based on Bayesian network reasoning

Abstract

The invention discloses a system reliability assessment method based on Bayesian network reasoning. The method comprises the following steps: (S1) establishing a system network structure chart based on the system structure; (S2) determining a start search node based on the system network topological property; (S3) obtaining a minimum path set based on a mixed first searching method combining breadth and depth; (S4) establishing a Bayesian network model; (S5) obtaining the prior probability of the root node according to the failure probability of a component corresponding to the node in the system; (S6) determining the conditional probability of non-root nodes; (S7) calculating the system failure probability; and (S8) solving the system reliability. The invention provides a searching method for increasing the minimum path searching speed and a method for converting a system reliability network model into a Bayesian network model; and the problem that the system reliability cannot be calculated by directly using the system reliability network in a complicated network model is solved.

Description

technical field [0001] The invention relates to the technical field of reliability engineering. More specifically, it relates to a system reliability assessment method based on Bayesian network reasoning. Background technique [0002] In the context of increasing demand for aviation, aerospace and military equipment, reliability theory has been further developed in technology on the one hand, and has also been widely used in engineering on the other hand. Reliability theory is of great significance in scientific experiments and production practice. The more commonly used system reliability analysis method is the fault tree analysis method. Fault tree analysis can calculate the reliability of the system and give the impact of bottom events on top events, but it cannot quantitatively give the position of certain bottom events or intermediate events in the entire system. Especially when the status of some links of the system changes, it is difficult to predict its impact on ...

AI Technical Summary

Problems solved by technology

However, these methods have the following limitations: 1) they are only suitable for analyzing unrepairable systems whose components are independent of each other and whose failure distribution follows an exponential distribution; 2) these system reliability evaluation methods are all static analysis methods, which are difficult to reflect the internal components of the system. The intricate coupling relationship between

Images