Multi-level system reliability analysis method based on Bayesian mixing

An analysis method and reliability technology, applied in the direction based on specific mathematical models, special data processing applications, instruments, etc., can solve the problems of ineffective processing of multi-source inconsistent information, insufficient analysis accuracy, etc., and achieve the goal of reducing accuracy Influence, expand the scope of application, improve the effect of accuracy

Inactive Publication Date: 2019-11-12
UNIV OF SCI & TECH BEIJING
View PDF1 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0014] The invention provides a multi-level system reliability analysis method based on Bayesian mixing. The technical problem to be solved is that the existing reliability analysis method cannot effectively deal with multi-source inconsistent information in a multi-level complex system, and the analysis accuracy Not tall enough

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Multi-level system reliability analysis method based on Bayesian mixing
  • Multi-level system reliability analysis method based on Bayesian mixing
  • Multi-level system reliability analysis method based on Bayesian mixing

Examples

Experimental program
Comparison scheme
Effect test

no. 1 example

[0052] This embodiment aims at the problem that the existing reliability analysis method cannot effectively deal with multi-source inconsistency information, and the accuracy of the analysis result is not high enough, and provides a multi-level system reliability analysis method based on Bayesian mixing. This embodiment The method includes the following steps:

[0053] S1, based on the system structure composition, calculate the system reliability function expression;

[0054] S2, based on the system reliability function expression, using the random variable conversion relationship to obtain the indirect prior distribution of the system parameters;

[0055] S3, according to the indirect prior distribution of system parameters and the preset direct prior distribution of system parameters, apply the Bayesian hybrid method to calculate the fusion prior distribution of system parameters;

[0056] S4, based on the prior distribution of system parameter fusion, calculate the updated ...

no. 2 example

[0082] This example will combine figure 2 The multi-level system shown will illustrate the application of the present invention in system reliability analysis;

[0083] Without loss of generality, the i-th unit E in / row (l，i) is the research object, and its parameter set is θ (l，i) . Then the parent node E of line / +1 (l+1，j) Then there is a parameter set θ (l+1，j) . Given its parameter direct prior distribution π D (θ (l+1，j) ), its reliability function can be generally described as R (l+1，j) (t)=f(t|θ (l+1，j) ). Among them, f() is a function determined by the specific physical background and failure mechanism. Then, the general form of the reliability function of the research object and the corresponding parameter probability density function is

[0084] R (l，i) (t|θ (l，i) )=Ψ (l，i) (R (l+1，j) (t|θ (l+1，j) ): j∈Q (l，i) ) (10)

[0085]

[0086] In the formula, Ψ (l，i) by the object unit E (l，i) and its parent node E (l+1，j) Determined structure func...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention provides a multi-level system reliability analysis method based on Bayesian mixing, which comprises the following steps of: selecting a description model according to structural characteristics of each unit, and determining direct prior distribution of parameters; establishing a likelihood function based on the available data set; applying bayesian updating to calculate parameter posteriori distribution; calculating a system reliability function expression based on the system structure composition; utilizing a random variable conversion relationship to obtain indirect prior distribution of system parameters; calculating system parameter fusion prior distribution by applying a Bayesian mixing method; calculating updated system parameter posteriori distribution based on the system parameter fusion prior distribution; and outputting various reliability indexes according to the posteriori distribution of the system parameters. The main innovation of the method lies in that anovel Bayesian mixing method is provided and used for processing multi-source non-consistency information in reliability analysis of the multi-level system, the application range of a traditional Bayesian method is expanded, and the accuracy of reliability analysis of the multi-level complex system can be improved.

Description

technical field [0001] The invention relates to the technical field of system reliability modeling and analysis, in particular to a Bayesian mixing-based multi-level system reliability analysis method suitable for multi-level complex systems containing multi-source uncertain information. Background technique [0002] Bayesian method is a mathematical statistical method widely used in system reliability analysis. The Bayesian method integrates subjective information into the prior distribution, establishes a likelihood function based on objective data, and makes probabilistic inference after comprehensively utilizing all available information. The general form of the classic Bayesian inference method is: [0003] [0004] where π(θ) is the prior distribution of the parameter θ, f(D|θ) is the likelihood function, and π(θ|D) is the parameter posterior distribution considering the data set D. [0005] However, for the reliability analysis of multi-level complex systems, the ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06F17/50G06N7/00
CPCG06N7/01
Inventor 杨乐昌王蔷贺可太
Owner UNIV OF SCI & TECH BEIJING
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap