In optical communications systems, the optical combs spectrum has been used to generate multiple carriers from a continuous wave laser or other optical source. Due to different parameters to find the best spectrum, one alternative is necessary to define. This paper proposes three fuzzy entropy measures by adequacy and non-adequacy in order to evaluate the optical combs spectrum performance. The goal is to group a set of data of optical combs spectra using fuzzy clustering, specify LAMDA, and to calculate the amount of information of each class. The fuzzy measures are used to analyze the phase and voltage bias variations for the second arm of Mach-Zehnder modulator and the changes of RF signal frequency in both arms. Obtained results show little changes of fuzzy entropies by adequacy average for phase and bias voltage cases, but an increase for frequency variation cases, as expected. With this result, the approach could be applied to the stability detection in Optical Frequency Comb Generation (OFCG) used for high capacity optical systems.