Reliability analysis for system life based on universal generating function

A novel technique for the reliability analysis of a hydraulic system is presented based on the universal generating μ function and useful information can be obtained for system analysts and design engineers to guide effective system improvement both in the design phase and during operation. First, according to the fault data, the states and their corresponding probabilities for the constitutive elements of the system are determined by the histogram technique. Then, based on the determined state probabilities for all elements, an individual μ function for each element is defined by means of μ function. Finally, by using the composition operators over the μ function of individual elements and their logic combinations in the entire system structure, the μ function of the entire system can be obtained by simple algebraic operations, and this μ function expresses the output performance distribution for the system. Due to the definition of reliability measures about the μ function, reliability analysis can be easily completed from this output performance distribution compared with traditional Monte Carlo(MC) method. As the method is performed without estimating the distribution of the elements, the errors of truncation and numerical fitting are reduced. The example of a certain hydraulic system is used to illustrate the correctness and effectiveness of the novel method.