In this paper we present a hierarchical planning architecture that enables a quadruped to walk over roughterrain.The planning architecture consisting two layers,one is a high-level planner that plans a set of footsteps across the terrain and the other is a low-level planner that plans trajectories for the robot's feet and center of gravity.In the high-level planner,we plan the reasonable path for the quadruped by A* algorithm and in the low-level planner,we plan the trajectory for the robot's feet by the cubic spline curve and optimize it by the convex optimization.In the MATLAB simulation, we validate the algorithm in different terrain,
and the experimental results show that the algorithm can make the robot smoothly through rough terrain.