Research Article
Hidden Semi-Markov Models for Predictive Maintenance
Figure 4
Condition monitoring using the Viterbi path. HSMMs can be effective to solve condition monitoring problems in time-dependent applications due to their high accuracy in hidden state recognition.
(a) State estimation with Viterbi path for continuous data and Gamma duration distribution |
(b) State estimation with Viterbi path for discrete data and Gaussian duration distribution |