5.5 Review/Preview

In this chapter, a simple GA made up of reproduction, crossover, and mutation was used to perform least-squares curve-fitting. Three curve-fitting problems were presented in which a GA rapidly converged to near-optimal solutions. The solutions were comparable to, or better than, solutions determined by more commonly used methods. Although they require some knowledge of the problem, GAs are readily adaptable to least-squares curve-fitting problems.

Determining the solution to least-squares curve-fitting problems is a challenging and worthwhile endeavor. However, we are more interested in using a GA-based search technique to select effective rules and tune membership functions in the fuzzy control and modeling systems we wish to implement. It is not readily apparent at this point how we can use a GA to accomplish these tasks. There are several points that must be addressed for the reader to be able to design and tune a fuzzy system using a GA.

The next three chapters will address many of the relevant issues that arise when using GAs to design fuzzy systems. Chapter 6 revisits the liquid level system first introduced in Chapter 2. The fuzzy controller is improved with a GA. Chapters 7 and 8 revisit the cart-pole system and the rendezvous system, respectively. These three chapters provide information on the techniques that can be employed when using a GA to select fuzzy rules, to alter triangular and trapezoidal membership functions, and to determine real-valued consequents. Then, Chapter 9 will change the game dramatically when the systems to be controlled are made more complex.

References

Davis, L. D. (1991) (Ed.), The genetic algorithms handbook. New York, NY: Van Nostrand Reinhold Company.

De Jong, K. A. (1975). Analysis of the behavior of a class of genetic adaptive systems. Dissertation Abstracts International, 36(10), 5140B.

Goldberg, D. E. (1983). Computer-aided gas pipeline operation using genetic algorithms and rule learning, Ph.D. Thesis, University of Michigan, Ann Arbor, MI.

Goldberg, D. E. (1989). Genetic algorithms in search, optimization, and machine learning. Reading, MA: Addison-Wesley.

Goldberg, D. E., and Lingle, R. L. (1985). Alleles, loci, and the travelling salesman problem. Paper in Proceedings of an International Conference on Genetic Algorithms and Their Applications, J.J. Grefenstette (Ed.), 154–159.

Goldberg, D. E., and Samtani, M. P. (1986). Engineering optimization via genetic algorithm. Paper in Proceedings of Ninth Conference on Electronic Computation, University of Alabama in Birmingham, 471–482.

Holland, J. H. (1975). Adaptation in natural and artificial systems. Ann Arbor, MI: University of Michigan Press.

Karr, C. L., and Goldberg, D. E. (1987). Genetic algorithms in mineral processing and machine learning. Paper in Proceedings of the Artificial Intelligence in Mineral and Material Technology Conference, University of Alabama, Tuscaloosa, AL, 127–141.

Stanley, D. A., and Scheiner, B. J. (1985). Mechanically induced dewatering of ion exchanged attapulgite flocculated by a high molecular weight polymer. Colloids and Surfaces, 14, 151–159.

Stanley, D. A., Webb, S. W., and Scheiner, B. J. (1986). Rheology of ion-exchanged montmorillonite clays. Bureau of Mines RI 8895.