core Estimated learning time: 9 h

OE-11.5 Fuzzy Logic System Components

You can build a fuzzy inference system end to end - fuzzification, membership assignment, rule base and decision making, then defuzzification - and compare defuzzification methods.

This unit assembles the previous one into a working controller: crisp input in, fuzzy inference through a rule base written in near-plain language, crisp output out. The rule base is the system's knowledge and it is usually written by a domain expert rather than learned, which is fuzzy control's great strength and its great limitation. Defuzzification methods differ in what they optimize - centroid is the common default because it uses the whole output set rather than a single point.

Work through these

  • Fuzzification

    Turning a crisp measurement into degrees of membership, which is the first block of any such system. It is where the membership functions from the previous topic are used.

  • Membership value assignment

    Deciding what degree a particular value gets, by method rather than by intuition. Several approaches exist and they disagree, which is a useful thing to know.

  • Development of the rule base

    Writing the rules, which is where the expert knowledge actually lives. This is the item that makes these systems explainable in a way networks are not.

  • The decision-making system and fuzzy inference

    Combining the rules that fired to produce a fuzzy answer. The inference method is a design choice with visible effects on the output surface.

  • Defuzzification to crisp sets

    Turning that fuzzy answer back into a single number, because an actuator needs one. It is the last block and the one with the most competing methods.

  • Defuzzification methods and their comparison

    The methods themselves, compared. Centroid is the default and the comparison shows you when it is the wrong default.

  • Applications of fuzzy logic systems

    Where these systems are used, which is heavily in control. It closes the course by connecting the fuzzy half back to the kind of problem the network half also solves.

Sign in to keep your progress.

Free resources

We haven't checked most of these for screen reader use yet.

Links last checked 30 Aug 2026.

Stuck here?

Ask a mentor. A real person answers, and they can see exactly which topic you're on. Usually within a couple of working days.

Checking your session…

Topics shown in module order.