1 @ McGraw-Hill Education Lecture 2 (III-LT1) Robot Kinematics (Ch. 5) by S.K. Saha Aug. 07, 2015 (F)@JRL301 (Robotics Tech.) PROPRIETARY MATERIAL. © 2014, 2008 The McGraw-Hill Companies, Inc. All rights reserved. No part of this PowerPoint slide may be displayed, reproduced or distributed in any form or by any means, without the prior written permission of the publisher, or used beyond the limited distribution to teachers and educators permitted by McGraw-Hill for their individual course preparation. If you are a student using this PowerPoint slide, you are using it without permission. 2 @ McGraw-Hill Education Pose Position + Rotation P Z W p΄ p Translation: 3 Rotation: 3 Total: 6 OM V M o O Y F U X A moving body Pose or Configuration 3 @ McGraw-Hill Education Position Description [p] F p x py p z . . . (5.8) p = px x + py y + pz z . . . (5.9) 0 0 1 [x]F 0 , [y ]F 1 , and [z ]F 0 0 0 1 . . . (5.10) 4 @ McGraw-Hill Education Orientation Description 1. Direction cosine representation 2. Fixed-axes rotations 3. Euler angles representation 4. Single- and double-axes rotations 5. Euler parameters representation I will illustrate the first TWO only 5 @ McGraw-Hill Education Direction Cosine Representation Refer to Fig. 5.12 p = puu + pvv + pww . . . (5.12) u = ux x + uy y + uz z . . . (5.11a) v = vx x + vy y + vz z . . . (5.11b) w = wx x + wy y + wz z . . . (5.11c) 6 @ McGraw-Hill Education Substitute eqs. (5.11a-c) into eq. (5.12) p = (puux + pvvx + pwwx)x + (puuy + pvvy + pwwy)y + (puuz + pvvz + pwwz)z . . . (5.13) px = uxpu + vxpv + wxpw . . . (5.14a) py = uypu + vypv + wypw . . . (5.14b) pz = uzpu + vzpv + wzpw . . . (5.14c) [p]F = Q [p]M . . . (5.15) 7 @ McGraw-Hill Education [p]F = Q [p]M . . . (5.15) p p uT x vT x w T x u v w x x x x u T T T [p] p , [p]M p , Q u v w u y v y w y F y y y y T Tz wTz v u z v u v w p z z z p w z .. . (5.16) Orientation description 1 uTu = vTv = wTw = 1, and uTv(vTu) = uTw(wTu) = vTw(wTv) = 0 Q is called Orthogonal … (5.17) 8 @ McGraw-Hill Education Due to orthogonality u v = w, v w = u, and w u = v . . . (5.18) QTQ = QQT = 1 ; det (Q) = 1; Q1 = QT . . . (5.19) 9 @ McGraw-Hill Education Interpretation 1 Example 5.6 Rotations [Elementary] (Fig. 5.13a) Cα [u]F Sα , 0 Sα [ v ]F Cα , 0 0 [ w ]F 0 1 . . . (5.20) 10 @ McGraw-Hill Education C Q Z S 0 C QY 0 S S C 0 0 1 0 0 0 1 S 1 0 ; Q X 0 0 C . . . (5.21) 0 C S 0 S C . . . (5.22) 11 @ McGraw-Hill Education Non-commutative Property: An Illustration Fig. 5.20 Successive rotation of a box about Z and Y-axes 12 @ McGraw-Hill Education Non-commutative Property (contd.) Fig. 5.21 Successive rotation of a box about Y and Z-axes 13 @ McGraw-Hill Education Lecture 3 Robot Kinematics (Ch. 5) by S.K. Saha Aug. 10, 2015 (M)@JRL301 (Rob. Tech.) PROPRIETARY MATERIAL. © 2014, 2008 The McGraw-Hill Companies, Inc. All rights reserved. No part of this PowerPoint slide may be displayed, reproduced or distributed in any form or by any means, without the prior written permission of the publisher, or used beyond the limited distribution to teachers and educators permitted by McGraw-Hill for their individual course preparation. If you are a student using this PowerPoint slide, you are using it without permission. 