towards gyroscopic balance assistance: proof of concept* · 2sensory-motor systems lab, eth zurich,...

1
Proceedings of the 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC14) August 26-30, 2014 Chicago, Illinois, USA. Towards Gyroscopic Balance Assistance: Proof of Concept* D. Lemus 1 , H. Vallery 1,2 Abstract— In this paper, the feasibility of a Control Moment Gyroscope (CMG) as balance assistance device is investigated. An inverted pendulum is used as experimental set-up. Our initial results show that it is possible to maintain balance within a 4 degrees range from vertical. I. INTRODUCTION Reducing falls is an urgent challenge in ageing societies, as falls are among the most frequent causes of hospitalization and death among the elderly [1]. A key factor leading to falls is degraded balance control. We proposed a minimalistic solution for balance assistance [2], which is based on a gyroscope assembly coupled to the upper body. By exploiting the effect of several control moment gyroscopes (CMGs), the required support is generated in the case of loss of balance. Here, we study technical feasibility of this approach using a single CMG mounted on an inverted pendulum (IP). II. METHODS The preliminary setup consists of a 5 kg CMG mounted on top of a 3 kg IP (Fig. 1). The CMG’s flywheel spins at 3500 rpm about the ˆ g s axis. In order to avoid singularities (i.e. ˆ g t ˆ e u ) the CMG is initially positioned in such a way that ˆ g s ˆ e u (i.e. γ =0 ). In addition, γ is restricted within a ±27.5 range. A spring-like behaviour (i.e. τ IP = with stiffness k, provided by the CMG about ˆ e v ), similar to the ankle strategy while maintaining balance in quiet stance was implemented. Tests were performed with the IP starting at an inclination angle φ =0 , and by manually applying small perturbations at the top. III. EXPERIMENTAL RESULTS AND DISCUSSION When the perturbations were applied, the CMG provided enough support to keep the IP upright (within a φ = ±2 range). Moreover, the CMG tracked the desired torque on the IP well, with an RMS error of 0.55Nm (Fig. 2). We presume that higher flywheel speeds (only a third of the flywheel angular velocity reported in [2] was used) will lead to higher moments, as needed for experiments with human subjects. REFERENCES [1] V. Scott, M. Pearce, and C. Pengelly,”Deaths due to falls among Canadians age 65 and over”, Public Health Agency of Canada, 2005. [2] D. Li, H. Vallery, ”Gyroscopic assistance for human balance”, 12th IEEE International Workshop on Advanced Motion Control (AMC),2012, pp. 1-6. *This research was supported by the U.S. Department of Education, National Institute on Disability and Rehabilitation Research, NIDRR-RERC, Grant Number H133E120010, and by the Marie-Curie career integration grant PCIG13-GA-2013-618899. 1 Faculty of Mechanical, Maritime and Materials Engineering, TU Delft, The Netherlands (e-mail: [email protected], [email protected]) 2 Sensory-Motor Systems Lab, ETH Zurich, Switzerland. Fig. 1. Experimental setup with a CMG balancing an inverted pendulum (IP). The pendulum’s inclination angle and the CMG’s rotation angle about the gimbal axis ˆ g g are denoted by φ and γ, respectively. Directions x,y and z are defined with respect to an inertial frame with z pointing vertical. Frames {ˆ eu, ˆ ev , ˆ ew} and {ˆ g s , ˆ g t , ˆ g g } are defined fixed to the IP and the CMG respectively, with ˆ eu and ˆ ew pointing in the anterior direction and along the IP’s longitudinal axis, respectively. Note that both frames fulfill the condition ˆ ew ˆ g g and ˆ ev y. Due to the gyroscopic effect, positive speeds about the CMG’s ˆ g g axis (within the operation range) will produce moments on the IP with a positive component τ IP about the ˆ ev axis (i.e. in the sagittal plane). 0 5 10 15 20 25 30 -20 0 20 -2 -1 0 1 2 3 -20 0 20 moment [Nm] moment [Nm] time [s] pole angle φ [ ] τ IP,ref τ IP,ref τ IP τ IP Fig. 2. Pendulum torque tracking over time with an RMS error of 0.55 Nm (top), and over angle illustrating the spring-like behavior (bottom). c 2014 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

Upload: others

Post on 31-Jul-2020

1 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Towards Gyroscopic Balance Assistance: Proof of Concept* · 2Sensory-Motor Systems Lab, ETH Zurich, Switzerland. Fig. 1. Experimental setup with a CMG balancing an inverted pendulum

Proceedings of the 36th Annual International Conferenceof the IEEE Engineering in Medicine and Biology Society (EMBC14)August 26-30, 2014Chicago, Illinois, USA.

