What is Fuzzy Logic in Artificial Intelligence
What is Fuzzy Logic in Artificial Intelligence

What is Fuzzy Logic?

The word fuzzy describes to phenomenon which are ambiguous or contain fraction of truth. Things which can not be described true or false with full conviction turn as fuzzy. We encounter such happenings in everyday activities where we do not in a situation to categorize the circumstances. For example, during Sunset a moment comes when we are unable to declare day or night.

What is Fuzzy Logic

Fuzzy logic (FL):

The concept of Fuzzy logic (FL) came in the early 1960s, Dr. Lotfi Zadeh from University of California promoted this idea while he was working on Natural language related problem to make computer understanding of human languages. Human languages are tough to translate into binary form (0 and 1) to assemble for computer. In practice, an immense amount of data needs to transform into computer understandable form from its original shape. With this picture, FL can refer as reasoning technique similar to human reasoning. FL emulates course of action of decision making by humans which concerns the entire set of intermediary possibilities in between binary values YES and NO. Below is the architecture of FL.

Figure 1 : Four main steps in Fuzzy Logic

What is Fuzzy Logic in Artificial Intelligence?

FL is a mathematical tool. Artificial intelligence is, well, intelligence through artificially. FL is one of the approaches to achieve intelligence because FL is quite resemble how the human brain works. So it’s logical and makes sense to utilize FL in AI. For the development of AI system, set of fuzzy rules can be incorporated. So, if you want to try and develop an AI system, you can set it up with a list of fuzzy rules (i.e. rules that use fuzzy logic). The issue with utilizing fuzzy logic solely in AI is that such a system can only execute these rules, the system will be unable to learn as goes along.

For decision making:

For decision making, AI can make use of numbers that are quite close instead of predetermined result that is close to the human  brain. When we observe a human face from a faraway distance, at first we might think it’s somebody else since data was fed  into human system was quite close enough. Human brain processes visual information and give judgement that person is our friend but as we approach closer enough we perceive that made presumption was wrong.

Fuzzy data permits AI to recognize resemble images (Both images of person for that matter) and then gives prediction to whom image belongs to although both images are fairly different (One is tall black person, one is a small white person.

Applications of Fuzzy Logic:

Undoubtedly, the applications of FL, once considered to be an obscure mathematical interest only, now applicable in a wide range of fields from scientific to engineering works. FL is being applied in immense applications such a pattern recognition in facial expression, washing machines, vacuum cleaners, abs braking system, management of subway systems, autonomous  helicopters, stock trading, computer vision, autonomous robotics, Energy engineering, automation in industry, electronics products for consumers, development of video games. To make the story short, FL has Inculcated ways in the field of science. Below is organization of Air conditioner with FL.

Figure 2 : Example

The major application areas of FL are as given :

  • Self-propelling Systems
  • Electronic product for Consumer
  • Domestic Goods
  • Environment Management

For nonlinear systems and multiple input and output systems, FL are quite suitable. It can accommodate fairly number of inputs and outputs easily. It systems are difficult to model, FL is here to work. FLs seem difficult to the complexity of the math behind.

Benefits of Fuzzy Logic system:

  • Generally, vague, noisy and contorted information can regard as input.
  • Also, FL are easy to constitute and intelligently understandable.
  • Primarily, easy solution for the complicated problem such as medicine.
  • FL can be related with mathematical concept
  • We can incorporate and remove rules in system due to the flexibility of FL

Examples from real world:

Nearly, any control mechanism can be altered with FL rooted control system. Although it might be overkill in many domains but complicated system can be modeled easily. FL cannot be answer to all the things but should be used to provide better control. If some PID control goes fine them controller with FL mechanism will be useless. To design a complex system, FL can be a better option.

Sendai underground system:

Sendai underground system in japan is famous applications of FL. This control system used a fuzzy control mechanism to drive the train all day which made the smoothest underground system and also improved the efficiency.

One of the great application of FL control have visible into consumer appliances. Specifically to ACs and heater systems. FL thermostats is being used in such systems to manage the temperature of the place which saves power consumption and makes the system efficient.

Figure 3 : Example

Automation industrial units:

Another important field of application of FL is automation industrial units. FL based PLCs are manufactured by companies. These PLCs, as well as other implementations of fuzzy logic, can be used to control any number of industrial processes. With the use of FL based PLCs, a controller can manage multiple industrial processes.

Animation of 3D content:

FL is also being used for animation of 3D content for producing crowds. This technique had also used for the production of animated movies The Lion and Lord of the Rings trilogy

As a concluding example of FL, it can be utilized in the fields not limited to control management. In signal processing or data analytics, FL Fuzzy logic can be incorporated in decision making process. An example of this is a fuzzy logic system that analyzes an energy system and medical diagnosis harmonic disturbance for that matter. Analysis of voltage, as third, fifth, seventh, ninth harmonics and temperature to estimate any potential concern.

Conclusion:

To conclude this, we can use FL to get AI in the system related to multiple domain. FL is reasoning approach, so any system which formulates logical reasoning, can utilize this approach. It is a flexible approach to design complex system.

You may also know Approach towards Self driving in Artificial Intelligence.

LEAVE A REPLY

Please enter your comment!
Please enter your name here

12 − 4 =