Reliability Assessment Of The System And Non-failure Data Processing

The problem of reliability is proposed that due to large-scale industrial production and the demand of using complex military equipment in the Second World War. With the continue development of science and technology, the components'reliability has been greatly improved. However, because of large-scale system's structure is getting more and more complex, the function which need be completed is also getting more widespread. Therefore, the quantitative analysis and the improvement of system's reliability have become an important topic.This article mainly study system reliability and non-failure data processing.Firstly, we study a class of series and parallel systems with controllers. The controller's function is to monitor the system, in order to the failed components being separated immediately from the system, and saving energy. As a result, the system can function more effective. Assume that the working time and the repair time of the components in the system are both exponential distribution and each component after be repaired is as good as new, by using the properties of Markov process, we can evaluate the important reliability indexes of the system.Secondly, we study the special systems with the load-sharing. This system also comes from the practice. For example, a system is composed of two parts, when the system works, the two components take on the load. If one part fails and the system can work normally, then the other will take on larger load. Based on the research of the failure rate increasing, we obtain the transfer rate matrix. Thus the reliability indexes of the system can be evaluated. And On the basis of the above, we put forward such a model that so- called Under Load-sharing m-out-of-n voting system with k Repair facilities and the failure rate changing with time. On the assumption that the component working time, controller working time and the component repair time are all exponential distribution. Using the properties of Markov process, we can obtain the evaluations of system availability and reliability.Thirdly, we research the warm standby redundant system with two categories of components and multiple repair facilities. Type I are the old parts and have the priority to be used, the old parts will be discarded if failure. Type II are the newly installed parts, and the repair time are exponential distribution. Through the analysis, we draw out the graph of the system transition rate, furthermore, obtain system transition rate matrix. According to the Kolmogorov-Feller forward equation, we obtain the evaluations of the important reliability indexes of system.Next, we study a repairable consecutive k-out-of-r-from-n:F system. It has n components arranged in a linear and fails if and only if at least k out of r consecutive components fail. To the best our knowledge, this system remains unexplored. In this article, it is assumed that the working time and the repair time of each component in system are both exponential distribution, and every component after repair is as good as new. The key component has priority of repair. The transition rate matrix is obtained by defining the generalized transition probability, furthermore, we get the equations to calculate the reliability and availability of system. Finally, the approach is descried by a numerical example.Finally, we study non-failure data processing. As the continue development of science and technology, the product lifetime is getting higher and higher. So we often encounter the case of non-failure data under fixed-time testing. In this paper, using Bayesian method and least squares theory, we estimate the reliability of the system.