Lecture 10 Forward Kinematics Katie DC Feb

Total Page：16

File Type：pdf, Size：1020Kb

lecture 10 forward kinematics Katie DC Feb. 25, 2020 Modern Robotics Ch. 4 Admin • Guest Lecture on Thursday 2/27 • Quiz 1 re-take is this week • HW3: PF Implementation due next week • HW5 due this week • I’ll be posting a course feedback form this week What is Forward Kinematics? • Kinematics: a branch of classical mechanics that describes motion of bodies without considering forces. AKA “the geometry of motion” • Forward kinematics: a specific problem in robotics. Given the individual state of each joint of the robot (in a local frame), what is the position of a given point on the robot in the global frame? Where is forward kinematics used? Forward Kinematics of a Simple Chain Assumptions • Our robot is a kinematic chain, made of rigid links and movable joints • No branches or loops (will discuss later) • All joints have one degree of freedom and are revolute or prismatic Review of Screw Motions We derived a representation for the screw axis: 휔 풮 = ∈ ℝ6 푣 where either • 휔 = 1 • where 푣 = −휔 × 푞 + ℎ휔, where 푞 is the point on the axis of the screw and ℎ is the pitch of the screw • 휔 = 0 and 푣 = 1 Screw Motions as Matrix Exponential The screw axis 풮푖 can be expressed in matrix form as: 휔 푣 풮 = 푖 ∈ 푠푒 3 푖 0 0 To express a screw motion given a screw axis, we use the matrix exponential: 푒 풮 휃 ∈ 푆퐸 3 Modeling Robot Joints as Screw Motions Case 1: Revolute Joint • 휔 = 1 • 푣 = −휔 × 푞 + ℎω • 푣 = −휔 × 푞 Modeling Robot Joints as Screw Motions Case 2: Prismatic Joint • 휔 = 0 • 푣 = 1 • Axis of movement defines 푣 Product of Exponentials Approach Let each joint have a configuration defined by 휃푖 Initialization steps: • Choose a fixed frame 푠 • Choose an end-effector (tool) frame attached to the robot 푏 • Put all joints in zero position • Let 푀 ∈ 푆퐸 3 be the configuration of 푏 in the 푠 frame when the robot is in the zero position Product of Exponentials Approach • Given zero position 푀 • For each joint , define the screw axis • For each motion of a joint, define the screw motion • These operations compose nicely through multiplication, giving us the Product of Exponentials (PoE) formula! 푇 휃 = 푒 풮1 휃1푒 풮2 휃2 ⋯ 푒 풮푛−1 휃푛−1푒 풮푛 휃푛푀 Visualizing the Product of Exponentials Example 1 Example 1: PoE Compute 푒 풮푖 휃푖 for each joint: 휔푖 휃푖 2 푒 풮푖 휃푖 = 푒 퐼휃푖 + 1 − cos 휃푖 휔푖 + 휃푖 − sin 휃푖 휔푖 푣푖 0 1 and compose with 푀: 푇 휃 = 푒 풮1 휃1푒 풮2 휃2푒 풮3 휃3푀 Example 2 Summary • Learned the basics of forward kinematics, which gives us a model for computing position and orientation of the end-effector • Uses the product of exponentials formula to define this transformation as composed matrix multiplications • Next time, we’ll learn about exponentials in the end-effector frame and modeling robots with the Universal Robot Description Format.

