Development of a Novel Omnidirectional Treadmill-Based Locomotion Interface Device with Running Capability
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
applied sciences Article Development of a Novel Omnidirectional Treadmill-Based Locomotion Interface Device with Running Capability Sanghun Pyo , Hosu Lee and Jungwon Yoon * Integrated Institute of Technology, Gwangju Institute of Science and Technology (GIST), 123 Cheomdan-gwagiro, Buk-gu, Gwangju 61005, Korea; pyopyo83@gist.ac.kr (S.P.); lakelee77@gist.ac.kr (H.L.) * Correspondence: jyoon@gist.ac.kr Abstract: To achieve an immersive virtual reality (VR) environment, omnidirectional treadmills (ODTs) allow users to perform locomotion in any direction. However, existing ODTs are heavy and complex, and operate at low speeds. This limits fast user motion and prevents natural interactions in real applications such as military training programs and interactive games. In this paper, we introduce a novel locomotion interface device with running capability, which uses an omnidirectional treadmill with a new power transmission mechanism and a locomotion controller that enables the user to make fast movements. As a result of the improved power transmission performance due to the simple and relatively lightweight structure, the proposed two-dimensional treadmill can generate a maximum speed of 3 m/s, with an acceleration of 3 m/s2 . Moreover, through a pilot test with the proposed locomotion interface device, we verified that the fast directional changes during walking and running with the designed speed adaptation controller do not exceed the acceleration performance of the proposed system. Due to its wide range of movement speeds and acceleration capabilities, and lack of any motion constraints, the proposed locomotion interface device with a novel ODT can be used as a representative platform in various VR environments to enhance the Citation: Pyo, S.; Lee, H.; Yoon, J. immersive experience. Development of a Novel Omnidirectional Treadmill-Based Keywords: human-machine interfaces; modeling and design of mechatronics systems; virtual reality Locomotion Interface Device with and human interface Running Capability. Appl. Sci. 2021, 11, 4223. https://doi.org/10.3390/ app11094223 1. Introduction Academic Editor: Yuichi Kurita A locomotion interface (LI) can support walking and running through appropriately Received: 23 March 2021 generated ground surfaces to provide active participation in virtual environments (VEs) Accepted: 3 May 2021 with realistic spatial sensations [1]. Therefore, an LI provides a sense of mobility based on Published: 6 May 2021 energy input/consumption of actual walking [2]. To create more immersive locomotion, a LI system should allow the user to arbitrarily change not only the walking speed, but also Publisher’s Note: MDPI stays neutral the walking direction. Such features may motivate a user to participate more actively in with regard to jurisdictional claims in VR experiences, such as military training programs, physical education programs, disaster published maps and institutional affil- preparedness training, and rehabilitation programs [3]. iations. To simulate omnidirectional locomotion, several types of devices have been suggested, such as balls [4], large spheres [5], mobile robots [6], a rotating one-dimensional (1D) tread- mill [7], and a robotic foot platform [8–10]. However, these devices have limitations with respect to omnidirectional walking. As the most natural mechanism for two-dimensional Copyright: © 2021 by the authors. (2D) LI, an omnidirectional treadmill (ODT) can simulate the over-ground walking of Licensee MDPI, Basel, Switzerland. humans [11–13]. An ODT or 2D treadmill usually consists of “unit segments” (transverse This article is an open access article treadmills) installed in such a way as to form a continuous loop (Figure 1). Each unit distributed under the terms and segment is a narrow treadmill with its own belt. conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). Appl. Sci. 2021, 11, 4223. https://doi.org/10.3390/app11094223 https://www.mdpi.com/journal/applsci
Appl. Sci. 2021, 11, Sci. 2021, x FOR PEER11, 4223 REVIEW 2 of 20 2 of 20 Appl. Sci. 2021, 11, x FOR PEER REVIEW 3 of 20 the following capabilities: (1) X- and Y-axis translational motion generation through a novel gear-based power transmission mechanism; (2) a distributed power scheme in which two motors installed on each axis are synchronized using low-level synchronized motor control [17]; and (3) high-level feedback using robust integral of the sign of the error (RISE) control [18] to allow fast movements while changing the walking direction. The remainder of the paper is organized as follows; the main design concepts for the locomotion interface device with high speed and acceleration are presented in Section 2, the high-level controller design for the proposed locomotion interface and pilot study re- sults are in Sections 3 and 4. Finally, the conclusions drawn are given in Section 5. 2. LI Device Design for Fast Motion 2.1. Actuation of Unit Segment Belt by Geared Transmission For 2D treadmills, the main issue is how to actuate the segment belt for Y-axis motion. In Torus and Cyberwalk treadmills, as shown in Figure 2a, each segment requires a dedi- Figure 1. Omnidirectional Figure 1. Omnidirectional treadmill (ODT)treadmill cated actuator, concept which(ODT) (see concept Figure increases (see 2the for Figure the 2 mass for thedue cross-sectional segment cross-sectional view), theview), gearedand andinclusion to the omni-pulley of the actuating the geared omni-pulley (GOP)-based actuation scheme of and parts (GOP)-based the proposed actuation ODTitto thus makes scheme a generate of the proposed infinite 2Ddesign disadvantageous ODT to generate infinite ground.for implementing fast X-axis motion. 2D ground. In an ODT, as shown in Figure 1, the X-axis translational motion is generated by the rotation of all the unit segments along AX vector, while the Y-axis translational motion is generated by rotational actuation of the belt of each segment along AY vector. Thus, the 2D treadmill combines small treadmills assembled orthogonally to create a single, large treadmill. An ODT provides an infinite ground plane by generating independent belt mo- tions along two orthogonal axes (X and Y). In the cases of Cyberwalk [12] and Torus tread- mill [14], Y-axis translational motion is generated by individual actuators attached directly to each unit segment. This increases the inertia of the segments; thus, this design requires a very large amount of power to rotate all the segments (X-axis motion). As the weight of each segment is increased, the acceleration performance is greatly reduced. The ODTs developed by the United States Army Research Institute (US ARL) ODT [15] and the InfinaDeck treadmill [16] use a different mechanism to generate Y-axis mo- tion. The transverse treadmill belt (segment belt) is actuated by stationary actuator(s) through a special transmission system. This design, due to the light weight of the individ- ual segments, may improve the maximum speed and acceleration of the systems during X-axis translational motion as compared to those of the Cyberwalk and Torus treadmill. To drive each segment belt while allowing free movement in the X-axis direction, these ODTs use a frictional Figure transmission mechanism composed of omni-wheels. However, since Transver- Figure2. 