OpenWHEEL i3r A new architecture for clearance performance

Transcription

1 LaMI Mechanical Engineering Research roup TIMS UBP IFMA Research Federation Blaise Pascal University Clermont-Ferrand II French Institute for Advanced Mechanics A new architecture for clearance performance Jean-Christophe.Fauroux@ifma.fr Belhassen-Chedli.Bouzgarrou@ifma.fr Frédéric.Chapelle@ifma.fr IFMA Campus de Clermont-Ferrand / Les Cézeaux, B.P AUBIERE Cedex FRANCE 1

2 Hybrid Locomotion Locomotion systems can be defined as poly-articulated mechanical systems that interact with environment via a set of unilateral adherent or slipping contacts to the ground. These contacts may change in nature and number according to time and space [Ben Amar, Bidaud 2007] Hybrid Hybrid locom. locom. Hybrid Hybrid robots robots Objectives Objectives Terrestrial vehicles & robots Hybrid locomotion Combining the advantages of wheel and leg Other solution: deformable frame Our objective: Improving locomotion Wheeled robots prevail (excellent energetic efficiency) Lack of agility / blocked on obstacles Legs / Tracks interesting for all terrain / climbing Wheeled robots That climb step obstacles With only four wheels And a stable behaviour 2

3 Existing Hybrid Mobile Robots Categories of Hybrid robots Hybrid Hybrid locom. locom. Hybrid Hybrid robots robots Wheels on legs vs. deformable frame Active / passive Difficulties: stiffness, power, control RobuROC 6 (150 kg, 1.5m long) Active deformable frame 3 tiltable axles with passive warping Able to turn on itself, climb obstacles WorkPartner (230 kg, 1.4m long, 7km/h) 4 wheels on legs (3 DOF per leg) Steering via central joint Many locomotion modes automation.tkk.fi Objectives Objectives Shrimp Passive deformable frame 6 wheels on 2 // bogies and 1 front linkage Excellent climbing abilities but requires 6 wheels 3

4 Objectives of the Work Within the project End of the car central-engine paradigm New articulated frames project, an open architecture for hybrid wheeled robots Wireless connection Hybrid Hybrid locom. locom. S 32 A3 Hybrid Hybrid robots robots Rear Control A2 W22 Wheel W31 S 21 W21 W12 Camera S22 S 12 A1 Control I2 S 31 Objectives Objectives CAN Bus W 32 Control I1 S 11 W11 nt Fro Z X Y Suspension mechanism Saw Swing arm Double wishbone Innovative suspension Inter-axle mechanism Ia Serial Parallel Innovative mechanism mechanism mechanism Objectives of this work Study the specific robot Analyse its behaviour during obstacle climbing Describe three implemented demonstrators 4

5 Architecture Exploring wheel (W12) Wheel (W12) joint R12 Architecture Architecture y12 Central warping joint R0 Rear frame (F2) Rear steering joint R2 Climbing Climbing proc. proc. A22 O22 W22 P22 Rear axle steering angle 2 B12 zf1 yf2 O ya2 B22 x ' O21 O11 P21 rw x0 hs Step Obstacle Advantages W21 O0 Front steering joint R1 Exploring front axle (A1) Lifting polygon Stability on three wheels B21 y0 A11 1 P11 A21 z0 OA1 2 W11 xa2 H Reconfigured rear axle (A2) xa1 B11 2 =xf F1 Front axle steering angle 1 2 =O F F1 T2 za2 za1 T1 ya1 zf2 OA2 x12 A12 O12 W12 Warping angle 0 Front frame (F1) z12 Only four wheels like most vehicles Active frame with only one actuator Stable when climbing obstacle 5 Precise steering via double Ackermann

