home *** CD-ROM | disk | FTP | other *** search
- Genetic Adapt Fuzzy Control System
-
- Why Fuzzy
- The real world is not black and white. The day in and day out human
- reasoning is all but EXACT. However, the classical logic and the
- traditional control systems are based on exact reasoning. Fuzzy logic,
- also known as imprecise reasoning, allows us solving problems with fuzzy
- boundaries as we handle daily routines. Fuzzy logic is a formulation of
- logic using mathematical framework to deal concepts such as few, very,
- much, and many. Fuzzy logic can help us make decisions when lack of
- specific facts or the knowledge is imprecise or far from complete.
-
- Human experts can handle situations beyond their knowledge with
- gracefully degraded accuracy, but the traditional control systems
- normally breakdown when the environment is out of their design scope.
- Many environments are difficult to model mathematically, especially for
- systems with uncertainties or non-linearties. When applying
- conventional control logic to these environments, it requires human
- interactions or interventions. On the other hand fuzzy control system
- is not based on strict mathematical model. Not only it is easier and
- quicker to build a control system but it also degrades gracefully.
-
- Apply Genetic Algorithm in Fuzzy Control System
- However, there are drawbacks in fuzzy logic control systems. The fuzzy
- sets and rules must be determined properly for the system to function
- correctly. This introduces a problem for some fuzzy systems and fuzzy
- chips. The problem is that the rules and sets must be coded up front
- instead tuned at site. Another problem is that because of its imprecise,
- it is difficult for users to come up with an optimal solution. As a
- result, users are spending more time in tuning up fuzzy sets and rules.
- While fuzzy logic mimics human's imprecise reasoning, genetic algorithm
- mimics the evolution of the nature. Applying genetic algorithm in
- fuzzy control eliminates problems associated with fuzzy system. The
- genetic algorithm uses randomization techniques to reach local optimum
- without exhausting searches through the solution state space. With
- randomization, genetic algorithm is also capable of escaping from local
- optimum and hopping into a new and better state.
-
- What Is Genetic Adapt Fuzzy Control System
- The Genetic-Adapt Fuzzy control system (GAF) combines powerful
- features from fuzzy logic and genetic algorithm. GAF allows users to
- generate a fuzzy control system by simply defining the inputs, outputs,
- measured data set, and initial rule sets. GAF uses genetic algorithm to
- derive proper rules and fuzzy sets from the initial rules. By changing,
- adding, deleting rules and fuzzy membership sets, the genetic algorithm
- automatically adapts and optimizes the fuzzy control system. Besides
- optimizes fuzzy rules, GAF also provides users the capability to
- simulate and debug their applications. GAF replaces complex
- conventional control algorithm with simple fuzzy rules contained in
- English like text files. It builds run time data directly from the rule
- files to eliminate intermediate compilation and achieve quick turn-
- around time.
-
- By R.P. Huang Jan 16, 1993
-