To improve the detection performance of glucose solutions, we first simulate and calculate the theoretical Raman spectrum of 99% (volume fraction) glucopyranose in the glucose solution. Next, we obtain experimental Raman spectra of glucose solutions at different mass concentrations using a Raman spectroscopy system. The theoretical calculations show that both types of glucopyranose exhibit characteristic peaks at 1125 cm^-1, which is consistent with the experimental results. We then select the 1640 cm^-1 characteristic peak of water as the internal standard peak for normalization. Finally, two quantitative analysis methods, the characteristic peak relative intensity method and the characteristic peak relative area method, are used for linear analysis. The linear correlation coefficient are 0.992 and 0.985, respectively. The lowest detection limits are 48.1 mg/dL and 52.3 mg/dL, and the lowest detection mass concentrations are all 12.5 mg/dL. The results indicate that the relative intensity method is simple and convenient, and provides a higher linear fitting, making it more suitable for accurate quantitative analysis of low mass concentration glucose solutions.