As I reflect on my graduate career, I am deeply grateful to the many people whose help, insight, and guidance have led me to where I am today. All members of the Center for Intelligent Mechatronics during my time were integral and deserve special recognition. The Lab fosters a collaborative environment that allows all members to push their boundaries.
The most advanced microprocessor knees available on the market include Ossur's Rheo knee (left), Otto Bock's C leg (middle), and Freedom Innovations' Plié knee (right). Proficient test adapter used in the development of prostheses and pre-controllers. References and Current Power Prosthetic Knee and Ankle Joint Rotations for a Step in.
Overview Overview
Design and Control of an Active Electrical Knee and Ankle Prosthesis
This article provides an overview of the design and operation of an electrically powered knee and ankle prosthesis. A valid test adapter used in the development of the prosthesis and controllers prior to participation of transfemoral amputees. References and actual knee and ankle joint moments of the powered prosthesis for one step at normal speed.
One of the primary limitations of the electrically powered knee and ankle prosthesis design is the power source. Measured mechanical strength of the prosthesis for ten consecutive walking cycles of treadmill running at slow speed. 20] Sup, F., Bohara, A., Goldfarb, M., “Design and Control of a Powered Transfemoral Prosthesis,” The International Journal of Robotics Research, pp.
Preliminary Evaluations of a Self-Contained Anthropomorphic Transfemoral Prosthesis
A finite-state-based impedance control method, previously developed by the authors, is used to control the prosthesis during walking and standing. Knee and ankle joint angle, torque and force data taken during walking experiments at various speeds demonstrate the ability of the prosthesis to provide a functional gait representative of normal gait biomechanics. Unlike any of the aforementioned previous works, this paper describes a transfemoral prosthesis with both a powerful knee and ankle.
The joint torque specifications required of the knee and ankle joints were based on an 85 kg user for a walking cadence of 80 steps per minute, Fig. 3-5, is designed to measure the ground reaction force components at the ball of the foot and heel. The general control architecture of the prosthesis consists of three layers, as diagrammed in Fig.
The subject's self-selected normal cadence was determined to be 75 steps per minute at 2.8 km/h. The interlayer control parameters of the motorized prosthesis were then set (with the prosthesis controlled in the tethered state) while walking on the treadmill at slow, normal and fast gait cadences (as determined by the daily use prosthesis), and also while standing. The resulting mid-layer controller parameters for the set impedance work for standing and for the various.
Measured joint angles, torques and forces of the motorized prosthesis for ten consecutive walking cycles at a self-selected speed (5.1 km/h at 87 steps per minute). References and actual torques of the knee and ankle joints of the motorized prosthesis for one step at a self-selected speed (5.1 km/h at 87 steps per minute) on normal ground. One of the main limitations in the design of electrically powered knee and ankle prostheses is the power source.
Measured electrical and mechanical power at the knee and ankle joints of the powered prosthesis over one walking cycle at self-selected speed (5.1 km/h at 87 steps per minute) on normal ground. To characterize battery requirements, the average electrical power required by prosthesis (i.e., the embedded system, knee joint, and ankle joint) during standing and walking over level ground (at the self-selected speed of 5.1 km/h) is shown in Fig . Average electrical power consumption of the powered prosthesis for standing and walking at self-selected speed (5.1 km/h at 87 steps per minute) on normal ground.
![Figure 3-1. Normal biomechanical gait data for an 85 kg subject walking at a cadence of 80 steps per minute [9]](https://thumb-ap.123doks.com/thumbv2/123dok/10740938.0/43.918.226.691.507.852/figure-normal-biomechanical-subject-walking-cadence-steps-minute.webp)
Powered Sit-to-Stand and Assistive Stand-to-Sit Framework For a Powered Transfemoral Prosthesis
State-of-the-art microcontroller-modulated braking knees adjust the damping at the knee joint to control resistance during the transition from standing to sitting [7]. In [11], an agonist-antagonist knee design is presented that exploits the passive dynamics of the knee during walking. The mid-level controllers generate torque references for the joints using a finite state machine that modulates the impedance of the joints depending on the phase of the activity.