14 Recap @ McGraw-Hill Education • Orientation representations – Non-commutative • Direction cosines: Has disadv. of 9 param. • Fixed-axes (RPY) rotations (12 sets) 15 @ McGraw-Hill Education Homogeneous Transformation P Z W p΄ p OM V M o O Y F U X Task: Point P is known in moving frame M. Find P in fixed frame F. Fig. 5.23 Two coordinate frames 16 @ McGraw-Hill Education p = o + p . . . (5.45) [p]F = [o]F + Q[p’]M [p] F Q 1 0 T . . . (5.46) [o]F [p] M . . . (5.47) 1 1 [p] F T [p] M . . . (5.48) Homogenous Transformation 17 @ McGraw-Hill Education T: Homogenous transformation matrix (4 4) TTT 1 T 1 or T1 TT Q T T 0 Q . . . (5.49) [o] F 1 T . . . (5.50) Example 5.10 Pure Translation 1 0 T 0 0 Fig. 5.24 (a) 0 2 0 1 1 0 0 1 . . . (5.51) 0 1 0 0 18 @ McGraw-Hill Education Example 5.11 Pure Rotation Fig. 5.24 (b) C 30o S 30o o o S 30 C 30 T 0 0 0 0 3 1 0 0 2 2 1 3 0 0 2 2 0 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 . . . (5.52) 19 @ McGraw-Hill Education Non-commutative Property Like rotation matrices homogeneous transformation matrices are non-commutative, i. e., TATB TBTA 20 @ McGraw-Hill Education Denavit and Hartenberg (DH) Parameters • Serial chain - Two links connected by revolute or prismatic joint • Four parameters – Joint offset (b) – Joint angle () – Link length (a) – Twist angle () Fig. 5.27 Serial manipulator 21 @ McGraw-Hill Education • Joint axis i: Link i-1 + link i • Link i: Fixed to frame i+1 (Saha) / frame i (Craig) DH Saha Variables bi and i [Screw@Z] XiXi+1@Zi Constants ai and i [Screw@X] ZiZi+1@Xi+1 Craig Xi-1Xi@Zi ZiZi+1@Xi Zi+1 Z’’’i 22 @ McGraw-Hill Education • bi (Joint offset): Length of the intersections of the common normals on the joint axis Zi, i.e., Oi and Oi. It is the relative position of links i 1 and i. This is measured as the distance between Xi and Xi + 1 along Zi. Zi+1 Z’’’i 23 @ McGraw-Hill Education • i (Joint angle): Angle between the orthogonal projections of the common normals, Xi and Xi + 1, to a plane normal to the joint axes Zi. Rotation is positive when it is made counter clockwise. It is the relative angle between links i 1 and i. This is measured as the angle between Xi and Xi + 1 about Zi. Zi+1 Z’’’i 24 @ McGraw-Hill Education • ai (Link length): Length between the O’i and Oi +1. This is measured as the distance between the common normals to axes Zi and Zi + 1 along Xi + 1. Zi+1 Z’’’i 25 @ McGraw-Hill Education • i (Twist angle): Angle between the orthogonal projections of joint axes, Zi and Zi+1 onto a plane normal to the common normal. This is measured as the angle between the axes, Zi and Zi + 1, about axis Xi + 1 to be taken positive when rotation is made counter clockwise. Zi+1 Z’’’i 26 @ McGraw-Hill Education Revolute Joint • DH@Z (Variable) • DH@X (Const.) – Joint offset (b) – Joint angle () – Link length (a) – Twist angle () Zi+1 Z’’’i Fig. 5.28 27 @ McGraw-Hill Education Mathematically • Translation along Zi Tb = 1 0 0 0 0 0 1 0 0 1 0 0 0 0 bi 1 . . . (5.60a) • Rotation about Zi T = Cθ i Sθ i 0 0 S θi Cθ i 0 0 0 0 1 0 0 0 0 1 . . . (5.60b) 28 @ McGraw-Hill Education • Translation along Xi+1 1 0 Ta = 0 0 0 1 0 0 0 0 1 0 ai 0 0 1 . . . (5.60c) • Rotation about Xi+1 T = 1 0 0 0 0 Cαi 0 S αi Sαi 0 Cαi 0 0 0 0 1 . . . (5.60d) 