Simulated Annealing (SA) has been used successfully with a broad spectrum of optimization problems, including the optimization of Fuzzy Logic Systems (FLS). A suitable formation of SA, however, is dependent on the selection of an appropriate cooling process and initial temperature. Previous attempts to optimize FLS using SA appear to have been based only on static cooling schedules, with little insight into the cooling schedule rule. The problem of determining the best cooling schedule should aid in the development of a suitable solution for balancing speed and performance when training fuzzy logic systems. In this study, we examined the application of three static cooling schedule variants and two variants of adaptive cooling schedules to determine how far the choice of cooling schedules can affect the final modeling accuracy of FLSs. Two widely-used benchmark problems in the field of fuzzy logic served as the basis for the study's application of these schedules. The results show that the choice of adaptive cooling schedules provides extra improvement to modeling accuracy.