2.Cross-sectional Cross-sectionalview view ofof the the Y-axis Y-axis motion of a 2D treadmill (unit segment): (a) Transversal the omni-wheels-based sal treadmill frictional actuated transmission directly by actuatormechanism an actuator (Torus has a low power or Cyberwalk transmission treadmill), (b) Frictional trans- treadmill actuated directly by an (Torus or Cyberwalk treadmill), (b) Frictional transmission efficiency, the maximum mission speed mechanism and based acceleration on the Omni-wheel of these ODTs for via the frame-fixed Y-axis motor translational (US army(c) ODT), (c) The mechanism based on the Omni-wheel via the frame-fixed motor (US army ODT), The proposed motion are limited. proposed Thus, due tobased mechanism the high inertia of the on gear-driven unit segment transmission via a treadmills frame-fixed for X-axis motor. mechanism based on gear-driven transmission via a frame-fixed motor. translational motion and the low efficiency of power transmission for Y-axis translational motion, the currently In In the US ARL anavailable ODT, as 2D ODT shown (United treadmills in FigureareStates 1,only Army the able X-axis Research Laboratory to translational accommodate is Omnidirectional slow human motion generated by the Treadmill), as shown in Figure 2b, a motor→ attached to theinsystem framean can drive the walking speeds and accelerations. This limitation is the main obstacle rotation of all the unit segments along AX vector, while the Y-axis translational motion isdeveloping immersive VR segment belt through environment an and with fast omni-wheel-based natural locomotion. power transmission mechanism. → The power A further generated fromcomplexity by inrotational the frame-fixedthe design actuation motor can of beoftransmitted a past the belt paced ODTof each to is thesegment that segment even afteralong belt AYY-axis for improvingvector.motion Thus, the by 2D the treadmill omni-wheels, combines while small the treadmills rollers of the assembled omni-wheels the power transmission efficiency for Y-axis motion and reducing the mass of moving orthogonally that are in to contactcreate witha single, the large segment componentsbelt treadmill. for passively Anrotate X-axis motion, ODTthe provides when an scheme infinite the segments actuation ground move along of the plane ODT X-axis.by This must generating independent transmission be carefully de- is basedbelt on motions line–contact along two friction, orthogonal which axes limits frame;(X and the ability Y). In the of the segmentcases of Cyberwalk belt should to follow [12]fast the andmove- Torus signed to reduce structural stress in the ODT’s additionally, the design min- treadmillof a [14], Y-axis translational motion theistransmission generated byefficiency individual of actuators attached imize energymentsloss from user along Y-axis. the actuators to theTopower enhance transfer components during motor segment ac- belt actu- directly to each unit segment. This increases the inertia of the segments; thus, this design tuation. ation, we propose the novel gear transmission method shown in Figure 2c to directly drive requires a very large amount of power to rotate all the segments (X-axis motion). As the each segment In this study, we have belt [19]. Since developed the locomotion a novel segment belts are driven interface device, bywhich geared-pulleys, is based the trans- weight of each segment is increased, the acceleration performance is greatly reduced. mission efficiency on an omnidirectional treadmill is with significantly enhanced a transmission as compared mechanism that to the frictional is referred to as transmission the by The ODTs developed by the United States Army Research Institute (US ARL) ODT [15] an omni-wheel. geared omni-pulley (GOP). To Moreover, overcome thetheproposed limitations concept is suitable ODTs of conventional to generate and tofast X-axis motion deal and the InfinaDeck treadmill [16] use a different mechanism to generate Y-axis motion. due to a lightweight with the complexities of fast ODT platform design, with low segment the proposed mass/inertia. locomotion interface device has The transverse treadmill belt (segment belt) is actuated by stationary actuator(s) through
Appl. Sci. 2021, 11, 4223 3 of 20 a special transmission system. This design, due to the light weight of the individual segments, may improve the maximum speed and acceleration of the systems during X-axis translational motion as compared to those of the Cyberwalk and Torus treadmill. To drive each segment belt while allowing free movement in the X-axis direction, these ODTs use a frictional transmission mechanism composed of omni-wheels. However, since the omni- wheels-based frictional transmission mechanism has a low power transmission efficiency, the maximum speed and acceleration of these ODTs for Y-axis translational motion are limited. Thus, due to the high inertia of the unit segment treadmills for X-axis translational motion and the low efficiency of power transmission for Y-axis translational motion, the currently available 2D treadmills are only able to accommodate slow human walking speeds and accelerations. This limitation is the main obstacle in developing an immersive VR environment with fast and natural locomotion. A further complexity in the design of a past paced ODT is that even after improving the power transmission efficiency for Y-axis motion and reducing the mass of moving components for X-axis motion, the actuation scheme of the ODT must be carefully designed to reduce structural stress in the ODT’s frame; additionally, the design should minimize energy loss from the actuators to the power transfer components during motor actuation. In this study, we have developed a novel locomotion interface device, which is based on an omnidirectional treadmill with a transmission mechanism that is referred to as the geared omni-pulley (GOP). To overcome the limitations of conventional ODTs and to deal with the complexities of fast ODT design, the proposed locomotion interface device has the following capabilities: (1) X- and Y-axis translational motion generation through a novel gear-based power transmission mechanism; (2) a distributed power scheme in which two motors installed on each axis are synchronized using low-level synchronized motor control [17]; and (3) high-level feedback using robust integral of the sign of the error (RISE) control [18] to allow fast movements while changing the walking direction. The remainder of the paper is organized as follows; the main design concepts for the locomotion interface device with high speed and acceleration are presented in Section 2, the high-level controller design for the proposed locomotion interface and pilot study results are in Sections 3 and 4. Finally, the conclusions drawn are given in Section 5. 