29 @ McGraw-Hill Education • Total transformation from Frame i to Frame i+1 Ti = TbTTaT . . . (5.61a) Do it yourself! Sθ i Cα i Sθ i Sα i Rotation Cθ Cα Cθ Sα Matrix Sα Cα i i i i 0 i 0 i ai C i ai Sθ i bi 1 Position Cθ i Sθ Ti = i 0 0 . . . (5.61b) 30 @ McGraw-Hill Education Spherical-type Arm • DH-parameters i bi 1 0 2 bFill-up 0 /2 the DH parameters 2 2 (JV) 3 b3 (JV) ai i Link 1 (JV) 0 /2 0 0 0 Fig. 5.32 A spherical arm 31 @ McGraw-Hill Education PUMA 560 Fig. 5.35 PUMA 560 and its frames i Variable DH bi i Constant DH ai i 1 0 1 0 -/2 2 0 2 a2 0 3 B3 3 a3 -/2 4 b4 4 0 /2 5 0 5 0 -/2 6 0 6 0 0 32 @ McGraw-Hill Education Lecture 4 (SIT Sem. Rm.) Forward and Inverse Kinematics (Ch. 6) by S.K. Saha Aug. 12, 2015 (W)@JRL301(Robotics Tech.) PROPRIETARY MATERIAL. © 2014, 2008 The McGraw-Hill Companies, Inc. All rights reserved. No part of this PowerPoint slide may be displayed, reproduced or distributed in any form or by any means, without the prior written permission of the publisher, or used beyond the limited distribution to teachers and educators permitted by McGraw-Hill for their individual course preparation. If you are a student using this PowerPoint slide, you are using it without permission. 33 @ McGraw-Hill Education Forward and Inverse Kinematics Forward: One soln. Multiply + Add Inverse: 1st soln. . Inverse: nth soln. Solve Non-lin. eqns. 34 @ McGraw-Hill Education Three-link Planar Arm • DH-parameters Link bi 1 2 3 0 0 0 i ai i 1 (JV) a1 the DH a2 Fill-up (JV) 2parameters 3 (JV) a3 0 0 0 • Frame transformations (Homogeneous) Ti = Cθ i Sθ i 0 a i C i with the elements SθFill-upCθ 0 a i Sθ i , i i 0 0 1 0 0 0 0 1 Fig. 5.29 A three-link planar arm for i=1,2,3 35 @ McGraw-Hill Education DH Parameters of Articulated Arm Link bi i ai i 0 π/2 1 0 1 (JV) 2 0 2 (JV) a2 0 3 0 3 (JV) a3 0 36 @ McGraw-Hill Education Matrices for Articulated Arm c1 s T1 1 0 0 c1 c 23 s c 1 23 T s 23 0 s1 0 c1 0 1 0 0 0 0 1 c3 s T3 3 0 0 c2 s T2 2 0 0 0 0 s3 c3 0 0 0 0 1 0 - c1 s 23 s1 s 23 s1 c1 c 23 0 0 0 s2 c2 0 0 0 0 1 0 a2 c2 a2 s2 0 1 a3 c3 a3 s3 0 1 c1 (a 2 c 2 a 3 c 23 ) s1 (a 2 c 2 a 3 c 23 ) … (6.11) (a 2 s 2 a 3 s 23 ) 1 37 @ McGraw-Hill Education Inverse Kinematics • Unlike Forward Kinematics, general solutions are not possible. • Several architectures are to be solved differently. 38 @ McGraw-Hill Education Two-link Arm p y a1s1 a2 s12 c2 p x2 p y2 a12 a22 2 a1 a2 s2 1 c22 2 = atan2 (s2, c2) s1 py Y3 2 Y2 a1 1 X3 RoboAnalyzer p x a1c1 a2 c12 Y1 a2 X 2 2 1 px X1 (a1 a2 c2 ) p y a2 s2 p x c1 2 2 2 2 Δ a a 2 a a c p p 1 2 1 2 2 x y Δ (a1 a2 c2 ) p x a2 s2 p y Δ 1 = atan2 (s1, c1) 39 @ McGraw-Hill Education Inverse Kinematics of 3-DOF RRR Arm φ θ1 θ2 θ3 … (6.18a) p x a1c1 a2 c12 a3 c123 … (6.18b) p y a1 s1 a 2 s12 a3 s123 … (6.18c) wx p x a3 c φ a1 c1 a2 c12 … (6.19a) w y p y a3 s φ a1 s1 a2 s12 … (6.19b) 40 w2x + w2y = a12+ a22 + 2 a1a2c2 … (6.20a) w12 w22 a12 a 22 s2 1 c22 c2 2 a1 a 2 2 = atan2 (s2, c2) wy (a1 a2 c2 )s1 a2 c1 s2 s1 Δ … (6.20b,c) . . . (6.21) wx ( a1 a2c2 )c1 a2 s1s2 (a1 a2 c2 ) wy a2 s2 wx @ McGraw-Hill Education c1 … (6.22a) … (6.22b) (a1 a 2 c 2 ) wx a 2 s 2 w y Δ a12 a22 2a1a2c2 wx2 wy2 Δ … (6.23a,b) 1 = atan2 (s1, c1) . . . (6.23c) 3 = - 1 2 . . . (6.24) 41 @ McGraw-Hill Education Numerical Example • An RRR planar arm (Example 6.15). Input T 1 2 3 2 0 0 3 2 1 2 0 0 Rotation0 Matrix 0 1 0 5 Origin 3 2 of end effector 3 1 frame 2 0 1 4.23 1.86 0 where = 60o, and a1 = a2 = 2 units, and a3 = 1 unit. Do it yourself Verify using RoboAnalyzer 42 @ McGraw-Hill Education Using eqs. (6.13b-c), c2 = 0.866, and s2 = 0.5, 2 = 30o Next, from eqs. (6.16a-b), s1 = 0, and c1= 0.866. 