6 Climbing Process Properties for stability Architecture Architecture Checking stability on three wheels Climbing Climbing proc. proc. Wheel W11 (front-right) Wheel W12 (front-left) Wheel W21 (rear-right) 1) 2) 3) W12 Wheel W22 (rear-left) 4) W11 W22 W21 5) Rear axle steering 2D model Stable if the lifted wheel is inside the turn Climbing process in 19 stages and 6 phases Front axle steering Can be stable on three wheels Axle steering - does not change the position of the centre of mass - changes the position of the contact points W11 6) W12 7) 8) W21 Stable Unstable Stable W22 Unstable 6

7 Climbing Process A - Prepairing Low W22 W11 W21 2 W12 B - W11 climbing High 1 Wheel center motion 6 C - W12 climbing Wheel lifting Architecture Architecture Wheel landing Support polygon For a very stable configuration (Four contact points) For a stable configuration (Three contact points) Climbing Climbing proc. proc. 10 D oing forward 11 E - W21 climbing 15 F - W22 climbing 16-7

8 Climbing Process Modeling and testing 2D model very helpful to build the complete climbing process Not acceptable for high pitch angles or strong warping Architecture Architecture Climbing Climbing proc. proc. 3D Validation required 3D Multibody model with Adams at several s Vi o de 1 8

9 at two s Full (V3 1.5m) V1 V1 Small Small V2 V2 Small Small V3 V3 Full Full Validates the climbing process (independent of the ) Safe work in a reduced space, Plastic elements lower cost Comparable to ATV size Metal components allow good stiffness (central joint) More features (clutch, better sensors) Price, complexity and development time increase Parameter Definition OW V1 OW V2 OW V3 b Wheelbase length T1T2 175 mm 260 mm 1210 mm t Track width Oa1Oa2 190 mm 151 mm 1250 mm rw Wheel radius 25 mm 25 mm 190 mm hl Leg height 72 mm 105 mm 500 mm m Mass 1530 g 2330 g < 200kg k0 Reduction ratio for central joint ,9 kw Reduction ratio for the wheels ,3 Small Full sc. High level control hardware (sensors, programmable logic controller) and software kept unchanged at all s Modular elements minimal re-design Small (V1 and V2 0.3m) 9

10 V1 Mindstorms RCX robotics kit 02 The first to validate the autonomous climbing process Open loop (no coders in the motors) 9V DC actuators (1.1W) Two controllers synchronized by IR message exchange V1 V1 Small Small mm V2 V2 Small Small 2 V3 V3 Full Full W12 b = 17 5 m m 90 1 t = h l = 72 mm 05 m m eo d Vi 2 06 W22 Overconstrained dual worm gear architecture with zero clearance 10

11 V2 Mindstorms NXT robotics kit Closed loop control with access to PID parameters Powerful 9V actuators (5W) 9V 8 Ah battery Motor multiplexer V1 V1 Small Small V2 V2 Small Small V3 V3 Full Full Wifi remote 9V actuator (5W) Overconstrained zero clearance twin transmission 11

12 V2 Powerful and expandable control architecture V1 V1 Small Small V2 V2 Small Small V3 V3 Full Full Opensource API Atmel 32-bit ARM processor at 48MHz, 256 Kb flash Bluetooth / USB2 3 motors outputs / 4 sensor inputs / multiplexers with daisy chain Only one controller required 3D accelerometer Many sensors : contact, angular, Rotation sensor distance (US), 3D accelerometer to measure tilting NBC assembler language NXC C-like language Sensor multiplexer Contact sensor 12

13 V3 Full 1.85m long, 1.38m wide, 0.98m high, 200kg 24V DC actuators 330W with peak torque of 100Nm + external encoder Two or four 12V 48Ah traction batteries V1 V1 Small Small 24V actuator (330W) V2 V2 Small Small V3 V3 Full Full Dual-stage 10.9x chain transmission ATV tire 13

