Abstract: Aiming at the limitations of current time-optimal trajectory planning algorithms for robots, such as low solution efficiency and incomplete constraints, we propose an improved time-optimal trajectory planning algorithm based on reachability analysis. Based on path discretization, the algorithm transforms the entire planning problem into a multi-stage iterative evaluation problem and efficiently executes it. The results of the algorithm are in line with the actual motion performance of robots under the promise of ensuring the time-optimal characteristic as a result of taking complete dynamics constraints, including friction and kinematic constraints, into account. Numerical simulation experiments confirm the feasibility and efficiency of the algorithm.