Code covered by the BSD License  

Highlights from
AutotunerPID Toolkit

image thumbnail
from AutotunerPID Toolkit by William Spinelli
Interactive simulation of a PID with autotuning.

Introduction to PID Autotuning (AutotunerPID Toolkit)
Introduction to PID Autotuning (AutotunerPID Toolkit)
AutotunerPID Toolkit Previous page   Next Page

Introduction to PID Autotuning


What is an Autotuner
Introduction to the PID Autotuning
The ISA-PID Regulator
The structure of the ISA-PID regulator
Process Identification
How to obtain a "description" of the plant represented by a simple model or by a set of characteristics
Synthesis Method
How to compute the parameters of the PID regulator from the available description of the plant
    Note    This introduction to autotuning has been extracted from the IFAC Professional Brief Hands-on PID Autotuning: a Guide to Better Utilisation (A. Leva, C. Cox, A. Ruano), 2002.

What is an Autotuner

An autotuner is something capable of computing the parameters of a regulator connected to a plant (i.e. tuning that regulator) automatically and, possibly, without any user interaction apart from initiating the operation.

The best way to understand autotuners is to analyze how a PID regulator may be manually tuned by a human. The basic steps of a tuning process may be summarized as follows

  • observing the process behavior, eventually stimulating it somehow and turning this knowledge into a description of the process behavior
  • establishing the desired closed loop behavior on the basis of the obtained process description;
  • computing the regulator parameters in order to achieve the desired closed loop behavior.

Formalizing these steps establishes a tuning method. An autotuner is then an implementation of a tuning method capable of running automatically.

The "description" of the process may be obtained in two ways:

  • the first approach identifies a model of the process from the data, where model means something that can be simulated and that captures the process dynamics precisely enough to allow forecasting once the regulator is tuned (model based autotuner);
  • the other approach is to employ the process data immediately for the subsequent tuning (non model based autotuner). The relevant fact is that these representations cannot be simulated. Non model based autotuners are also identified as characteristics based autotuners.

Tuning rules can be constructed in several ways:

  • choosing the regulator parameters so that the expected closed loop behavior (forecasted using the process model) is as similar as possible to that of a reference model (model based, model following autotuner);
  • computing the regulator parameters so that some time or frequency domain properties are satisfied by the closed loop system (model based, characteristics following autotuner).

Previous page  Reference The ISA-PID Regulator Next page

Contact us at files@mathworks.com