When the Ziegler-Nichols PID tuning method doesn’t do the trick

Posted by Michiel Huizer on Jul 8, 2020 9:40:51 AM

Many tuning methods are proposed for PID controllers, of which trial and error is the best known and used method. Despite its popularity, the biggest downside is that it’s time-consuming and it doesn’t guarantee a robust and stable solution. The Ziegler-Nichols method is a good alternative but doesn’t always provide optimal performance. In this blog, you will read why Ziegler-Nichols isn’t always the right choice to achieve stable and robust control loops.

The Ziegler-Nichols PID tuning method explained

Ziegler-Nichols is an often applied rule-based method that assumes a certain process response to obtain easy mathematical formulas for PID tuning. It typically results in aggressive control performance. The process characteristics are derived from simple experiments and used to calculate the PID parameters. The Ziegler-Nichols tuning method provides two different methods (the step response method and the frequency response method) and you might end up with undesired overshoot.

In general Ziegler-Nichols leads to fast, sometimes too aggressive tuning, including overshoot.
Ziegler-Nichols method only guarantees a quarter amplitude decay damping ratio (height of the second overshoot peak to the first overshoot peak).

 

The Ziegler-Nichols PID tuning method is simple and intuitive

Basically, Ziegler-Nichols works well enough when the dead time is small compared to the time constant of the process. It’s also simple, intuitive and it obtains reasonable performance for simple loops. DCS engineers often use this method when they need to tune new loops. New loops get always first a set of default tuning parameters. When this works well enough, there is no reason to retune the loop. When the loop needs to be retuned it might be smart to use a tuning rule such as Ziegler-Nichols. This method is also applicable to existing loops. When a closed-loop becomes active and doesn’t work properly, you might not know where to start. It can be helpful to use the Ziegler-Nichols method to reach an initial estimation of the tuning rules followed by manual tweaking of the parameters.

The assumptions of Ziegler-Nichols lead to shortcomings

Small discrepancies between estimated and actual process characteristics (gain or process delay) can result in an extremely oscillatory or even unstable control loop. The loop might not be robust enough against changes in the process dynamics, including non-linearities.
 
The Ziegler-Nichols PID tuning method gives reasonable results as long as all assumptions are respected. When the assumptions are not respected the closed loop behavior will be unknown and often not desired.
 
Learn more — PID Tuning
 

A different PID tuning method for robustness

So, the Ziegler-Nichols PID tuning method is a perfect tool to use when you don’t know exactly how to tune the loops. But, you should consider a different PID method when you want to have control over the loop objectives. When you don’t allow a quick response behavior, overshoot, or oscillations a model-based approach may be a better fit. This enables you to calculate the right values and reach your goals faster.
 
 

Model-based tuning leads to a safer plant

The main benefit of the model-based PID tuning method is that the engineering constraints and the robustness of the loop are captured and warranted. And a robust plant means a safe plant. External variables that cause changes in the plant are compensated which creates more margin to absorb these changes. Tuning software — which is grounded on the model-based tuning method — enables you to set the parameters right the first time.
 
Would you like to learn more about the different PID tuning methods and their benefits and downsides? Download the ebook: “The guide to PID tuning” and discover which method suits your plant best.
 
We are happy to help you with your plant optimization.

Download ebook

 

Do you like to know what INCATools PID tuning software can offer you?
Request your demo here.

Topics: PID tuning methods