2. LI Device Design for Fast Motion 2.1. Actuation of Unit Segment Belt by Geared Transmission For 2D treadmills, the main issue is how to actuate the segment belt for Y-axis motion. In Torus and Cyberwalk treadmills, as shown in Figure 2a, each segment requires a ded- icated actuator, which increases the segment mass due to the inclusion of the actuating parts and thus makes it a disadvantageous design for implementing fast X-axis motion. In the US ARL ODT (United States Army Research Laboratory Omnidirectional Treadmill), as shown in Figure 2b, a motor attached to the system frame can drive the segment belt through an omni-wheel-based power transmission mechanism. The power from the frame-fixed motor can be transmitted to the segment belt for Y-axis motion by the omni-wheels, while the rollers of the omni-wheels that are in contact with the segment belt passively rotate when the segments move along X-axis. This transmission is based on line–contact friction, which limits the ability of the segment belt to follow the fast movements of a user along Y-axis. To enhance the transmission efficiency of segment belt actuation, we propose the novel gear transmission method shown in Figure 2c to directly drive each segment belt [19]. Since the segment belts are driven by geared-pulleys, the transmission efficiency is significantly enhanced as compared to the frictional transmission by an omni-wheel. Moreover, the proposed concept is suitable to generate fast X-axis motion due to a lightweight platform with low segment mass/inertia. The omnidirectional rack described in [20] may be considered a suitable mechanism for orthogonal translation motion. However, it is not appropriate for creating infinite ground because the rack gear is made using a rigid material, which is not flexible such as timing belt. The proposed transmission mechanism for Y-axis translation can use commercially
Appl. Sci. 2021, 11, x FOR PEER REVIEW 4 of 20 The omnidirectional rack described in [20] may be considered a suitable mechanism Appl. Sci. 2021, 11, 4223 for orthogonal translation motion. However, it is not appropriate for creating infinite 4 of 20 ground because the rack gear is made using a rigid material, which is not flexible such as timing belt. The proposed transmission mechanism for Y-axis translation can use com- mercially available timing belts as segment belts (Standardized name: T10 Urethane belt, available timing belts as segment belts (Standardized name: T10 Urethane belt, pitch pitch specification: 10 mm). specification: 10 mm). 2.2. 2.2.Transmission TransmissionDesign DesignforforOmnidirectional OmnidirectionalMotion Motion Figures Figures1 1and and2c2cshow showthetheconceptual conceptualdesign designofofthe theproposed proposed2D 2Dtranslational translationalmotion motion with geared-pulley transmission. This simple holonomic design with geared-pulley transmission. This simple holonomic design allows generationallows generation of infi-of nite motion in both axes. In addition, this actuation method only drives infinite motion in both axes. In addition, this actuation method only drives the segmentthe segment belts that beltsare onare that theonactive surface the active where surface the user where can can the user locate andand locate walk. This walk. Thiscancanreduce reducethe the required required motor power as less than half of the segment belts are actuated at any giventime. motor power as less than half of the segment belts are actuated at any given time. However, However,when whenthe theproposed proposedmechanism mechanismcreates creates2-dimensional 2-dimensionalground groundby byactuating actuatingbothboth XXandandYYaxes, axes,the thecoupled coupledsurfaces surfacesofofthethegeared-pulleys geared-pulleysand andsegment segmentbelts beltsget getfrictional frictional forces forcesbecause becausethe segment the segment belt should belt should alsoalso move movein aindirection perpendicular a direction perpendicular to the todi-the rection of the transmitted force of the input geared pulley. Thus, the geared direction of the transmitted force of the input geared pulley. Thus, the geared pulley needs pulley needs totobebemodified modifiedtotoinclude includea apassive passiverotation rotationmechanism mechanismfor forreducing reducingthis thisfriction. friction.InInthis this paper, paper, we implement the Y-axis motion using a geared omni-pulley set (GOPS)equipped we implement the Y-axis motion using a geared omni-pulley set (GOPS) equipped with withtoothed toothedrollers rollersinstead insteadofofthe theconventional conventionalgeared gearedpulley, pulley,as asshown shownin inFigure Figure3a.3a. (a) (b) Figure Figure3.3.(a)(a)GOPS GOPSdesign designincluding includingtoothed toothedrollers, rollers,(b) (b)The Thefront frontprojection projectionofofGOPS GOPSdesign designwith with a a normal normal geared-pulley geared-pulley parameter, parameter, and and GOPS GOPS configuration configurationwith withits itsgeometric geometricanalysis. analysis. The Thedesign designconcept conceptofofGOPS GOPSisisorganized organizedbybyan anomni-wheel omni-wheelwith withgear. gear.The Thehigh high → transmission transmissionefficiency efficiencyofofthe thegear gearisisused usedfor formotion motionin the AA inthe Y direction, while the friction Y direction, while the friction on the tooth surface caused by motion in the AX direction→ is reduced by the passive rota- on the tooth surface caused by motion in the AX direction is reduced by the passive tion of the toothed rollers, similar to an omni-wheel. The toothed roller for reducing the rotation of the toothed rollers, similar to an omni-wheel. The toothed roller for reducing frictional force is designed by body of rotation of an involute toothed part of a normal the frictional force is designed by body of rotation of an involute toothed part of a normal geared pulley along the rotational vector → shown in Figure 3b. This rotational vector geared pulley along the rotational vector N shown in Figure 3b. This rotational vector also represents the axis of the passive rotation of the toothed rollers when segments move along the X-axis. In the presented system, the design of the GOPS uses the specifications of the com- mercial product called the T10 type geared pulley that works with the T10 Urethane belt.
Appl. Sci. 2021, 11, x FOR PEER REVIEW 5 of 20 also represents the axis of the passive rotation of the toothed rollers when segments move Appl. Sci. 2021, 11, 4223 5 of 20 along the X-axis. In the presented system, the design of the GOPS uses the specifications of the com- mercial product called the T10 type geared pulley that works with the T10 Urethane belt. The The parameters parameters of of the thenormal normalgeared gearedpulley, pulley, which which areare independent independent of number of the the number of of teeth, teeth, pitch (p = 10 mm), tooth height (t = 3.2 mm), tooth width (w pitch (p = 10 mm), tooth height (th = 3.2 mm), tooth width (wh = 2.76 mm) and input angle h h = 2.76 mm) and input angle of theoftooth the tooth (λ = 25 (λdeg). = 25 deg). Moreover, Moreover, GOPSGOPS designdesign also considers also considers the parameters the parameters de- depended pended on the number on the number of teethof(Z teeth = 36)(Zsuch = 36)as such as radius radius of theof the pitch pitch circle circle (rp = 57.295 (rp = 57.295 mm),mm), total total radius radius (r =(r56.375 = 56.375 mm)mm) of of thethe GOPS GOPS andand theradius the radiusofofthethebase basecircle circle (r (rbb == 53.175 53.175mm)mm) because the frontal projection of the GOPS is identical to a normal geared-pulley because the frontal projection of the GOPS is identical to a normal geared-pulley profile. profile. Among Amongthe theGOPS GOPSconfiguration configurationparameters parametersshown shownininFigure Figure3b, 3b,the thenumber, number,n, n,ofof toothed toothedrollers rollersisisthe themain maindeterminant. determinant.Figure Figure44shows showsthe theexample exampleofofhow howto toproperly properly setup setupthetheGOPS GOPSconfiguration configuration in in thethe case thatthat case the the number number of toothed rollers of toothed is 6. The rollers is 6.rela- The tionship between relationship the total between number the total of teeth number (Z) and of teeth (Z) the andnumber the numberof toothed rollers of toothed (n) de- rollers (n) defines fines to the to the module module of the of the GOPSGOPS(mGOP(m)GOP ) as follows: as follows: mGOP Z / n(n (=n =2a2a≥≥6,6,aa≥≥ 33)),, ((a,a,Z, mGOP==Z/n Z ,n, n, m GOP ∈ mGOP ),, ∈ N) (1) (1) where wherea∈ℕa∈N isis aa positive positive natural natural number number greater than or greater than or equal equal toto 3. 