1 = 0o. Finally, from eq. (6.17) , Therefore 3 = 30o. 1 = 0o 2 = 30o, and 3 = 30 …(6.30b) The positive values of s2 was used in evaluating 2 = 30o. The use of negative value would result in : 1 = 30o 2 = -30o, and 3 = 60o …(6.30c) 43 @ McGraw-Hill Education Watch • Forward and Inverse Kinematics: Watch 3/3 of IGNOU Lectures [29min] https://www.youtube.com/watch?v=duKD8cvtBTI • For more clarity: Watch 12 of Addis Ababa Lectures [77 min] [https://www.youtube.com/watch?v=NXWzk1toze4 • Robotics (13 of Addis Ababa Lectures): Inverse Kinematics [82 min] https://www.youtube.com/watch?v=ulP3YiJLiEM 44 @ McGraw-Hill Education Velocity Analysis Jacobian maps joint rates into end-effector’s velocities. It depends on the manipulator configuration. ω e twistof end - effector : t e ; ve t e Jθ 1 Joint rates : θ n where J j1 e1 J e1 a 1e jn and j2 e2 e 2 a 2e e n a ne en . . (6.86) 0 ei , if Joint i is prismatic ji , if Joint i is revolute ji e i a ie ei aie 45 @ McGraw-Hill Education Jacobian of a 2-link Planar Arm J e1 a1e e 2 a 2e where e1 e2 [0 0 1]T a1e a1 a 2 [a1c1 a2c12 a1s1 a2 s12 0]T a2e a2 [a2c12 a2 s12 0]T a1 s1 a2 s12 Hence, J a1 c1 a2 c12 a2 s12 a2 c12 46 @ McGraw-Hill Education Example: Singularity of 2-link RR Arm a1 s1 a 2 s12 J a1 c1 a 2 c12 a 2 s12 a 2 c12 2 = 0 or 47 @ McGraw-Hill Education Lecture 4 (SIT Sem. Rm.) Forward and Inverse Kinematics (Ch. 6) by S.K. Saha Aug. 12, 2015 (W)@JRL301(Robotics Tech.) PROPRIETARY MATERIAL. © 2014, 2008 The McGraw-Hill Companies, Inc. All rights reserved. No part of this PowerPoint slide may be displayed, reproduced or distributed in any form or by any means, without the prior written permission of the publisher, or used beyond the limited distribution to teachers and educators permitted by McGraw-Hill for their individual course preparation. If you are a student using this PowerPoint slide, you are using it without permission. 48 @ McGraw-Hill Education Statics and Manipulator Design (Ch. 7) PROPRIETARY MATERIAL. © 2014, 2008 The McGraw-Hill Companies, Inc. All rights reserved. No part of this PowerPoint slide may be displayed, reproduced or distributed in any form or by any means, without the prior written permission of the publisher, or used beyond the limited distribution to teachers and educators permitted by McGraw-Hill for their individual course preparation. If you are a student using this PowerPoint slide, you are using it without permission. 