# Working Inside the Fuzzy Logic System | Artificial Intelligence

In this article, we are going to thoroughly study about how the processing of data and inference making from the available data takes place inside the Fuzzy Logic System? We will also learn about the algorithm which is followed by the Fuzzy Logic System for decision making.
Submitted by Monika Sharma, on June 17, 2019

As discussed earlier, the Fuzzy Logic System consists of 4 components: the Knowledge Base, Fuzzification Module, Inference Engine, and the Defuzzification Module. We know how the data and information flow between these components, but we do not know how the processing of that information takes place. Here, we are going to study the same.

## Membership Function

The membership function is the backbone of the Inference Engine. It is a function which quantifies the data and represents a Fuzzy Set, which is defined over the range 0 to 1 (both inclusive). The input space that the Membership Function works in is known as the Universe of Discourse and the data that it takes as input are usually linguistic terms.

The Linguistic terms can be defined as the words which define the physical characteristics of a function. For example, if we are defining the temperature of a body, then we use the terms which define the characteristics of it, like high, low, very high, moderate, etc. These are the linguistic terms here.

The Membership function can be defined as: µP: X → [0, 1]

Where,

• 'µ' denotes the membership function,
• 'P' denoted the Fuzzy Set,
• and 'X' denotes the universe of discourse, i.e. input space.

Algorithm:

The algorithm on which the Fuzzy logic system is as follows:

1. Define the Knowledgebase by feeding the Fuzzy set Rules into it.
2. Define the universe of discourse for the membership function.
3. Construct the membership function (By any method, Triangular, Singleton or Gaussian).
4. Perform Fuzzification to convert the input information into data in the form of Fuzzy sets.
5. Process the Fuzzy data set and draw the inference using the rules defined in the Knowledgebase (This process takes place inside the Inference Engine).
6. Perform Defuzzification to convert the fuzzy data into the user-understandable form and produce the output.