Adaptive neurofuzzy inference system for an active transfemoral prosthetic leg using in-socket sensory system / Nur Hidayah Mohd Yusof

• Main Author: Nur Hidayah , Mohd Yusof
• Format: Thesis
• Published: 2019
• Subjects: T Technology (General)
• Online Access: http://studentsrepo.um.edu.my/10599/1/Nur_Hidayah_Mohd_Yusof.jpg; http://studentsrepo.um.edu.my/10599/8/hidayah.pdf; http://studentsrepo.um.edu.my/10599/

Prosthetic leg is known as one of the solutions to help lower limb amputees to regain their ambulation ability. However, most of the current existing knee components still lack in the ability to provide active body propulsion, which in turn results in higher metabolic energy consumption required by the amputee in doing locomotion movement. This study was motivated by the idea of developing a microprocessor controlled prosthetic leg with user intention recognition for transfemoral amputees which is able to mimic a manner of ambulation close to that of healthy individuals. In particular, this study focused on developing the control input framework in order to derive the used intention, as read by the in-socket sensory system, as control input into the main controller. The first part of this work is the development of an Adaptive Neurofuzzy Inference System (ANFIS) - based algorithm to human gait phase recognition. The gait phase is necessary for cadence and torque control required by the knee joint mechanism. In the second part, the sensing method was adopted using the in-socket sensors embedded into the transfemoral socket of the prosthetic leg. A transfemoral amputee walked with in-socket sensorised prosthetic leg performed several gait cycles and the in-socket sensor signals were used as an input into the newly developed ANFIS system. Physical simulation of the controller presented a realistic simulation of the actuated knee joint in terms of knee mechanism, proving that the novel ANFIS system was validated with real amputee experimental signals. The fuzzy system successfully replicated human gait cycle by categorizing the cycle into seven gait phases. In conclusion, this study had established an ANFIS-based control input framework using in-socket sensory system based on amputee user’s seven phases of gait cycle.