Neural network based gait phases of above knee prosthesis

Abstract

Two gait phase estimation method to control the above knee prosthesis is discussed in this paper. A rule base quantization and an ANN based system is preferred for controlling various parameters such as motion, torque required in walking with the help of prosthetic leg. Microcontroller based semi-active knee prosthesis in order to respond patients demands and adapt environmental conditions such as whether are considered. A design is suggested to measure experimental environment in which gait data is collected for both inertial as well as image based measurement systems. The inertial measurement system consist of MEM accelerometers as well as gyroscopes to identify direct motion measurement of controlling parameter using microcontroller. The image based measurement system is used to verify the above measured data from the prosthetic leg. Various advantages of proposed system is discussed in this paper.