Research on an Improved Genetic Algorithm Based Knowledge Acquisition

LM Su,H Zhang,CZ Hou,XQ Pan

DOI: https://doi.org/10.1109/icmlc.2002.1176795

2002-01-01

Abstract:Based on an optimization model, an improved genetic algorithm applied to knowledge acquisition of a network fault diagnostic expert system is described. The algorithm applies operators such as selection, crossover and mutation to evolve an initial population of diagnostic rules. Especially, a self-adaptive method is proposed to regulate the crossover rate and mutation rate. Finally, a knowledge acquisition problem of a simple network fault diagnostic expert system is simulated, and the results of simulation show that the improved approach can solve the convergence problem better.