3. Thus, Thus, mmGOP GOP represents represents the the number of teeth in one toothed roller. It can also define an appropriatenumber number of teeth in one toothed roller. It can also define an appropriate numberfor fornn because becausethe thenumber numberofofteeth teethmust mustbe beaanatural naturalnumber. number.All Alltoothed toothedrollers rollershave haveananindex index number number(i(ith) according th ) according to the to range the range 1 ≤1 i∈≤ℕ ≤ i n, ∈N as ≤ shown n, as in shownFigure in 4. Odd Figure 4. and Oddeven andnum- even bers in thein numbers toothed roller roller the toothed indexindex (i) make up each (i) make up separate geared each separate omni-pulley geared (GOP)(GOP) omni-pulley (see also (seeFigure 3), and also Figure 3),two andGOPs are combined two GOPs are combined to form one GOPS. to form one GOPS. . GOPSdesign Figure4.4.GOPS Figure designparameters parametersand andrange rangeofoftoothed toothedroller rollersize. size. InInthe thedesign designofofGOPS, GOPS,the thetotal totalnumber numberofofteeth teeth(Z)(Z)ofofthe theGOPS GOPSisis 36,36, andanditsits 66 toothed rollers are configured with a uniform angular spacing of 60 ◦ . To configure the toothed rollers are configured with a uniform angular spacing of 60°. To configure the frontalprojection frontal projectionofofthis thisGOPS GOPSaccording according toto the the normal normal geared-pulley geared-pulley profile, profile, oneone GOP GOP is is placed behind the other with a phase offset of 60 ◦ between the toothed rollers of both placed behind the other with a phase offset of 60° between the toothed rollers of both the the GOPs. GOPs. Oncethe Once theGOP GOParrangement arrangementisisset, set, the the next next stepstep is is selecting selecting the the size of rrroller size of asthe rolleras the GOPS design parameter, whose range is calculated by the general geared-pulley parameters GOPS design parameter, whose range is calculated by the general geared-pulley parame- (p, th , etc.) and the GOPS configuration parameter (n) that has already been determined. ters (p, th, etc.) and the GOPS configuration parameter (n) that has already been deter- Due to the shape of the toothed roller, the radius at its end (r ) is used to define its size. mined. Due to the shape of the toothed roller, the radius at itsroller end (rroller) is used to define Moreover, rroller_thick is defined as the largest radius at the center of the toothed roller, as its size. Moreover, rroller_thick is defined as the largest radius at the center of the toothed roller, shown in the right-side image of Figure 4. r has the following range: as shown in the right-side image of Figure 4.roller rroller has the following range: th r < rroller < . (2) sin π 1.5n−1 + p/4r p 4 sin(2π/n) n Thus, the minimum rroller is calculated to maintain the proper tooth shape, and its maximum size corresponds to the circumcenter point to avoid interference between the
th r < rroller < . Appl. Sci. 2021, 11, 4223 ( sin π 1.5nn−1 + p / 4rp ) 4sin ( 2π / n) (2) 6 of 20 Thus, the minimum rroller is calculated to maintain the proper tooth shape, and its maximum size corresponds to the circumcenter point to avoid interference between the arrangements of each toothed roller installed in a GOP. The selected rroller defines the value arrangements of each toothed roller installed in a GOP. The selected rroller defines the value of rroller_thick according to the following relationship: of rroller_thick according to the following relationship: ( ( rroller _ thick==rrroller ++r(r−−r sin rroller_thick roller π 1.5n −1 / n . )) |r sin((π (1.5n )− 1)/n)|). (3)(3) In In thethe presented presented design, the design, value the ofof value rroller is 10.82 rroller mm, is 10.82 mm,which whichis selected considering is selected considering thethe range of rroller calculated using Equation (2) (3.6~16.27 mm), and do is selected as 28.5 range of rroller calculated using Equation (2) (3.6~16.27 mm), and do is selected as mm to mm 28.5 avoidtointerference between avoid interference the toothed between rollers rollers the toothed of the GOPs. of the GOPs. 2.3.2.3. Realization of Stable Realization Omnidirectional of Stable Motion Omnidirectional Motion Figure Figure5a5ashows shows howhow thethe ODT ODT can generate can generate X-axis X-axismotion motion through through segment segment rotation rotation → along alongA Adefined in Figure 1. The segments are pin-constrained to the rib of the X-axis X defined in Figure 1. The segments are pin-constrained to the rib of the X- drive chain. As compared axis drive chain. As compared to a timing to belt mechanism a timing [19], the chain belt mechanism [19],mechanism improves the chain mechanism theimproves transmission efficiency of X-axis drive by allowing the precise positioning the transmission efficiency of X-axis drive by allowing the precise positioning of each seg- ment without collisions between segments at high speeds. In the unit of each segment without collisions between segments at high speeds. In the unit segment segment (transverse treadmill) design (transverse with a timing treadmill) design belt withdriven a timingusing beltGOPS, driventhe segment using GOPS, belt thehas been turned segment belt has inside beenout so that turned the toothed inside side the out so that of the belt is side toothed coupled of thewith the belt is teeth of the coupled GOPS, with while of the teeth a belt tensioner the GOPS, retains while a beltthe segmentretains tensioner belt tension. A frame-fixed the segment motor(s) belt tension. drives the A frame-fixed seg- motor(s) ment timing drives belt continuously the segment timing belt through the GOPS continuously to achieve through Y-axistomotion, the GOPS achievewhile Y-axisallow- motion, ingwhile translational allowingmotion of segments translational motion along the X-axisalong of segments throughthe the passive X-axis rotation through the of the passive toothed rollers [19,21]. It should be noted that the number of contact teeth rotation of the toothed rollers [19,21]. It should be noted that the number of contact teeth between one segment between belt oneand one GOPS segment belt is setone and to be the same GOPS is set as to mbeGOP . same as mGOP . the (a) Figure 5. Cont.
