The Reliability Evalution Of The LJ6010 Multi-function Series Standard Source

Evaluating reliability of products scientifically can reduce the cost, check design reasonability, optimize system and shorten research cycle. At present, there was almost no method of reliability evaluation for the complex and expensive metrological instrument. On the basis of the background described above, the subject named the Reliability Analysis and Research on standard AC and DC resource (serial number: 2002QK30) was set up by State Bureau for Quality Testing, which mainly deals with the key projects. In this paper, the reliability evaluation is carried out based on the research on the LJ6010 (multifunctional standard source).Firstly, the reliability prediction of the LJ6010 is conducted with the combination of counting method and stress method, the result of which is its characteristic values, with which the initial evaluation for the whole machine can be achievedSecondly, in order to carry out the qualitative analyze of the LJ6010, a method of fault tree is adopted. Monte Carlo simulation method is used to conduct quantitative analyze. The intolerant errors in the calculation by the utilization of the counting method and stress prediction method to analyze complex system under the non-exponential distribution condition are corrected.Finally, by using the characteristics of neutral network self study and non-linearity fitting. We can facilitate the reliability initial evaluation. On the basis of the failed data of related products we collected, LJ6010's neutral network model can be established, the touchy problems of reliability prediction can be overcome, and simulation carried out, the outcome of which can prove the method is infeasible.In this paper, the reliability evaluation of LJ6010 is conducted with different ways under varied situations, the main purpose of which is to explore reliability evaluation methods for engineering use, integrate the theory of reliability evaluation and engineering practice and offer a reference method of reliability prediction for metrological instrument.