Recently, robot assisted therapy devices are increasingly used for spinal cord injury (SCI) rehabilitation in assisting handicapped patients to regain their impaired movements. Assistive robotic systems may not be able to cure or fully compensate impairments, but it should be able to assist certain impaired functions and ease movements. In this study, the control system of lower extremity orthosis for the body weight support gait training system which implements pneumatic artificial muscle (PAM) is proposed. The hip and knee joint angles of the gait orthosis system are controlled based on the PAM coordinates information from the simulation. This information provides the contraction data for the mono- and bi-articular PAMs that are arranged as posterior and anterior actuators to simulate the human walking motion. The proposed control system estimates the actuators' contraction as a function of hip and knee joint angles. Based on the contraction model obtained, input pressures for each actuators are measured. The control system are performed at different gait cycles and two PMA settings for the mono- and bi-articular actuators are evaluated in this research. The results showed that the system was able to achieve the maximum muscle moment at the joints, and able to perform the heel contact movement. This explained that the antagonistic mono- and bi-articular actuators worked effectively.