Appl. Sci. 2021, 11, 4223 7 of 20 Appl. Sci. 2021, 11, x FOR PEER REVIEW 7 of 20 Appl. Sci. 2021, 11, x FOR PEER REVIEW 7 of 20 (b) (b) Figure 5. Figure 5. (a) Drive chain of segment segment treadmills treadmills for for X-axis X-axismotion, motion,(b) (b)Design Designand andoperation operationofofthe the Figure 5. (a) Drive synchronizer chain of segment treadmills for X-axis motion, (b) Design and operation of the mechanism. synchronizer mechanism. synchronizer mechanism. When Whenaasegment segmentbelt beltenters entersthetheactive activesurface, surface, asasshown shown in in Figure Figure5b 5b (See alsoalso (See Figure 1), Figure When aspeed a1),rotational segment belt enters mismatch occurs thebetween active surface, the the as shown segment beltin Figure 5b entering the(See alsosurface active Figure a rotational speed mismatch occurs between segment belt entering the active sur- 1), a rotational without GOPS speed mismatch coupling and the occurs between the segment belt entering the Therefore, active sur- face without GOPS coupling andsegment the segment beltsbelts already in the already inactive surface. the active surface. There- face speed without GOPS synchronization coupling should and the segment be considered belts to prevent already damage in the active to the to surface. teeth, and toandThere- guar- fore, speed synchronization should be considered to prevent damage the teeth, to fore, antee speed smooth synchronization gear coupling should during be considered the re-coupling to prevent stage.stage. damage to the teeth, and to guarantee smooth gear coupling during the re-coupling guarantee For smooth gear coupling during the re-coupling stage. For the the proposed proposed F-ODT,F-ODT, aa speed speedsynchronizer synchronizer is is used used totoaccelerate accelerate thethere-coupling re-coupling segment For the proposed F-ODT, a speed synchronizer isAs used to accelerate the re-coupling segment beltbelt in in advance advance through through frictional frictional actuation. actuation. As aa unit unit segment segment approaches approaches the the segment edge of belt the in advance GOPS, a cone through shaped frictional actuation. synchronizer increases As its a unit belt segment speed to approaches match that the of edge of the GOPS, a cone shaped synchronizer increases its belt speed to match that of the edge the of belts the GOPS, already a cone coupled shaped with synchronizer the GOPS. increases Thus, the its belt synchronizer speed to match minimizes that the of the speed belts already coupled with the GOPS. Thus, the synchronizer minimizes the speed mis- belts already mismatch betweencoupled with the GOPS. Thus, therecoupled. synchronizer minimizes the speed mis- match between the the GOPS GOPSandandthethe beltbelt beingbeing recoupled. match between the GOPS and the belt being recoupled. 2.4. 2.4. Design Design of of Actuation Actuation System System forfor Desired DesiredPerformance Performance 2.4. Design When a human try to run orDesired of Actuation System for Performance stop quickly during straight walking, a maximum When a human try 2 to run or stop quickly during straight walking, a maximum accel- acceleration When aof 32 m/stryistogenerated human run or stop[22]. Thus, quickly the target during straightperformance of velocityaccel- walking, a maximum and eration of 3 m/s is generated [22]. Thus, the target performance 2 , respectively, of velocity and accelera- acceleration eration of 3 m/s of the 2 isODT were set generated [22].toThus, 3 m/sthe and 3 targetm/s performance of to simulate velocity and running accelera- tion of the ODT were set to 3 m/s and 3 m/s2, respectively, to simulate running and stop- and tionstopping. of the ODT To were validate the3target set to m/s and performance, dynamic analysis 3 m/s2, respectively, to simulate wasrunning performed andusing stop- ping. To validate the target performance, dynamic analysis was performed using a com- aping. commercial multibody dynamics software (ADAMS), as shown To validate the target performance, dynamic analysis was performed using a com- in Figure 6. To obtain mercial multibody dynamics software (ADAMS), as shown in Figure 6. To obtain realistic realistic analysis results, mercial multibody dynamicsthe boundary conditions,asincluding software (ADAMS), shown inthe mass Figure 6. and inertia, To obtain were realistic analysis results, the boundary conditions, including the mass and inertia, were set based on set basedresults, analysis on 3D modeling, the boundary gravity, the initial conditions, X-axisthe including drivemasschain andtension inertia, (3000 were setN) based and the on 3D modeling, gravity, the initial X-axis drive chain tension (3000 N) and the Coulomb fric- Coulomb 3D modeling, friction due to gravity, contact the initial between X-axis drivethe segment chain tension belt and (3000 theN)Teflon-coated and the Coulomb segmentfric- tion due to contact between the segment belt and the Teflon-coated segment structure [19]. structure tion due to [19]. contact between the segment belt and the Teflon-coated segment structure [19]. Figure 6. Dynamics simulation for X and Y axes. Figure6.6.Dynamics Figure Dynamicssimulation simulationfor forXXand andYYaxes. axes. The X-axis motion simulation was done by actuating all 64 unit segments (total TheX-axis The X-axis motion simulation waswas done by actuating all 64 unit segments (total weight: 576 kg)motion simulation and human mass (150 done by actuating kg) through all 64 unit the rotation segments of the (totalsprocket, chain and weight: weight: 576 576 kg) and mass human mass (150 kg) through the rotation of the chain and sprocket, and Y-axis motion was performed with one segment among the 27 segment belts onand kg) and human (150 kg) through the rotation of the chain and sprocket, the and Y-axis Y-axis motionmotion was performed was performed with with one one segment segment amongamong the 27 the 27 segment segment belts onbelts on the the active active surface loaded with 150 kg to simulate a human mass. The additional information surface loadedloaded active surface with 150 kg150 with to simulate a human kg to simulate mass.mass. a human The The additional information additional informationof of the proposed ODT is summarized in Table 1, which is used to determine the actuator of the proposed ODT is summarized in Table 1, which is used to determine the actuator
Appl. Sci. 2021, 11, 4223 8 of 20 the proposed ODT is summarized in Table 1, which is used to determine the actuator power. The X-axis and Y-axis motor instantaneous powers required to maintain a speed of 3 m/s were determined to be 28 kW and 8 kW, with average values of 8.8 kW and 4.2 kW, respectively. Table 2 summarizes the power requirements for each axis actuation. Table 1. The ODT Specifications for Dynamic Simulation. Item Specifications System frame dimensions 2780 mm × 3310 mm × 640 mm Active surface area 2.5 m × 2.5 m Unit segment dimensions 100 mm × 2577 mm × 70.5 mm Unit segment weight 9 kg Number of segments 64 units Number of active segments 27 units Number of GOPS in 1 GOP shaft 54 units per 1 GOP shaft Chain and timing belt X-axis drive chain Y-axis segment belt Pitch 18.875 mm 10 mm Width 9.4 mm 96 mm Actuation part specification Sprocket GOP shaft Pitch diameter 396.375 mm 114.59 mm The number of teeth 21 36 Table 2. Power Requirements. Required Pulley Axis Required Pulley Torque Power Angular Velocity 1768 Nm (max.) 28 kW (peak) X 15.63 rad/s 563 Nm (avg.) 8.8 kW (nominal) 148.5 Nm (max.) 8 kW (peak) Y 52.35 rad/s 81 Nm (avg.) 4.2 kW (nominal) In the presented system, 3-rows of GOP shafts were installed to connect 6 GOPS per one segment, as shown in Figure 7a, to secure the performance of the geared-coupling between the segment and GOPS. Thus, this mechanism reduces the power concentration on the contacted teeth of a segment belt by increasing the number of the contacted teeth by 3 times. Therefore, it can guarantee the power transmission performance because the generated motor torque required to drive the segment belt is distributed over the 3-row GOP shaft. The motors (Motor1y , Motor2y ) actuate the power transmission belts, which in turn drive the gearboxes. The gearboxes actuate timing belts for rotation of the 3 GOP shafts simultaneously as a mechanically coupled power transmission system. The actuation mechanism for the X-axis also uses a distributed power design, in which the four drive chains are mechanically coupled. In Figure 7b, Motor1x and Motor2x simultaneously actuate the drive chains by actuating the sprockets.
