Fuzzy logic gives computers the ability to emulate human reasoning and solve certain types of problems efficiently. Coined in 1965, the word "fuzzy" refers to the imprecise type of logic used to manage real-world tasks. This imprecision, which may seem like a limitation, is actually a benefit for computers used in fields like engineering, process control, medicine and other fields where fuzzy logic may work better than regular logic. There are, however, limitations to fuzzy logic technology.
PCs, Web servers and most devices you work with use binary logic. Microchips, which power computers, work like light switches. A light can be on or off depending on its switch's position. If you could see what's going on inside your computer, you would discover numbers similar to those shown here:
Computers that use binary logic can only recognize ones and zeros. When a computer sees a "1," it considers that to be a switch that is on. A zero means that the switch is off. By processing these switches, a computer can display a text file, play a movie or draw an image.
Fuzzy Logic and Approximation
Computers using binary logic are precise. If you ask one to add 10 and 20, it will tell you that the sum is 30. Thirty is not an approximation but instead an exact data value. Computers that use fuzzy logic approximate values and may work with ideas instead of numbers. Instead of being pure black or pure white, for instance, a variable in a fuzzy logic program may be a shade of gray. While regular computers might understand terms such as hot and cold, a fuzzy logic computer may have the ability to work with intermediate temperatures that lie between hot and cold. Some washing machines even use fuzzy logic to manage their washing cycles.
Fuzzy Logic Processing
Because fuzzy logic computers deal with ranges of possible values instead of precise numbers, they can evaluate conditions in a way that emulates human logic. For example, a computer using fuzzy logic might create a set of processing rules similar to the following:
When event1 happens, perform task 1
When event2 happens, perform task 2
Suppose you wanted to compute a restaurant tip by assuming that it equals 15 percent of the bill. In this example, the computer might consider event1 to be "superior service" and add an extra percentage to the bill. Event2 might represent "superior food," another fuzzy concept. The computer could then come up with a final amount using logic that a human might use when calculating a tip.
Because fuzzy logic computers use approximations, they do not make good candidates for managing systems that require extreme precision. You might be concerned, for example, if a computer managing your bank account approximated your checking account balance. Computers that use fuzzy logic do not have the ability to learn and adapt after solving a problem as some expert systems can. Another limiting factor is the inability of fuzzy logic to solve problems when no one knows the solution. Experts must exist who know how to create the rule sets needed to make a fuzzy logic system work. If you cannot find an expert, you cannot create a fuzzy logic controller. Fuzzy logic systems may also be expensive to develop because they often require extensive testing.
- Production Engineering and Management Under Fuzziness; Cengiz Kahraman, Mesut Yavuz
- Center for Spatially Integrated Social Science: CSISS Classics - Lotfi Zadeh - Fuzzy Logic-Incorporating Real World Vagueness
- PAControl.com: Fuzzy Logic
- Fuzzy Logic for Embedded Systems Applications; Ahmad M. Ibrahim
- Safari Books Online: Knowledge-Based Systems > Fuzzy Logic > 5.16 LIMITATIONS OF FUZZY SYSTEMS
- LG Electronics Australia: Top Loader Washing Machine - WT-H650
- Photo Credit Andrew Olney/Digital Vision/Getty Images