To avoid the "over-maintenance" or "under-maintenance" in generator set maintenance decision, a maintenance strategy based on Markov chain state estimation model is presented.
Based on wavelet domain Hidden Markov model, a novel speckle suppression method for medical ultrasound images is presented which combines Bayesian estimation and homomorphic filtering.
The three basic problems of two-dimensional (2-d) hidden Markov models (HMMs) are studied, including probability evaluation, optimal states and parameter estimation.