Feedforward Control with Deadtime Observers Including a Multi-Lag Loop

This paper is concerned with the addition of (signal) feedforward control to open and closed-loop observer-based deadtime compensation schemes. The open-loop form is the well-known Smith predictive model and compensation scheme. For the closed-loop observer form, a controller is applied to the predictive model part. It has been shown that such a control system can be tuned to remain stable in the face of major process parametric upsets when compared to other methods. However, under these conditions, sluggish responses are obtained for signal load or disturbance changes. When applicable, the addition of feedforward control could help alleviate this problem. Several schemes for feedforward compensation are presented. The first order process with deadtime is the main basis of the analysis presented which indicates that only conventional feedforward need be considered. Response results are shown for the application of the deadtime compensators to a multi-lag process.