Preliminary results on energy efficient 3D prosthetic walking with a powered compliant transfemoral prosthesis

This paper extends bipedal trajectory tracking methods to prostheses to enable construction of a class of model-dependent prosthesis controllers using locally available sensor information. The rapidly exponentially stabilizing control Lyapunov functions (RES-CLFs) developed for bipedal robots guarantee stability of the hybrid zero dynamics in the presence of impacts that occur in walking. These methods cannot be directly applied to prostheses because of the unknown human dynamics. We overcome this challenge with two RES-CLFs, one for the prosthesis subsystem and another for the remaining human system. Further, we outline a method to construct these RES-CLFs for this type of separable system by first constructing separable CLFs for partially feedback linearizable systems. This work develops a class of separable subsystem controllers that rely only on local information but provide formal guarantees of stability for the full hybrid system with zero dynamics.

show abstract

“…and hence forms a CLF satisfying the sub-CLF conditions (17). We write V sep (x) as a composite CLF of sub-CLFs:…”

Section: Definitionmentioning

confidence: 99%

“…Therefore any set {µ i ∈ R} i=1,...,k that satisfies each respective sub-CLF condition (17), will also satisfy the CLF condition (16) for the separable CLF.…”

Section: Definitionmentioning

confidence: 99%

Separable Control Lyapunov Functions With Application to Prostheses

Gehlhar

Ames

2021

IEEE Control Syst. Lett.

Self Cite

View full text Add to dashboard Cite

show abstract

“…We first present an overview of the methodology for synthesizing PD controllers for powered prostheses -specifically, AMPRO3 (Fig. 1), a powered transfemoral prosthesis developed at Caltech [21]. In particular, this prosthesis with an actuated knee and ankle joint is modeled as a hybrid dynamical system.…”

Section: A Ampro With Pd Controllermentioning

confidence: 99%

“…AMPRO3 Fig. 4 shows the control architecture, including sensing and computation, of AMPRO3 [21]. It is run on a Beagle-Bone Black (BBB) micro-controller on-board the prosthetic.…”

Section: Hardware Implementationmentioning

confidence: 99%

Recurrent Neural Network Control of a Hybrid Dynamical Transfemoral Prosthesis with EdgeDRNN Accelerator

Gao

Gehlhar

Ames

et al. 2020

2020 IEEE International Conference on Robotics and Automation (ICRA)

Self Cite

View full text Add to dashboard Cite

Lower leg prostheses could improve the life quality of amputees by increasing comfort and reducing energy to locomote, but currently control methods are limited in modulating behaviors based upon the human's experience. This paper describes the first steps toward learning complex controllers for dynamical robotic assistive devices. We provide the first example of behavioral cloning to control a powered transfemoral prostheses using a Gated Recurrent Unit (GRU) based recurrent neural network (RNN) running on a custom hardware accelerator that exploits temporal sparsity. The RNN is trained on data collected from the original prosthesis controller. The RNN inference is realized by a novel EdgeDRNN accelerator in real-time. Experimental results show that the RNN can replace the nominal PD controller to realize endto-end control of the AMPRO3 prosthetic leg walking on flat ground and unforeseen slopes with comparable tracking accuracy. EdgeDRNN computes the RNN about 240 times faster than real time, opening the possibility of running larger networks for more complex tasks in the future. Implementing an RNN on this real-time dynamical system with impacts sets the ground work to incorporate other learned elements of the human-prosthesis system into prosthesis control.

show abstract

“…While feedback linearization and CLFs can enforce the trajectories on walking robots [3], [1], these methods cannot be applied to prostheses in the same way because they require full model information. example, in [30] a method was developed to apply CLF-QP to a prosthesis, but in a model independent fashion. Modeldependence is desired in a prosthesis controller to improve tracking performance and robustness to perturbations.…”

Section: Introductionmentioning

confidence: 99%

Data-driven Characterization of Human Interaction for Model-based Control of Powered Prostheses

Gehlhar

Chen

Ames

2020

2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

Self Cite

View full text Add to dashboard Cite

This paper proposes a data-driven method for powered prosthesis control that achieves stable walking without the need for additional sensors on the human. The key idea is to extract the nominal gait and the human interaction information from motion capture data, and reconstruct the walking behavior with a dynamic model of the human-prosthesis system. The walking behavior of a human wearing a powered prosthesis is obtained through motion capture, which yields the limb and joint trajectories. Then a nominal trajectory is obtained by solving a gait optimization problem designed to reconstruct the walking behavior observed by motion capture. Moreover, the interaction force profiles between the human and the prosthesis are recovered by simulating the model following the recorded gaits, which are then used to construct a force tube that covers all the interaction force profiles. Finally, a robust Control Lyapunov Function (CLF) Quadratic Programming (QP) controller is designed to guarantee the convergence to the nominal trajectory under all possible interaction forces within the tube. Simulation results show this controller's improved tracking performance with a perturbed force profile compared to other control methods with less model information.

show abstract

Preliminary results on energy efficient 3D prosthetic walking with a powered compliant transfemoral prosthesis

Cited by 34 publications

References 27 publications

Separable Control Lyapunov Functions With Application to Prostheses

Separable Control Lyapunov Functions With Application to Prostheses

Recurrent Neural Network Control of a Hybrid Dynamical Transfemoral Prosthesis with EdgeDRNN Accelerator

Data-driven Characterization of Human Interaction for Model-based Control of Powered Prostheses

Contact Info

Product

Resources

About