Posted By
Dennis Nash President Control Station, Inc. Manchester, CT
Blog Categories

PID Tuning Fundamentals: Simple ‘Recipe” for Tuning PIDs

PostedTuesday, January 29,2019 at 2:05 PM

PID Tuning Fundamentals: Simple ‘Recipe” for Tuning PIDs

Here’s a little secret. The majority of industry practitioners tune PID controllers manually versus using software. The split isn’t even close. As the head of a software company specializing in controller tuning software, I say this: It’s all good…as long as it’s done correctly.

PIDs have been in use since the 1920s when the first controller using a PID equation was introduced. Since that time engineers and technicians alike have been tuning them manually. Many of those manual methods are little more than ‘poke and hope’ approaches. They’re the ones where a new parameter value is chosen and implemented almost at random. The change pokes the process, stimulating a reaction; and the engineer hopes that the loop’s performance doesn’t get worse. Such a method doesn’t exactly inspire confidence. Fortunately there are better ways of tackling this challenge.

Highly effective and repeatable methods for tuning PID controllers manually do indeed exist. Among the many methods used in industry, ours is taught each year to engineering candidates at over 200 colleges and universities. The ‘recipe’ taught to those future practitioners involves the following six (6) repeatable steps which are explained more fully in our PID Tuning Guide (Download Here):

  1. Specify the PID’s Design Level of Operation and Control Objective

Simply put: The Design Level of Operation (DLO) is the range within which a given controller usually operates. It’s important to know this as tuning the controller for operation outside of this range can be counterproductive. Indeed, doing so can result in a PID that’s either too slow or too fast in its response to upsets, or even one that is prone to excessive oscillations. As for the Control Objective it’s essential to know what constitutes ‘good’ control. It’s always good to start with the end objective in mind.

  1. Perform a Bump Test and Collect Dynamic Process Data

There are a variety of tests that are commonly applied such as the Step, Doublet, and PRBS. As should be expected each test has its strengths as well as its weaknesses. Ultimately it’s necessary to drive the process outside of its noise band so that the dynamic behavior of the process can be revealed. Equipped with insights from a good test, it’s then possible to tune the controller so that normal variability in the process can be regulated effectively.

  1. Fit a Model to the Process Data

The behavior of most industrial processes can be accurately described through the use of a First-Order Plus Dead-Time (FOPDT) model. While that seems like a mouthful, the FOPDT is straight forward and easy to calculate. Using a printed trend of the controller’s bump test response, simple formulas reveal model parameters that describe “How Far”, “How Fast”, and with “How Much Delay” the process reacts to upsets.

  1. Use Model Parameters and Correlations to Calculate Tuning Values

Whereas model parameters describe the dynamics of a process, it is tuning parameters that are uploaded into the PLC or DCS for the purpose of controlling a process. Vendors of PLC and DCS technologies use their own unique algorithms. As a result it’s necessary to use the model parameters in tuning correlations that are specific to the vendor’s algorithm. These correlations are usually available in the vendor’s documentation or journal articles that focus on PID optimization. Those correlations can be used to tailor the tuning parameters to deliver appropriate responsiveness and achieve more precise control.

  1. Implement the New Tuning Parameters and Test for Results

As an industry best-practice new tunings should always be tested after they’re implemented. By performing a bump test similar to the one applied earlier in this approach it’s possible to see if the controller response as both expected and needed. Manual tuning methods often require multiple iterations. Each new set of parameters with the associated test shifts the loop closer to the desired state of control.

  1. Document the Decision-Making Process and Associated Results

Establishing a record of the tuning process is an important final step. With access to a detailed record there’s clear evidence of important considerations that can be referenced in the future. What’s more, for integrators documentation provides proof of the work that was done. Staple that to your next invoice.

The beauty of this procedure is its simplicity. Again, download a free copy of our PID Tuning Guide for additional insights and industry best-practices.

Dennis Nash is President and CEO of Control Station, Inc., an award-winning supplier of process diagnostic and optimization solutions. For more information about Control Station, visit the company’s profile here on the Industrial Automation Exchange.


Related Tags:

  • Process Manufacturing
  • Process Engineering
  • Training and Education
  • Process/Batch
  • DCS