Optimization of non-periodic inspection of structural components by Bayesian approach

抄録

This paper presents an advanced Bayesian analysis method to determine the appropriate non-periodic inspection intervals of fatigue-sensitive structures. The calculation procedure of the posterior distribution is improved compared to the previous methods. The method is based on assumptions about the probability density functions of the time until crack initiation, a law of crack propagation, the probability of crack detection and the failure rate before and after crack initiation. A major feature of this method is that even when there are uncertain parameters in the equations, the next inspection interval can be determined. Moreover, the probability density function of the uncertain parameters is updated according to the results of each inspection. A simulation study was conducted to evaluate the proposed method. The proposed method is evaluated and it is statistically shown that this Bayesian approach allows (a) evaluation the inspection interval accurately even with incorrect prior knowledge about the parameters and, (b) estimation of the reliability of a system accurately, even when some of the parameters are uncertain.