Design & evaluation of a sensor minimal gait phase and situation detection Algorithm of Human Walking

A Prosthetic Shank With Adaptable Torsion Stiffness and Foot Alignment

“…Basic tests with non-amputees showed individual detection rates between 91.4 and 100 % in a course around obstacles. Functional tests with participants with amputation confirm the reliable applicability of the detection algorithm (Schuy, 2016), which is also underlined by a previous study considering 3, 000 steps of 15 able-bodied personens and 2 persons with amputation on a predefined parcours achieving comparable detection rates distinctly above 90% and similar reliability (Schuy et al, 2015).…”

“…A gait detection algorithm processes signals from an inertial measurement unit (BNO 055, Robert Bosch GmbH, Gerlingen, Germany) implemented in the structure of the prosthesis. In preliminary investigations with healthy users (Schuy et al, 2015), angular shank velocity in sagittal plane ω ML and frontal plane ω AP acquired from gyroscopes were sufficient to determine gait direction, detect stance and swing phase and estimate gait velocity. Gait phase estimation is based on the identification of characteristic events of stance and swing phase.…”

“…The neural network facilitates considering motor inertia as well as nonlinear effects such as friction and hysteresis as identified in bench-testing. Ten hidden neurons were fed with mid swing and heel strike amplitude date and trained using a Levenberg-Marquardt algorithm (Schuy et al, 2015). Training and evaluation was performed with different data subsets from the preliminary study by Schuy et al (2015) considering different gait situations and speeds.…”

“…Ten hidden neurons were fed with mid swing and heel strike amplitude date and trained using a Levenberg-Marquardt algorithm (Schuy et al, 2015). Training and evaluation was performed with different data subsets from the preliminary study by Schuy et al (2015) considering different gait situations and speeds. The current control of the motor controller EPOS 24/5 (Maxon Motor AG, Sachseln, Switzerland) is tuned to generate the motor torque τ m based on the input τ s while considering the gear ratio.…”

A user-specific human-machine interaction strategy for a prosthetic shank adapter

“…The high-level control strategy contains a sensorminimal approach to determine the gait parameters [9]. A fuzzy-based evaluation of the shank velocity signals of an inertial measurement unit in the sagittal and frontal plane is developed to detect type and direction of motion as well as velocity [10]. The high-level control scheme detects the gait parameters and transfers the corresponding optimal torsional stiffness and foot alignment to the low-level control algorithm.…”