Improving the performance of professional cyclists during individual time trials has been a challenge for several decades. In this study, the fatigue power profiles of male and female professional cyclists were characterized and a novel optimal pacing strategy model was established to determine the output power distribution based on the simulated annealing algorithm. Moreover, the model was implemented on the male 2021 UCI ITT course and the female 2021 Olympic ITT course. Compared with the actual performance of the professional cyclists, the race scores predicted by this model are among the best, which indicates that the model is highly practical and instructive. Considering riders are unlikely to follow the detailed plan exactly and their output power can deviate from the optimal power, a sensitivity analysis was conducted and the results demonstrated the high stability and reliability of the model.