Robotic assistive ankle-foot prosthesis based on sensorless fuzzy logic control system / Mouaz Al Kouzbary

• Main Author: Mouaz , Al Kouzbary
• Format: Thesis
• Published: 2019
• Subjects: TA Engineering (General). Civil engineering (General)

In the last decade, design and development of lower-limb robotic prostheses were arisen. The electrical actuated prostheses were the optimal solution when torque generation and portability are considered. A great deal of prototypes with wide variety of operating mechanisms were introduced, the elastic component (spring) was the backbone of all prostheses’ mechanical models. Springs are able to store energy during early stance phase and release it prior swing phase, which reduce system power consumption. A proper control system which can mimic biological ankle-foot behavior in different operation condition is one of the most challenging tasks in powered ankle-foot prostheses’ design. The most common approach based on two-level control system; the high-level controller is a finite state machine which regulates the low-level controller behavior accordingly to gait cycle’s feature. In contrast to high-level controller, there is no common approach for low-level controller, and many control strategies were introduced to imitate ankle-foot behavior. The main objective of this dissertation is to assess the hypothesis that an advanced control system without state switching could enhance the powered ankle-foot performance, and restore symmetric characteristic for gait cycle. To evaluate the hypothesis a fuzzy logic control system was developed, the fuzzy inference system’s knowledgebase was constructed after analyzing ankle-foot behavior during walking gait. The control systems together with the dynamic model of powered ankle-foot based on series elastic actuator were modeled using C-code and tested in MATLAB/SIMULINK. The fuzzy logic controller was able to provide a control law to compensate the effect of environment interaction torques, which influenced the powered ankle-foot’s performance with finite state machine impedance base control system. Moreover, the proposed controller eliminates phase shift in powered ankle-foot respond.