Appl. Sci. 2021, 11, 4223 9 of 20 Appl. Sci. 2021, 11, x FOR PEER REVIEW 9 of 20 (a) (b) Figure7.7. Mechanically Figure Mechanicallycoupled coupledactuation actuationsystem systemdesign: design:power powertransmission transmissionmechanisms mechanismsofof(a)(a)the the Y-axis and (b) the X-axis. Y-axis and (b) the X-axis. As the As thedesigned designedODTODTrequires requires(1) (1)aapower powertransmission transmissionmechanism mechanismfor fordriving drivingthethe 3-row GOP 3-row GOP shafts shaftssimultaneously, simultaneously,(2) (2)aachain-sprocket chain-sprocketmechanism mechanismfor forconstraining constrainingand and carryingthe carrying thesegments, segments,and and(3) (3)aa frame frame stiffness stiffnesssuitable suitableforfor high high velocity velocity and and acceleration, acceleration, the theinterior interiorofofthe theODT ODT has very has limited very limitedspace. Therefore, space. motion Therefore, actuation motion alongalong actuation each axis each isaxis generated using using is generated two synchronized two synchronized motors, as shown motors, in Figure as shown in 7. Under Figure 7. this Underdistributed this dis- actuation, the motors’ tributed actuation, thepower motors’ can be distributed power uniformly can be distributed to all thetopower uniformly all thetransmission power trans- components mission components without excessive stress. Moreover, this actuation designcompara- without excessive stress. Moreover, this actuation design provides provides tively faster command comparatively response than faster command responsea single thanmotor a singledesign motor[23]. design [23]. Since Since the the system uses uses aa distributed distributedpower powerscheme, scheme,ititisisimportant important to to achieve achieve pre- precise cise speedspeed control control of both of both actuators actuators to prevent to prevent damagedamage to the to the power power transmission transmission com- components ponents [17]. To simultaneously [17]. To simultaneously minimizeminimize the differences the differences in speed andin speed torqueand torque of both of botha actuators, actuators, a cross-couple control scheme [24] was incorporated into the cross-couple control scheme [24] was incorporated into the low-level control of the pro- low-level control of the proposed ODT, as shown posed ODT, as shown in Figure 8. in Figure 8.
mission components without excessive stress. Moreover, this actuation design provides comparatively faster command response than a single motor design [23]. Since the system uses a distributed power scheme, it is important to achieve precise speed control of both actuators to prevent damage to the power transmission components Appl. Sci. 2021, 11, 4223 [17]. To simultaneously minimize the differences in speed and torque of both actuators, 10 of 20a cross-couple control scheme [24] was incorporated into the low-level control of the pro- posed ODT, as shown in Figure 8. Appl. Sci. 2021, 11, x FOR PEER REVIEW 10 of 20 Figure 8. Low-level speed controller for motor synchronization. Figure 8. Low-level speed controller for motor synchronization. Thelow-level The low-levelcontroller controller synchronizes synchronizes the the motor motor speeds speeds according according to the to the desired desired speed speed command i (vci), where vcx= [vycx,Tvcy]T is the desired speed of the active surface, and vi 1i command (vc ), where vc = [vc , vc ] is the desired speed of the active surface, and v1 andvv2i i correspond and correspond to to the thevelocity velocityof ofeach eachmotor, motor,determined determinedvia viaencoder encoderfeedback feedbackwith with 2 respecttotothe respect theXXororY-axis, Y-axis,Tc1 i andTTc2c2i are Tc1i and i are the the torque torque values. values. Synchronization Synchronizationfor forreducing reducing the difference of each motor’s speed (vvi) and torque (vTi i) in each axis was achieved by i the difference of each motor’s speed (vv ) and torque (vT ) in each axis was achieved by proportional-derivative(PD) proportional-derivative (PD)control, control,asasfollow: follow: d (v −− v2i v) i i vv = Cv v = C ( v , v ) = k ( v − v )+ i i i i i i i i 1 1i i id v +k 2, i = X , Y , i i vv 1 i , v2v i 1= k 2 Pv vPv1 i −1 v2 i2 , i = X, Y, kDv (4) (4) Dv dt dt wherekPv where kPvand andkkDv Dv are the positive gains of the speed speed controller. controller.The Thetorque torquesynchronization synchronization controllerisisalso controller alsoimplemented implementedusing usingPD PDcontrol, control,asasfollows: follows: ddTcT1i c1−iT− ( i c 2 Tc2 i ) ( ) ( ) ii iii i i ii i ii i v T = Ct i Tvc1T ,=TCc2t Tc1=,Tkc2Pt = kTPtc1 T− c1 T−c2Tc2 + + kkDtDt , i = X, ,Y i=, X, Y, (5) (5) dt dt wherekPt where kPti i and andkkDtDti are the positive positive gains gains of ofthe thetorque torquesynchronization synchronizationcontroller, controller,andandTc1 i i Tc1 and and Tc2 Tc2i are the torque values for the control input of C i i. To secure control the torque values for the control input of Ct . To secure control stability of t stability of the low-level the low-level controller, suitable controller, suitablegains were gains selected were based selected on the based Nyquist on the Nyquist criterion by con- criterion by sidering thethe considering backlash backlash model model[19,25]. [19,25].InInterms termsof of the criterion, the the Nyquist criterion, theintersection intersection showsaamarginally shows marginallystable stablecondition, condition, i.e., i.e., within within 0.40.4 Hz.Hz. Regarding Regarding thethe gaingain of the of the control- controller, i = 0.1i and 0.5, k i = 0.02 kPvas; i = 0.2i and 0.4, and the ler,Xthe andXYand values were set Y values wereas;set kPv = 0.1 andDv 0.5, kDvi =and and k0.04, 0.020.04, Pt kPt = 0.2 and 0.4, kandi = 0.08 and 0.1, respectively. Dt kDt = 0.08 and 0.1, respectively. i 2.5. 2.5.Fabricating FabricatingthetheODT ODTandandVerifying VerifyingthethePerformance Performance The proposed ODT fabricated as The proposed ODT fabricated as shown shown inin Figure 9 has Figure thethe 9 has following major following character- major charac- istics: (1) high transmission efficiency via the novel GOPS for driving the segment teristics: (1) high transmission efficiency via the novel GOPS for driving the segment beltsbelts for Y-axis motion, (2) low-weight unit segments for fast X-axis motion, (3) distributed for Y-axis motion, (2) low-weight unit segments for fast X-axis motion, (3) distributed mo- motor actuation to reduce tor actuation the motion to reduce response the motion timetime response and the andstrain on transmission the strain parts.parts. on transmission Figure9.9.The Figure Thedeveloped developedODT ODTusing usingGOP GOPactuation actuationon onthe theactive activesurface, surface,distributed distributedactuation, actuation, and and the dynamics simulations results. the dynamics simulations results. To verify the performance of the proposed ODT, we measured the maximum accel- eration and velocity of the active surface. The motion command was given as a sinusoidal signal of a magnitude of 3 m/s at 0.16 Hz for generating the maximum acceleration of 3 m/s2. In addition, to verify the low-level controller, the velocity and torque transmitted from the motors were also measured.
