Process Monitoring Using Hidden Markov Models

Abstract

Autocorrelated data arise in a variety of processes. To statistically monitor such processes, special statistical tools are needed to account for these correlations. The most common set of tools for such purpose is the autoregressive integrated moving average (ARIMA) models. Implementation of ARIMA models requires a fair amount of background understanding of how these models work, because the model selection step is essential. In this paper, we propose a new monitoring technique based on the use of hidden Markov models. The proposed monitoring method is powerful yet simple to use technique because it only requires a basic knowledge in statistics. Simulation results show that the proposed method performs similar to the ARIMA models in terms of average run length for detecting out of control processes and false alarms. Copyright © 2013 John Wiley & Sons, Ltd.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 24 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.