Reliability Assessment of Large Systems That Fail as a Result of Damage Accumulation

The problem of assessing the reliability for systems containing a large number of elements is considered in a general statement. A subclass of systems in which system elements are exposed to the flow of damage is distinguished. Accumulation of defects resulting from damage leads to system failure. The assessing the reliability of such systems problem is reduced to solving the non-trivial problem of finding the random value for the sum of a random terms number distribution density. A general model of this problem is proposed, which determines the structure of the computational procedure for its solution. The solution is brought to the final results for a special case when random values of defects from damage are distributed according to the exponential law. For the general case, a solution is proposed based on the approximation of the desired relation for the distribution density of the random terms number sum. The approximating model is a normal distribution, the parameters of which are estimated by standard statistical processing of experimental data. The numerical values of these parameters depend on the type of random defects distribution and on the damage flow characteristics. The values of the numerical distribution density characteristics of the total defect are determined by the duration of the observation interval. Therefore, the obtained distribution density can be used to predict the reliability of the system. Possible directions of further research are considered, related to taking into account errors in estimating the parameters of the problem, which are a consequence of the available source data limited amount