PID Controllers, 2nd Edition

Chapter 6.1 - Automatic Tuning and Adaptation: Introduction

By combining the methods for determination of process dynamics (described in Chapter 2) with the methods for computing the parameters of a PID controller (described in Chapter 4), methods for automatic tuning of PID controllers can be obtained. By automatic tuning (or auto-tuning) we mean a method where the controller is tuned automatically on demand from a user. Typically the user will either push a button or send a command to the controller. An automatic tuning procedure consists of three steps:

  • Generation of a process disturbance.
  • Evaluation of the disturbance response.
  • Calculation of controller parameters.

This is the same procedure that an experienced operator uses when tuning a controller manually. The process must be disturbed in some way in order to determine the process dynamics. This can be done in many ways, e.g., by adding steps, pulses, or sinusoids to the process input. The evaluation of the disturbance response may include a determination of a process model or a simple characterization of the response.

Industrial experience has clearly indicated that automatic tuning is a highly desirable and useful feature. Automatic tuning is sometimes called tuning on demand or one-shot tuning. Commercial PID controllers with automatic tuning facilities have only been available since the beginning of the eighties. There are several reasons for this. The recent development of microelectronics has made it possible to incorporate the additional program code needed for the automatic tuning at a reasonable cost. The interest in automatic tuning at universities is also quite new. Most of the research effort has been devoted to the related, but more difficult, problem of adaptive control.

Automatic tuning can also be performed using external equipment. These devices are connected to the control loop only during the tuning phase. When the tuning experiment is finished, the products suggest controller parameters. Since these products are supposed to work together with controllers from different manufacturers, they must be provided with quite a lot of information about the controller in order to give an appropriate parameter suggestion. The information required includes controller structure (standard, series, or parallel form), sampling rate, filter time constants, and units of the different controller parameters (gain or proportional band, minutes or seconds, time or repeats/time). The fact that PID controllers are parameterized in so many ways creates unnecessary difficulties.

Tuning facilities are also starting to appear in the distributed control systems. In this case it is possible to have a very powerful interaction of the user because of the graphics and computational capabilities available in the system.

Even when automatic tuning devices are used, it is important to obtain a certain amount of process knowledge. This is discussed in the next section. Automatic tuning is only one way to use the adaptive technique. Section 6.3 gives an overview of several adaptive techniques, as well as a discussion about the use of these techniques. The automatic tuning approaches can be divided into two categories, namely model-based approaches and rule-based approaches. In the model-based approaches, a model of the process is obtained explicitly, and the tuning is based on this model. Section 6.4 treats approaches were the model is obtained from transient response experiments, frequency response experiments, and parameter estimation. In the rule-based approaches, no explicit process model is obtained. The tuning is based instead on rules similar to those rules that an experienced operator uses to tune the controller manually. The rule-based approach is treated in Section 6.5.

Some industrial products with adaptive facilities are presented in Section 6.6. Four single-station controllers are presented: Foxboro EXACT (760/761), Alfa Laval Automation ECA400, Honeywell UDC 6000, and Yokogawa SLPC-181 and 281. Three tuning packages to be used within DCS systems, Fisher-Rosemount Intelligent Tuner and Gain Scheduler, Honeywell Looptune, and ABB DCS Tuner are also presented, as well as the process analyzer Techmation Protuner.

Adaptive techniques are closely related to diagnosis procedures. Section 6.7 gives an overview of both manual and automatic on-line diagnosis procedures. The chapter ends with conclusions and references in Sections 6.8 and 6.9

UNLIMITED FREE
ACCESS
TO THE WORLD'S BEST IDEAS

SUBMIT
Already a GlobalSpec user? Log in.

This is embarrasing...

An error occurred while processing the form. Please try again in a few minutes.

Customize Your GlobalSpec Experience

Category: Motion Controllers
Finish!
Privacy Policy

This is embarrasing...

An error occurred while processing the form. Please try again in a few minutes.