Fault Diagnosis Using Wavelet-Domain HMMs

The HMT model can also be extended from the single-tree structure to the multi-tree structure (MHMT), which is then used in the multivari- 

The root nodes of each tree structure are connected, corresponding to each variable under consideration. In each single tree, the deterministic trend information and the random factors are all accounted for. The rationale behind using the multivariate tree structure is to be able to capture the correlations among variables. Here, the connection among variables is arbitrary, and the apparent parent-child connection does not really imply the parent-child dependence, but it is just a way to model the relation be- 

While on-line implementation of the trend detection strategy is indeed feasible, the training of HMT models may be rather time consuming and inefficient. We note the following on the implementation issues  

Moreover, combining detail and approximation coefficients to build HMT models would be a natural next step in process trend detection, but this extension is hampered by computational difficulties as mentioned before. 

Two case studies are presented next to illustrate this strategy 

---

Fault Diagnosis Using Wavelet-Domain HMMs... 

---