A stochastic process model for life cycle cost analysis of nuclear power plant systems

Homogeneous Continuous-Time, Finite-State Hidden Semi-Markov Modeling for Enhancing Empirical Classification System Diagnostics of Industrial Components

“…For the assessment of the degradation state of DrE bearing affected by balls failure mode, Table 2 reports the classification performance on the real CWRU bearing dataset as developed in our earlier work [22]. (1,1) indicates that the probability of observing o = 1 when i = 1 is 0.887, whereas this probability decreases to 0.082 if the actual state is 2. As mentioned above, this confusion matrix B will be used as the observation matrix within the HCTFSHSMM framework.…”

“…Multistate degradation modeling has been receiving considerable attention for supporting dynamic maintenance paradigms based on condition monitoring (CM), such as condition-based maintenance and predictive maintenance (examples in References [1][2][3][4]). In fact, multistate models describe the degradation evolution more realistically than binary models, as the evolution of many degradation processes proceed in successive phases characterized by different physical degradation mechanisms.…”

Genetic Algorithms in the Framework of Dempster-Shafer Theory of Evidence for Maintenance Optimization Problems

“…Each of these distributions (e.g., Weibull) depends on the set of parameters (e.g., scale and shape parameters), whose values are provided by experts in the form of intervals to account for the uncertainty in their values. This degradation model has been used in different contexts; for example, to optimize maintenance strategies [21], [22], in support to reliability analysis [31]- [33], and in prognostics applications [34], as it provides an estimation of the remaining useful life of the component.…”

“…• Probability distributions have been used to represent the uncertainty in the parameters of the stochastic models of the degradation mechanisms (e.g., [5], [6]). However, the capability of the probabilistic approach to represent the epistemic uncertainty associated with the expert judgments has been questioned [7], [8].…”

Genetic algorithms for condition-based maintenance optimization under uncertainty

“…Multi-state degradation modelling has recently received considerable attention in the domain of reliability and maintenance engineering [2], [7], [36], [45], [46], as it pragmatically allows setting advanced maintenance paradigms such as Condition-Based Maintenance (CBM) and Predictive Maintenance [62], [63]. In practice, the parameters governing the stochastic transitions among the states of these models are first estimated, based on the available data; then, the degradation model is embedded into the maintenance model to estimate the performance indicators of interest (e.g., unavailability [34], [47], profitability [1], quality in production [11], total costs, risk [19], etc.…”