Answers are grounded in this paper and cite their sources.
A simple time-delay estimation approach that requires temporarily retuning the existing PID controller is presented. Closed-loop output data are fitted with autoregressive moving average models at two different controller tunings. The time delay is determined by using the feedback-invariance principle that the closed-loop impulse response coefficients do not change inside the time-delay window. This approach requires no intervention to plant operation and does not estimate the open-loop process model but the time delay only. The estimated time delay is particularly suitable for use in control performance monitoring.
No takes yet. Share an insight, caveat, or question.
Harrison et al. (2012) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: