What Is a Fuzzy PID Controller?

A fuzzy PID controller is a proportional-integral-derivative controller that uses fuzzy logic software principles as a form of error control in industrial systems. PID controllers are widely used in automation to adjust valves and other process controls based on the accumulation of errors over time. Where a fuzzy PID controller differs from a standard PID controller is in their ability to provide more accurate results in non-linear situations. Fuzzy systems are built upon a type of programming logic that attempts to deal with the gray areas of uncertainty in processes in a more successful way than standard controls do.

Control systems have to have some built-in method to drive a value that is out of range back toward an acceptable range. Typical PID controllers will have an on/off response to this variability at set points in the process, which can be compared in some ways to a digital method of looking at a process, by breaking it up into discrete values and assigning predetermined actions to those values. A fuzzy PID controller, on the other hand, calculates deviation along an analog range where there is an optimal value and increasingly non-optimal values, but no predefined point at which action is always taken.

The design of a fuzzy PID controller and a standard PID controller both utilize historical values to calculate future responses. The letters in PID in fact stand for this, where P represents present errors, I for past errors, and D as future error states. Fuzzy systems attempt to map the errors in terms of persistence and assign them to various membership sets for different ranges of logic conditions. This allows a fuzzy PID controller to also set the rate of change for bringing a system back under control. This rate of change is based on inference rules, where accumulating data and error states suggests a most correct course of action.

One of the problems with a fuzzy PID controller is that it can arrive at conclusions that are mutually contradictory and take no action. This requires a process of conflict resolution in the code that is commonly referred to in fuzzy systems as defuzzification. Defuzzification is done by giving certain parameters in a PID controller greater weight than others to tip the balance in a certain direction of action, and this is analogous to an adjusted gain in a standard PID controller.

In cases of standard variation, a fuzzy PID controller and a standard PID controller may both regulate the control system in the same exact way. This does not suggest that they are identical control systems or that the advantages of fuzzy control have been disproved. It simply means that the situation is easily handled by any basic control system. Standard PID controllers can be seen as a subset of the fuzzy PID controller, which has a more robust, capable ability to handle unpredictable deviation. In situations where standard PID controllers are failing, a well-designed fuzzy PID controller will work better.

The advantages that a fuzzy PID controller offers include that it can provide different response levels to non-linear variations in a system, and, at the same time, it can function as well as a standard PID controller in a system where variation is predictable. A fuzzy PID controller can also keep a system more stable. It can be weighted towards types of responses just like the gain settings on a standard PID controlled system as well.