In this work, the design of the supervisory intention recognizer and the finite state impedance based controller for sitting and standing mode will be presented. In the weight-bearing phase, the user's weight is supported on the joints with high impedance. The transition phases, sit-to-stand and stand-to-sit, modulate knee stiffness as a function of knee angle.
The database generation for the landing mode was more complicated, since the limited-state-based finite-impedance controller for landing includes SU and SD transitions. The parameterization of GMMs for all desired modes of activity is achieved based on the training data in one. The activity mode goal recognizer is a component of the supervisory controller for the powered prosthesis and has two performance objectives.
The combination of the voice length, l, and the frame length, f, determines the delay of activity mode intent recognition. The respective locations of the two different activities in the reduced feature space can be seen in this figure. The knee prosthetic angle and the activity mode for one of the trials are shown in Fig.
Further work includes a comprehensive biomechanical evaluation of the sit-to-stand and stand-to-sit control assist frame in multiple amputee subjects.
Slope Ascent with a Powered Knee and Ankle Prosthesis
Third, an overview of the powered prosthesis used in this study and the experimental protocol is presented. In finite-state impedance-based control, the impedance behavior of healthy biomechanical gait is mimicked by modifying the joint impedances of the prosthesis according to the gait phase. The length of Phase 0 is directly related to the threshold value for contact with the ball of the foot.
Due to the similarity of the gait patterns for walking on level ground and uphill, the previously developed gait impedance control structure is used for walking uphill. All experiments used the subject's daily use socket, with the prototype powered prosthesis attached in place of the daily use prosthesis. The impedance parameters of the powered prosthesis are tuned using a combination of user feedback and visual inspection.
In addition, the device's electrical energy consumption in the knee and ankle was measured via built-in sensors. Average joint angles, torques, and power of the motorized prosthesis for ten consecutive gait cycles at self-selected speed for level, 5, and 10 degrees of gait. Measured joint angles of the motorized prosthesis for five consecutive gait cycles at a self-selected speed for a 10-degree uphill walk.
The average electrical power required by the prosthesis (i.e. embedded system, knee joint and ankle joint) during level, 5 and 10 degree uphill walking at a self-selected cadence is shown in the figure. Average power consumption of a powered prosthesis for walking level, 5 degrees and 10 degrees up at a self-selected speed on normal ground. This work presented an extension of the hill-climbing control framework for an electric knee-ankle prosthesis developed by the authors in previous works.
Future work includes extensive biomechanical evaluation of the motorized prosthesis and the incline ascent controller on multiple amputees.
Contributions and Future Work
Describes the extension and refinement of the finite-state impedance control structure to enable walking in three cadences: standing, ramp ascent, sitting, and associated sit-to-stand transitions. Presents the validation of the device on a treadmill and on a flat surface, tested by a unilateral transfemoral amputee and demonstrating the ability to restore a near-normal biomechanical gait pattern. The physical hardware consists of the mechanical and electrical components that make up the powered prosthesis.
An important requirement of the current design is that a gear ratio of the order of 160:1 is required to generate the required torques at the connections. To overcome this problem, a useful area of research is to develop a compact servo amplifier that is better matched to the state-of-the-art brushless DC motors that are tuned for quiet operation. To increase the population of amputees assisted by such a powered prosthesis, modularity of design should be pursued.
Development of separate powered knee and ankle devices that could work both together and independently could better meet the needs of the transfemoral and transtibial amputee communities. A disadvantage of the current finite-state control method is the significant amount of manual parameter setting. To further increase the efficiency of the tuning process, a more quantitative means of addressing the performance of the sound side limb and whole body mechanics to provide a more optimal set of impedance parameters.
Another area is the development of error detection algorithms to ensure proper operation of the device and prevent user injury. A short-term study is planned to assess the amputee's biomechanics, as used in Manuscripts 2 and 3 and Chapter V to study the effects of the powered knee and ankle prosthesis on the amputee's gait. The measures to be investigated are the walking symmetry at three walking speeds (self-selected speed and ±15% of the self-selected speed) in terms of joint angles, moments, forces and stride length and walking posture determined by the movement of the trunk.
Long-term future areas of research include assessing the benefits and effects of a powered knee and ankle prosthesis on walking and general mobility.
