J. Kinzel, F. Klawoon, and R. Kruse. Modifications of genetic algorithms for designing and optimizing fuzzy controllers. In Proc. First IEEE Conference on Evolutionary Computation (ICEC'94), pages 28--33, Orlando, FL, USA, 1994.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....problems. Fuzzy logic controllers tuned by genetic algorithms. Some fundamentals of theory and description of individual components, as well as the early implementations of fuzzy logic controllers tuned by genetic algorithms can be found in a series of publications [57] 68] 69] 70] 71] [72], 61] 22] 73] 74] and [75] Fuzzy logic controllers learning by genetic algorithms. In [76] two different approaches to apply genetic algorithms to fuzzy logic controllers are described. The first approach uses the knowledge base as the individual of the genetic system, while the second ....

Kinzel, J., Klawoon F. and Kruse R., Modifications of genetic algorithms for designing and optimizing fuzzy controllers, Proc. 1 IEEE Conf. on Evol. Computation, ICEC'94, 28-33, 1994

....used in the rule antecedent (state vector) and consequent (control action) The population was the concatenation of all rules so represented. A customized (max min arithmetical) crossover operator was also proposed. The fitness function was a sum of quadratic errors. Kinzel, Klawon and Kruse (Kinzel et al. 1994) tuned both rules and termsets. They departed from the string representation and used a (cross product) matrix to encode the rule set (as if it were in table form) They also proposed customized (point radius) crossover operators which were similar to the twopoint crossover for string encoding. ....

, pages 28--33, Orlando, FL., USA, 1994.

....of the membership values used in the rule antecedent (state vector) and consequent (control action) The population was the concatenation of all rules so represented. A customized (max min arithmetical) crossover operator was also proposed. The fitness function was a sum of quadratic errors. Kinzel et al. 1994) tuned both rules and termsets. They departed from the string representation and used a (cross product) matrix to encode the rule set (as if it were in table form) They also proposed customized (point radius) crossover operators which were similar to the two point crossover for string encoding. ....

Kinzel, J , Klawoon, F. , and Kruse, R. (1994). Modifications of genetic algorithms for designing and optimizing fuzzy controllers. In Proc. First IEEE Conf. on Evolutionary Computing (ICEC'94), pages 28--33, Orlando, FL., USA.

....a string of numerical genes whose alleles belong to binary or n ary alphabets, and or the set of real or natural numbers, rather than the collection of linguistic variables, fuzzy sets and fuzzy logic connectives that actually make up the rule. An exception is the representation proposed in [12] where a chromosome is encoded as a matrix whose elements (alleles) are fuzzy sets. Furthermore, for approaches that use the simple GA [13] the fixed length chromosome restricts each individual to have the same pre specified number of rules. The contention here is that the genetic programming ....

....complex behavior. This view is shared by Feldman [11] who has developed a technique that encodes fuzzy control rules as a fuzzy network, a connectionist extension to fuzzy linguistic systems. The GA is used to synthesize or modify the rules of the fuzzy network controller. Finally, Kinzel et al. [12] deemed it necessary to modify the GA (using the matrix rule base representation mentioned earlier) by taking the properties of fuzzy controllers into account to facilitate fast convergence. As a departure from the Darwinian approach Grefenstette [20] added Lamarckian mechanisms to the SAMUEL ....

Kinzel, J., F. Klawonn and R. Kruse "Modifications of genetic algorithms for designing and optimizing fuzzy controllers", 1st IEEE Conference on Evolutionary Computation, Orlando, FL, pp. 28-33, June 1994.

....during the last years since it does not require the difficult knowledge acquisition task from scratch. There are of course a number of approaches to learning fuzzy rules from data based for instance on techniques of neural (for an overview see [18] or evolutionary computation (see for example [9]) mostly aiming at optimizing certain parameters of a fuzzy controller. However, fuzzy clustering seems to be a very appealing method for learning rules since there is a close and canonical connection between fuzzy clusters and fuzzy rules. Intuitively, each if then rule of a Mamdani type fuzzy ....

J. Kinzel, F. Klawonn, R. Kruse, Modifications of Genetic Algorithms for Designing and Optimizing Fuzzy Controllers. Proc. IEEE Conference on Evolutionary Computation, IEEE, Orlando (1994), 28--33.

No context found.

J. Kinzel, F. Klawoon, and R. Kruse. Modifications of genetic algorithms for designing and optimizing fuzzy controllers. In Proc. First IEEE Conference on Evolutionary Computation (ICEC'94), pages 28--33, Orlando, FL, USA, 1994.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC