High-speci c-impulse electric propulsion technology is promising for future space robotic debris removal in sun-synchronous orbits. Such a prospect involves solving a class of challenging problems of low-thrust orbital rendezvous between an active spacecra and a free-ying debris. This study focuses on computing optimal lowthrust minimum-time many-revolution trajectories, considering the e ects of the Earth oblateness perturbations and null thrust in Earth shadow. Firstly, a set of mean-element orbital dynamic equations of a chaser (spacecra ) and a target (debris) are derived by using the orbital averaging technique, and speci cally a slow-changing state of the mean longitude di erence is proposed to accommodate to the rendezvous problem. Subsequently, the corresponding optimal control problem is formulated based on the mean elements and their associated costate variables in terms of Pontryagin's maximum principle, and a practical optimization procedure is adopted to nd the speci c initial costate variables, wherein the necessary conditions of the optimal solutions are all satis ed. A erwards, the optimal control pro le obtained in mean elements is then mapped into the counterpart that is employed by the osculating orbital dynamics. A simple correction strategy about the initialization of the mean elements, speci cally the di erential mean true longitude, is suggested, which is capable of minimizing the terminal orbital rendezvous errors for propagating orbital dynamics expressed by both mean and osculating elements. Finally, numerical examples are presented, and speci cally, the terminal orbital rendezvous accuracy is veri ed by solving hundreds of rendezvous problems, demonstrating the e ectiveness of the optimization method proposed in this article.