FSM-Based Volitional Control of Robotic Transfemoral Prosthesis for Walking on Uneven Ground

To effectively control the robotic transfemoral prosthesis equipped with active actuators, it is necessary to recognize the user's state and intentions and generate an appropriate joint trajectory. In this paper, a walking trajectory generator based on a finite-state machine (FSM) is proposed as a simple algorithm to implement walking, starting, and stopping motions. We define the thigh and knee angles in a one-to-one correspondence by dividing the walking cycle into several states and using FSM controller. Through this, it is possible to linearize the relationships between the thigh and knee, and by setting gait trajectory generation parameters, it is possible to generate the knee angle according to the thigh angle and the FSM state. In addition, a variable impedance control that performs actual ankle movements to enable walking on uneven ground. Finally, the proposed method was successfully verified through walking experiments with the robotic transfemoral prosthesis on treadmill with ground obstacles.