Magnetorheological (MR) fluid has recently been used in a variety of applications, including vibration control, prostheses, and tactile devices. A magnetic field is used to control the rheological properties of the MR fluid, allowing for a high-pressure drop, high shear stress, and reversibility of these features. The modifiability imparts smartness to the material, and recent studies have investigated its behaviour extensively. Active prosthetic ankles represent a critical advancement in prosthetic technology, aiming to enhance user comfort and mobility. This seminar will present a novel computational approach that integrates musculoskeletal modeling, control system design, and biomechanical analysis within the OpenSim framework to achieve the design and validation of a Smart Powered Prosthetic Ankle.
The work involves the design and control of an ankle-foot prosthesis using an MR damper and a series elastic actuator (SEA). The MR damper absorbs the vibration during the heel strike event of the stance phase of the gait cycle and the SEA provides desired actuation at the ankle at push-off. Together, the system provides 40 degrees range of motion to help tackle walking in different terrain. Firstly, SEA parameters have been optimized to get the desired ankle moment for human walking. Secondly, the geometric dimensions of the MR valve are optimized by integrating a multi-objective genetic algorithm (MOGA) in MATLAB and FEMS in ANSYS-APDL software. In the coupling method of optimization, the simulation time was more than eight hours. To reduce that, an Artificial Neural Network (ANN) model was utilized to forecast the magnetic flux density (MFD) across different regions of the MR valve. Subsequently, by predicting the value of MFD using ANN to calculate the objective function, GA has been used to optimize the MR damper. Furthermore, to validate the optimal design parameters in real-life scenarios, an innovative model for the human-prosthesis interaction was implemented utilizing OpenSim and the results match the simulated MATLAB results.