A scheme of investigating the intracellular metabolic fluxes in central metabolism of Saccharomyces cerevisiae based on isotope model and tracer experiment was developed. The metabolic model applied in this study includes the Embden-Meyerhof-Parnas pathway, the pentose phosphate pathway, the tricarboxylic acid cycle, CO2 anaplerotic reactions, ethanol and acetate formation, and pathways involved in amino acid synthesis. The approach of hybridized genetic algorithm combined with the sequential simplex technique was used to optimize a quadratic error function without the requirement of the information on the partial derivatives. The impact of some key parameters on the algorithm was studied. This approach was proved to be rapid and numerically stable in the analysis of the central metabolism of S.cerevisiae.