Towards Gyroscopic Balance Assistance: Proof of Concept*

D. Lemus1, H. Vallery1,2

Abstract— In this paper, the feasibility of a Control MomentGyroscope (CMG) as balance assistance device is investigated.An inverted pendulum is used as experimental set-up. Ourinitial results show that it is possible to maintain balancewithina 4 degrees range from vertical.

I. INTRODUCTION

Reducing falls is an urgent challenge in ageing societies, asfalls are among the most frequent causes of hospitalizationand death among the elderly [1]. A key factor leading tofalls is degraded balance control. We proposed a minimalisticsolution for balance assistance [2], which is based on agyroscope assembly coupled to the upper body. By exploitingthe effect of several control moment gyroscopes (CMGs), therequired support is generated in the case of loss of balance.Here, we study technical feasibility of this approach usingasingle CMG mounted on an inverted pendulum (IP).

II. METHODS

The preliminary setup consists of a 5 kg CMG mountedon top of a 3 kg IP (Fig. 1). The CMG’s flywheel spins at3500 rpm about thegs axis. In order to avoid singularities(i.e. gt ‖ eu) the CMG is initially positioned in such a waythat gs ‖ eu (i.e. γ = 0

◦). In addition,γ is restricted withina ±27.5◦ range. A spring-like behaviour (i.e.τIP = −kφ

with stiffnessk, provided by the CMG aboutev), similar tothe ankle strategy while maintaining balance in quiet stancewas implemented. Tests were performed with the IP startingat an inclination angleφ = 0

◦, and by manually applyingsmall perturbations at the top.

III. EXPERIMENTAL RESULTS AND DISCUSSION

When the perturbations were applied, the CMG providedenough support to keep the IP upright (within aφ = ±2

range). Moreover, the CMG tracked the desired torque on theIP well, with an RMS error of 0.55 Nm (Fig. 2). We presumethat higher flywheel speeds (only a third of the flywheelangular velocity reported in [2] was used) will lead to highermoments, as needed for experiments with human subjects.

REFERENCES

[1] V. Scott, M. Pearce, and C. Pengelly,”Deaths due to fallsamongCanadians age 65 and over”, Public Health Agency of Canada, 2005.

[2] D. Li, H. Vallery, ”Gyroscopic assistance for human balance”,12th IEEE International Workshop on Advanced Motion Control(AMC),2012, pp. 1-6.

*This research was supported by the U.S. Department of Education,National Institute on Disability and Rehabilitation Research, NIDRR-RERC,Grant Number H133E120010, and by the Marie-Curie career integrationgrant PCIG13-GA-2013-618899.

1Faculty of Mechanical, Maritime and Materials Engineering, TU Delft,The Netherlands (e-mail: [email protected], [email protected])

2Sensory-Motor Systems Lab, ETH Zurich, Switzerland.

Fig. 1. Experimental setup with a CMG balancing an inverted pendulum(IP). The pendulum’s inclination angle and the CMG’s rotation angle aboutthe gimbal axisgg are denoted byφ and γ, respectively. Directionsx,yandz are defined with respect to an inertial frame withz pointing vertical.Frames{eu, ev, ew} and{gs, gt, gg} are defined fixed to the IP and theCMG respectively, witheu and ew pointing in the anterior direction andalong the IP’s longitudinal axis, respectively. Note that both frames fulfillthe conditionew ‖ gg and ev ‖ y. Due to the gyroscopic effect, positivespeeds about the CMG’sgg axis (within the operation range) will producemoments on the IP with a positive componentτIP about theev axis (i.e.in the sagittal plane).

0 5 10 15 20 25 30−20

0

20

−2 −1 0 1 2 3−20

0

20

mom

ent[

Nm

]m

omen

t[N

m]

time [s]

pole angle φ [◦]

τIP,ref

τIP,ref

τIP

τIP

Fig. 2. Pendulum torque tracking over time with an RMS error of 0.55 Nm(top), and over angle illustrating the spring-like behavior (bottom).

c©2014 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current orfuture media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for

resale or redistribution to servers or lists, or reuse of anycopyrighted component of this work in other works.