Appl. Sci. 2021, 11, 4223 11 of 20 To verify the performance of the proposed ODT, we measured the maximum accelera- tion and velocity of the active surface. The motion command was given as a sinusoidal signal of a magnitude of 3 m/s at 0.16 Hz for generating the maximum acceleration of 3 m/s2 . In addition, to verify the low-level controller, the velocity and torque transmitted from the motors were also measured. Figure 10 shows the results of the maximum speed and acceleration. The root mean square (RMS) value of the speed difference was 0.0087 m/s for the Y-axis, and 0.0013 m/s Appl. Sci. 2021, 11, x FOR PEER REVIEW 11 of 20 for the X-axis, respectively, as shown in Figure 10a,c. Thus, the speed synchronization and command-following performance are considered to be stable. In the Y-axis case, the RMS value of the torque bias was 27%, while in the X-axis case, the RMS value of torque was bias20%, was as shown 20%, FigureFigure as shown 10b,d. 10b,d. Thus, the low-level Thus, controller the low-level can overcome controller the nonlin- can overcome the earities presentpresent nonlinearities in the in power transfer the power process transfer and adequately process distribute and adequately the load distribute to the the load to the motors motors while while executing executing the motion the motion commands. commands. It also It can can also adequately adequately compensate compensate for for the the torque torque difference. difference. Figure10. Figure Verification of 10.Verification of the the performance performance using using aa sinusoidal sinusoidal motion motion command command for for surface surface acceler- accelera- tion of 3 m/s 2 . (a) Y-axis speed synchronous performance, (b) Y-axis torque synchronous performance, ation of 3 m/s . (a) Y-axis speed synchronous performance, (b) Y-axis torque synchronous perfor- 2 (c) X-axis mance, (c) speed X-axissynchronous performance, speed synchronous and (d)and performance, X-axis (d)torque X-axis synchronous performance. torque synchronous perfor- mance. In the maximum performance, both axes showed a torque requirement of about 1.75 In times the the rated motor maximum capacity, due performance, bothtoaxes the showed implicit friction a torqueofrequirement the real system. However, of about 1.75 the servo motors guarantee operation at 200% of the rated torque capacity times the rated motor capacity, due to the implicit friction of the real system. However, for a period of 5~10-min. the servo motors Therefore, guaranteetheoperation system can safelyofachieve at 200% the ratedhigh velocities torque andfor capacity accelerations a period of exceeding 3 m/s and 3 m/s 2 , respectively. 5~10-min. Therefore, the system can safely achieve high velocities and accelerations ex- ceedingTable 3 compares 3 m/s and 3 m/sthe specifications of the existing 2D treadmills and the proposed 2, respectively. ODT. The proposed ODT has a workspace Table 3 compares the specifications of 2.5 of the m × 2.5 existing 2Dm, which isand treadmills sufficient for safe the proposed running and various other types of locomotion, such as crawling. ODT. The proposed ODT has a workspace of 2.5 m × 2.5 m, which is sufficient for It can achieve the higher safe velocities and accelerations than the others. Also, due to the use of the distributed running and various other types of locomotion, such as crawling. It can achieve the higher actuation system, the velocities and overall system height accelerations than theis
Appl. Sci. 2021, 11, 4223 12 of 20 Table 3. Comparison with Existing 2D Treadmills. Y-axis Drive Active Surface Max. vel. Max. acc. Actuator Specification Mechanism Area/Thickness (km/h) (m/s2 ) Frame stationery motor 1.3 × 1.3 m2 X-axis 4 kW (1 EA) US army ODT 1 7.2 Under 1 with omni-wheel /0.46 m Y-axis 4 kW (1 EA) Segment attached 6.5 × 6.5 m2 X-axis 40 kW (4 EA) 7.2 0.5 Cyber Walk motor /1.5 m Y-axis 37.5 kW (25 EA) 10.8 0.75 Segment attached 1 × 1 m2 X-axis 200 W (1 EA) 4.3 1 Torus treadmill motor /0.5 m Y-axis 960 W (12 EA) 4.3 0.8 Frame stationery motor 2.5 × 2.5 m2 X-axis 8.8 kW (2 EA) 10.9 3 Proposed ODT Appl. Sci. 2021, 11, x FOR PEER GOPS 2 withREVIEW /0.64 m Y-axis 5.8 kW (2 EA) 10.9 3 12 of 20 ODT 1 : omnidirectional treadmill, GOPS 2 : geared omni-pulley set. 3. LI 3. LI Control Control for for Omnidirectional OmnidirectionalRunning Running 3.1. 3.1. Design Design of of High-Level High-Level Controller Controller For Foreffective effectivetreadmill-based treadmill-basedgait gaitexercise, exercise,aauser userspeed speedadaptation adaptationcontroller controllershould should converge converge toto an an intentional intentional speed speed of of aauser userfast fastand andprecisely precisely[26]. [26]. Moreover, Moreover, ititshould should effectively effectivelycompensate compensatethe theposition positionerror error from from the reference the referenceposition to prevent position to prevent feetfeet twisting twist- within turning ing within motions turning in 2Din motions treadmill simulations 2D treadmill [15]. This simulations [15].isThis because if a userifperforms is because a user per- a step turning while too away from the reference position, an excessively forms a step turning while too away from the reference position, an excessively unin- unintended motion of the active tended motionsurface of the inactive ODT occurs surfaceininlateral direction. ODT occurs in lateral direction. To verify that the fast directional changes To verify that the fast directional changes duringduring walking and running walking and runningis possible using is possible to the proposed using ODT byODT to the proposed conducting turn walking, by conducting the high-level turn walking, controller the high-level facilitating controller is facili- designed by using the tating is designed by 1D treadmill using the 1Dsystem treadmillmodel shown system in Figure model shown 11.inWhen Figurewalking 11. Whenor running, walking the position the or running, error is derived position errorasisfollows derived[27]: as follows [27]: . d( x d−( xd−)/dt xd ) /=dt−=v−cv+ vww, ,vcv= c +v c =a ac ,c , (6) (6) wherexxisisthe where thecurrent currentuser userposition positionwith withrespect respecttotothe thereference position,xxd disisthe referenceposition, thedesired desired position, v is the intentional velocity of the position, vww the intentional velocity of the user, and vc anduser, and v c and a c the treadmill beltbelt ac areare the treadmill ve- velocity locity and and acceleration acceleration commands, commands, respectively. respectively. In the In the modelmodel shownshown in Equation in Equation (6), v(6), c is vapplied c is applied to treadmill to the the treadmillservoservo motors, motors, where where its dynamics its dynamics are compensated are compensated by thebylow- the low-level controller. level controller. To avoid To avoid sudden sudden variations variations in belt in belt