Cnc Machines Electrical System Reliability Evaluation Based On Bayes Theory

This paper is proposed based on the national high technology research and development program (863 program) subject of "Facing Life Cycle Of Heavy NC Equipment Credibility Modeling And Analysis Technique" project.Along with the development of numerical control machine,its product varieties and function is becoming more and more complex, makes the reliability evaluation of nc machine tools become an important index. In the study,the number of the sample size is very important, large sample cost crucial time,It increase costs. Restricted by many conditions, often experimental data are small sample. This paper solve the binomial distribution data model based on the small sample reliability assessment questions, using Bayes method and pyramid system evaluation method of combining the CNC machine electrical system to solve the problems of reliability assessment.In Bayes method, firstly we should fully utilize priori information as much as possible, so we can save manpower and material resources and can get more accurate estimation results. The Paramount problem is the establishment of a priori information. In this paper the prior distribution function is based on matlab toolbox , it use the optimal distribution fitting principle—namely using simulation results unreliability value, fitting for a priori function unknown super parameter values (choose two groups of common fitting function Beta distribution and negative logarithmic Gama distribution) compared to its fitting effect Beta distribution, so as to obtain the most for reliability R preferred check distribution functions.Secondly,it mainly discussed the binomial distribution of electric system for data model Bayes reliability R assessment, that based on the fuzzy reliability theory, which give the electrical system reliability comprehensive formula an ambiguity of system reliability formula, each parameter through quantitative and qualitative analysis, "numerically-controlled machine tool's electrical system" as simple series structure system, the reached the prior distribution function, r. draws Bayes posterior robability density function of the electricity system, system reliability assessment, get its parameters point estimation and interval estimate. Because the number of prior information to the number of the test sample has certain influence to a priori information, finally, this paper analyzed on credibility after the confidence lower limit get point estimation and based on the analysis of the reliability, known more reasonable.This paper aimed at nc machine tools, namely the electrical system CNC machine failure of each subsystem is the highest frequency subsystems, it has the reliability evaluation of representative, this remarkable reliability evaluation methods for other subsystems too is universal significance.