Asteroids, as the primitive building blocks for the formation of our solar system, could reveal its evolution mechanism, and have attracted more and more attention from the public and professional institutions in recent years. Their physical properties, such as rotational period, spin axis and overall shape, can be inverted from ground- and space-based photometric observations. Since the inversion process is very time-consuming, this paper combines the genetic algorithm with the Levenberg–Marquardt (LM) algorithm, and presents a hybrid optimization algorithm based on a Cellinoid shape model for the inversion of rotational periods, which greatly improves the inversion efficiency. The proposed hybrid algorithm is applied to the synthetic lightcurves generated for an assumed Cellinoid shape model and the inverted rotational period results are consistent with the preset ones with a reduced search time, compared with the LM algorithm. Finally, multiple numerical experiments on the periods are performed on lightcurves and sparse observations of real asteroids to confirm that the proposed method can perform well in improving computational efficiency.