acceleration, acceleration, the the dynamics dynamics areareex- extended using a . The final control input (v ) can be obtained by time-domain tended using ac. The final control input (vc) can be obtained by time-domain integration of c c integration ac.aTo of c . To maintain maintain thethe user user at the at the center center position, position, the the desired desired position position (xd) (x ) is to isdset setzero to zero (i.e., (i.e., x = 0) and saturation is applied to xd = 0)d and saturation is applied to vc to remove vc to remove the oscillatory command due to thetonega- the oscillatory command due the negative position error when a user does not tive position error when a user does not intend to walk. intend to walk. Figure11. Figure 11.High-level High-levelLI LIcontroller controllerfor foraa1D 1Dtreadmill. treadmill. As shown As shown in Figure 12, for for expanding expandingto tothe the2D 2Dtreadmill treadmillmodel, model,the user the should user should be able be ableto maintain to maintainforward locomotion forward locomotionwhile changing while the the changing walking direction. walking Thus, direction. the con- Thus, the controller uses user troller uses thethe measured measured userorientation user orientationangle angle(θ) (θ) to to create the the user usercoordinates coordinates(X(X user,, Y ). The X -axis is set up along the anterior-posterior (AP) direction of the user’s body, user user and the Yuser-axis is along the medio-lateral (ML) direction. The walking intention ( ), referred to in treadmill coordinates, is converted to in the user coordinates. To in- terface with the walking intention , the control command = [ , , ] is generated by the high-level controller with respect to the user coor-
Appl. Sci. 2021, 11, 4223 13 of 20 Yuser ). The Xuser -axis is set up along the anterior-posterior (AP) direction of the user’s body, and the Yuser -axis is along the medio-lateral (ML) direction. The walking intention (vw ), referred to in treadmill coordinates, is converted to vuser w in the user coordinates. To interface h iT X, user with the walking intention vuser w , the control command vc user = vc vY,user c is generated by the high-level controller with respect to the user coordinates. Then, this control command is transformed into treadmill coordinates as vc . According to the adaptive treadmill controller, which can be considered to be a cascade system [28], the linear growth rate of the interconnection term and the convergence property represented by Equation (6) guarantee the stability of the entire system. Thus, the control command (vc ) for interfacing Appl. Sci. 2021, 11, x FOR PEER REVIEW with the 2D gait information is derived as follows: 13 of 20 vc =Rz,θ vuser c = Rz,θ (v̂user w +µ user ), (7) where , is the orientation matrix of the Z-axis according to the orientation of the user, where Rz,θ is the orientation matrix of the Z-axis according to the orientation of the user, is a continuous input using the RISE controller [29] to compensate for the user po- µuser is a continuous input using the RISE controller [29] to compensate for the user position sition error due to error in the estimation of user , and is the 2D feed-forward term error due to error in the estimation of vw , and v̂w is the 2D feed-forward term based on based on the user’s coordinates, which is simply expanded from [27], as follows: the user’s coordinates, which is simply expanded from [27], as follows: w = ko ( p vˆ user -ξ ) ., ξ = -vuser +ko ( puser -ξ ) , user c (8) v̂user w =ko (p user -ξ), ξ=-vuser c +ko (p user -ξ), (8) user, yuser]T represents the position error in the user coordinate system, ko∈ where user = [xuser where × p = [x , yuser ]T represents the position error in the user coordinate system, is a diagonal positive constant matrix which determines the convergence rate of the ko ∈ R2×2 is a diagonal positive constant matrix which determines the convergence rate observer output to the true value of stably [30], and ξ is the state of the feed-forward of the observer output to the true value of v̂user stably [30], and ξ is the state of the feed- w the term, Finally, the total control law including feedback command ( ) in the user forward term, Finally, the total control law including the feedback command (µuser ) in the coordinate system can be derived as follows: user coordinate system can be derived as follows: p user +α1puser dt +Zβ t sgn (. puser user +α1puser) dt t t t vuser c = vˆ user Z q( t ) dt + ( ks + IZ) αt2 .user w +tka user = v̂user q t0 dt + (ks p0 user 0 user (9) vc w + ka + I) α2 +α1 p dt + β sgn p +α1 p dt, (9) 0 0 , 0 ×2 whereαα11∈ where ∈ R 2× andkkss∈ R2××2 are and are diagonal positive matrices, matrices, β, α22 ∈ R are β, kaa and α arepositive positive user − RT v. The applied constants, constants,and andqqisisthe theerror of of error thethe velocity velocitycommand command defined as vas defined c − z,θ , v. The ap- RISE pliedcontrol schemescheme RISE control for the for uncertainty compensation the uncertainty reducesreduces compensation positionposition error viaerror a closed- via a loop system while closed-loop system guaranteeing stability stability while guaranteeing and asymptotic convergence and asymptotic of the position convergence of theerror posi- by the tion estimation error property property by the estimation of the RISE control of the schemescheme RISE control with the with applied observer the applied [22]. observer Similar [22]. Similar procedure of the high-level control design for a user-driven treadmill suchas procedure of the high-level control design for a user-driven treadmill such as observer-based observer-basedcontrol controlisisalso alsoperformed performedin inthe theother otherresearch research[30]. [30]. Figure12. Figure 12. Expansion of high-level high-level LI LI controller; controller;2D 2Dlocomotion locomotioncontrol controlcommands commandsand andthe thedead dead zonein zone inthe thesaturation saturationdirection directionof ofthe themotion motioncommand commandininthe theuser usercoordinates. coordinates. Moreover,the Moreover, thestability stabilityissue issueby bythe theapplied appliedsaturation saturationisiseliminated eliminatedby byaasupervisory supervisory algorithmwhich algorithm whichperform performinitializing initializing to to thethe integral integral term term of RISE of the the RISE controller controller whenwhen a usera user stays behind the reference position [22]. It also solves the chattering problem that usually affects sliding mode control (SMC) systems [29]. To apply the designed controller to the 2D treadmill, the dead zone should be defined with respect to the user’s ML direction (Yuser-axis), as shown in Figure 12. Although the user walks straight ahead, the waist shows a swaying motion in the ML direction. The ML motions of a user during walking can affect the treadmill controller by inducing a contin-
You can also read