The Active Above-Knee Prosthesis Gait Control Based on the EMPC

On-line optimization control has been recognized as a superior algorithm for constrained systems. To deal with various disturbances and serious nonlinearity, active above-knee prosthesis (AKP) is treated as constrained switching system in this paper. The kinetic characters of AKP were analyzed when an amputee was walking on level-ground. A piecewise affine model can be built, and the AKP system was divided into 15 subsystems. Explicit model predictive control algorithm was applied to switch between subsystems. State-space partitions and explicit control laws are calculated off-line. On-line control signal can be obtained by looking up table. This algorithm can make the prosthesis to track the flexion curve of health limb accurately. Finally, results are shown that this method can tune the control impedance for powered knee prostheses with less energy consumption and high operation speed.