49 @ McGraw-Hill Education Principle of Virtual Work T w e x τ θ T … (7.28) • Relation between two virtual displacements (Can be derived from velocity expression) x Jθ T w e Jθ … (7.29) w J τ … (7.31) T e τ θ T τ J we T T … (7.32) 50 @ McGraw-Hill Education Example: 2-link RR Planar Arm τ1 [e ] [n01 ]1 T 1 1 a1 f x sθ2 (a2 a1cθ2 )f y τ 2 [e ] [n12 ]2 a2 f y T 2 2 τJ f T τ1 T a1sθ 2 a1cθ2 a2 τ J 0 a τ 2 2 fx 0 f f y 0 0 51 @ McGraw-Hill Education Two Jacobian Matrices • From Statics a1 sθ 2 J a1cθ 2 a2 0 0 a2 0 • From a1 s1 a 2 s12 Kinematics J a1 c1 a 2 c12 a 2 s12 a 2 c12 52 @ McGraw-Hill Education Jacobian from Statics in Frame 1 0 c1 s1 0 c 2 s 2 0 a1sθ 2 [J ]1 s1 c1 0 s 2 c 2 0 a1cθ2 a2 a2 0 0 1 0 0 1 0 0 a1sθ1 a2 sθ12 a2 sθ12 a1cθ1 a2 cθ12 a2 cθ12 0 0 … (7.34) • Without the last row, it is the same as the one from kinematics Should be! 53 @ McGraw-Hill Education Manipulator Design • High investment in robot usage low technological level of mechanical structure • Functional Requirements • Kinetostatic Measures • Structural Design and Dynamics • Economics 54 @ McGraw-Hill Education Functional Requirements of a Robot • Payload • Mobility • Configuration • Speed, Accuracy and Repeatability • Actuators and Sensors 55 @ McGraw-Hill Education bmin b bmax, for 0o 360o 56 @ McGraw-Hill Education Dexterity and Manipulability • Dexterity wd det(J) • Manipulability wm … (7.44) T det(JJ ) • Nonredundant manipulator square Jacobian wm det(J ) wd wm 57 @ McGraw-Hill Education Motor Selection (Thumb Rule) • Rapid movement with high torques (> 3.5 kW): Hydraulic actuator • < 1.5 kW (no fire hazard): Electric motors • 1-5 kW: Availability or cost will determine the choice 58 Simple Calculation @ McGraw-Hill Education 2 m robot arm to lift 25 kg mass at 10 rpm • Force = 25 x 9.81 = 245.25 N • Torque = 245.25 x 2 = 490.5 Nm • Speed = 2 x 10/60 = 1.047 rad/sec • Power = Torque x Speed = 0.513 kW • Simple but sufficient for approximation 59 Practical Application @ McGraw-Hill Education Trapezoidal Trajectory Subscript l for load; m for motor; G = l/m (< 1); : Motor + Gear box efficiency 60 @ McGraw-Hill Education Accelerations & Torques Ang. accn. during t1: Ang. accn. during t2: Zero (Const. Vel.) Ang. accn. during t3: Torque during t1: T1 = Torque during t2: T2 = Torque during t3: T3 = 61 RMS Value @ McGraw-Hill Education 62 @ McGraw-Hill Education Motor Performance 63 @ McGraw-Hill Education Final Selection • Peak speed and peak torque requirements , where TPeak is max of (magnitudes) T1, T2, and T3 • Use individual torque and RMS values + Performance curves provided by the manufacturer. • Check heat generation + natural frequency of the drive. 64 @ McGraw-Hill Education Dynamics and Control Measures • Rule of Thumb 1 n r 2 … (7.51) n : closed-loop natural frequency r : lowest structural resonant frequency 65 @ McGraw-Hill Education Manipulator Stiffness 1 1 1 2 ke k1 k2 … (7.48) ke equivalent stiffness gear ratio 66 @ McGraw-Hill Education Link Material Selection • Mild (low carbon) steel: Sy = 350 Mpa; Su = 420 Mpa • High alloyed steel Sy = 1750-1900 Mpa; Su = 2000-2300 Mpa • Aluminum • Sy = 150-500 Mpa; Su = 165-580 Mpa 67 @ McGraw-Hill Education Driver Selection • Driver of a DC motor: A hardware unit which generates the necessary current to energize the windings of the motor • Commercial motors come with matching drive systems 68 @ McGraw-Hill Education Summary • Forward Kinematics • Inverse kinematics – A spatial 6-DOF wrist-portioned has 8 solutions • Velocity and Jacobian • Mechanical Design 69 @ McGraw-Hill Education Thank You saha@mech.iitd.ac.in http://sksaha.com