Controller Tuning with Autotune

A controller turning wizard called Autotune was introduced with CADSIM Plus v2.5. Autotune utilizes the Lambda tuning method, which was first documented in a book published by TAPPI PRESS titled Process Control Fundamentals for the Pulp & Paper Industry by Nancy J. Sell, Frederick Y. Thomasson et al. This method allows proportional constant, integral reset time and differential time to be estimated using PID controller tuning equations.

Autotune automates the Lambda tuning method by breaking the process down into three steps:

Step 1: Bump Test

The bump test consists of two step changes in the Controller Output (CO) while tracking the changes in the measured Process Variable (PV). The Autotune wizard analyzes the results of the bump test, attempts to do a process identification, and then compares those results to the response from the actual process model.


Step 2: Process Model Identification

For the purposes of controller tuning, processes generally fall into one of the following categories: first order, second order, first order plus delay and integrating. These process groups are further broken into two types: self-regulating vs. integrating. A step change (or bump) in the CO of a self-regulating process will cause a change in the measured PV that will eventually settle out. A step change in the CO of an integrating process will cause the PV to continue to ramp.

The tuning approach differs between integrating and self-regulating processes. That is why the process model identification step is important. The Autotune wizard makes an attempt to identify the process model type based on its response to the bump test. This step gives the user the opportunity to review those automatic choices, change parameters and compare the resulting model response with the CO response (integrating model shown superimposed on PV, below).


Step3: Tune and Test

When the 3rd step is reached, Autotune has suggested a new set of tuning parameters for the controller. The user may compare these new tuning parameters with the original values. The user may also edit any of the parameters and then run additional tests to see the results. Each test consists of two step changes (bumps) to the Setpoint (SP).

The user can adjust the automated tuning suggestions by changing the Lambda value (in minutes) which makes calculated adjustments to the other fields automatically. Increasing the value in the Lambda field provides the controller with a slower response and stabilizes an unstable process. Decreasing the Lambda value provides a faster response.