Regarding the problem that the bionic suspension pendulum robot can not grasp the target rod reliably in continuous brachiation process, we propose a staged control strategy, which includes an optimization method to maximize potential energy by adjusting the robot arms’ configuration, a swinging motion optimization method, and a big damping switch and grasp method. For each motion stage, we deduce the desired trajectory of the PID controller of each active joint and the smooth switching condition. Simulations of continuous brachiation between bars with different distance are conducted in the ADAMS software and the MATLAB/Simulink software. Then brachiation experiments are conducted by using the BARDAH robot developed by our lab. In the experiments, the initial bar distance is 1 m and the target bar distance is respectively chosen from 0.5m, 0.6 m, 0.7 m and 0.8m. For each target bar distance, three experiments are repeated and the success rate of the total 12 brachiation experiments is 100%, which indicates that our control method can effectively improve the success rate of the robot continuous brachiation motion.