14 V3 Mechanical improvement V2 V2 Small Small V3 V3 Full Full Same high level NXT controller enerates 8kHz PWM Pulse Width Modulated motor signal Used as an average tension to control low level Curtis DC controler 12V 48Ah traction battery Electromagnetic clutch Control architecture V1 V1 Small Small Central clutch no overconstraint when rolling no suspension required Fast 45 central warping : only one second (15x faster than V1) Battery and payload front cradle Curtis programmable DC controller 14

15 Steering reconfiguration Problem The rear axle must be orientated to improve stability before wheel lifting S2 The rear axle steering angle 2 must reach π/ Rotation sensor Steering Steering Non-Symmetry Non-Symmetry Coupling Coupling Solution Steering condition in function of track width t and wheel radius rw 2rw S2 S2 S2 21= 22 =t 2 /2 r w Unsafe because relies on non-slipping hypothesis More robust solution : yf2 Adding a rotation sensor for each axle ya2 S2 2 t/2 15

16 Front-Rear Non-Symmetry Real testing revealed a non-symmetric behaviour Axle A1 climbs easily whereas axle A2 has difficulties mm W12 Non-Symmetry Non-Symmetry b = 17 5 Coupling Coupling 2 CW 150 g m m 90 1 t = h l = 72 mm Steering Steering m m Mass: 1.5 kg W22 Solved by adding a counter-weight CW of 150 g Best obstacle height : 55 mm, 67% of the height of the centre of mass 16

17 Front-Rear Non-Symmetry Explanation a) Non-symmetry was not predicted because of 2D approximation On flat ground, the 2D model is exact With some pitch - In 2D, stability margin P 2 ' =b cos /2 P 2 ' =b cos /2 hl hs /b - In 3D Stability is favoured at stages 3 and 7 and penalized at stages 12 and 16 Adding CW equilibrates the climbing capacities of front and rear axles 2D model: On flat ground b) 2D model: On a step 3D model: stability margin nullifies b Steering Steering Non-Symmetry Non-Symmetry Coupling Coupling ' hl z0 O0 rw ' x0 P2 P1 P1 z0 x0 P2 O0 3D model with counterweight CW: stability is back hs ' 3D model: On a step c) f) e) ' O0 CW P1 rw x0 ' hl z0 Example: Stage 12 2D model: supposed stable d) P2 ' 17

18 Steering-Warping Coupling Problem When the robot is no more horizontal, steering reconfiguration generates induced axle warping Without warping correction With warping correction Steering Steering Non-Symmetry Non-Symmetry Coupling Coupling Solutions To cancel this phenomenon, the central joint must be unwarped With a central clutch, another solution is to declutch and let the central joint undergo passive rotation 18

19 Steering-Warping Coupling Analytic solution Steering Steering Front frame (F1) is supposed fixed to the ground and submitted to a roll angle α and a pitch angle β Wheel center point O21 comes from the following combination of rotation and translations O21=R y. R x.t x b /2. R x 0.T x b/ 2.R z 2. T z hl. T y t / 2.T 1 The altitude of O21 is z O21=cos [cos Non-Symmetry Non-Symmetry Coupling Coupling t t t sin 0 cos 2 hl cos 0 sin cos 0 cos 2 h l sin 0 ] sin [ sin 2 b] (2) The difference of altitude of O21 and O22 is: zo22 zo21=t [ cos cos 2 sin 0 sin sin 2 ] A condition for non warping is that O21 and O22 have the same altitude This gives the non warping condition: sin 0 = tan tan 2 (1) (3) (4) The roll and pitch α and β can be measured by inclinometers so θ2 θ0 19

20 Main results : a hybrid mobile robot with deformable frame Can climb 2/3 of centre of mass altitude with only 4 wheels Only 1 supplemental actuator minimal actuation, good stiffness Problems & solutions : front-rear non-symmetry, steering-warping coupling 3 small and full implementations / 1 control architecture Towards new robots & vehicles for agile all-terrain mobility & clearance performance IFMA / UBP students for the project Laurent ENEVAY et al. Romain CARTAILLER & Farhat HZA 20