Copy Link

Recommended publications

KINEMATIC ANALYSIS of a THUMB-EXOSKELETON SYSTEM for POST STROKE REHABILITATION By
KINEMATIC ANALYSIS OF A THUMB-EXOSKELETON SYSTEM FOR POST STROKE REHABILITATION By Vikash Gupta Thesis Submitted to the Faculty of the Graduate School of Vanderbilt University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE in Mechanical Engineering August, 2010 Nashville, Tennessee Approved: Professor Nilanjan Sarkar Professor Derek Kamper Professor Robert Webster III ACKNOWLEDGEMENTS I will like to extend my sincere thanks to Dr. Nilanjan Sarkar for his continuous support and guidance, for his encouragement and advice throughout my studies. He has always been a source of inspiration for me. I would also like to thank Dr. Derek Kamper from Rehabilitation Institute of Chicago, for providing valuable comments on the thesis and being supportive throughout the research. My deep gratitude and sincere thanks goes to my parents and my sister, Swati Gupta. Their invaluable support, understanding and encouragement from far home has always helped me to continue my research. I would also like to thank Hari Kr. Voruganti, Mukul Singhee, Abhishek Chowdhury, Sayantan Chatterjee and Raktim Mitra for their support. My sincere thanks goes to Nino Dzotsenidze, who has always supported and believed in me and helped me make the right decisions at all times. Without her support and encouragement, it would have not been possible. I would also like to thank and express my sincere gratitude to my colleagues and friends Milind Shashtri, Furui Wang, Yu Tian, Jadav Das and Uttama Lahiri at Robotics and Autonomous Systems Laboratory for their support. Last but not the least, I want to express my humble thanks and appreciation to Suzanne Weiss from the Mechanical Engineering department for always being supportive.
[Show full text]
Arxiv:2102.12942V3 [Cs.RO] 30 Mar 2021 Drawback of These Strategies Is That They Are Extremely Sensitive to Inaccuracies in the Model
Structured Prediction for CRiSP Inverse Kinematics Learning with Misspecified Robot Models Gian Maria Marconi∗,1,2 Raffaello Camoriano∗,2 Lorenzo Rosasco2,3,4 Carlo Ciliberto5 [email protected] raff[email protected] [email protected] [email protected] Abstract With the recent advances in machine learning, problems that traditionally would require accurate modeling to be solved analytically can now be successfully approached with data-driven strategies. Among these, computing the inverse kinematics of a redundant robot arm poses a significant challenge due to the non-linear structure of the robot, the hard joint constraints and the non-invertible kinematics map. Moreover, most learning algorithms consider a completely data-driven approach, while often useful information on the structure of the robot is available and should be positively exploited. In this work, we present a simple, yet effective, approach for learning the inverse kinematics. We introduce a structured prediction algorithm that combines a data-driven strategy with the model provided by a forward kinematics function – even when this function is misspecified – to accurately solve the problem. The proposed approach ensures that predicted joint configurations are well within the robot’s constraints. We also provide statistical guarantees on the generalization properties of our estimator as well as an empirical evaluation of its performance on trajectory reconstruction tasks. 1 Introduction Computing the inverse kinematics of a robot is a well-known key problem in several applications requiring robot control [20]. This task consists in finding a set of joint configurations that would result in a given pose of the end effector, and is traditionally solved by assuming access to an accurate model of the robot and employing geometric or numerical optimization techniques.
[Show full text]
Forward and Inverse Kinematics Analysis of Denso Robot
Proceedings of the International Symposium of Mechanism and Machine Science, 2017 AzC IFToMM – Azerbaijan Technical University 11-14 September 2017, Baku, Azerbaijan Forward and Inverse Kinematics Analysis of Denso Robot Mehmet Erkan KÜTÜK 1*, Memik Taylan DAŞ2, Lale Canan DÜLGER1 1*Mechanical Engineering Department, University of Gaziantep Gaziantep/ Turkey E-mail: [email protected] 2 Mechanical Engineering Department, University of Kırıkkale Abstract used Robotic Toolbox in forward kinematics analysis of A forward and inverse kinematic analysis of 6 axis an industrial robot [4]. DENSO robot with closed form solution is performed in This study includes kinematics of robot arm which is this paper. Robotics toolbox provides a great simplicity to available Gaziantep University, Mechanical Engineering us dealing with kinematics of robots with the ready Department, Mechatronics Lab. Forward and Inverse functions on it. However, making calculations in kinematics analysis are performed. Robotics Toolbox is traditional way is important to dominate the kinematics also applied to model Denso robot system. A GUI is built which is one of the main topics of robotics. Robotic for practical use of robotic system. toolbox in Matlab® is used to model Denso robot system. GUI studies including Robotic Toolbox are given with 2. Robot Arm Kinematics simulation examples. Keywords: Robot Kinematics, Simulation, Denso The robot kinematics can be categorized into two Robot, Robotic Toolbox, GUI main parts; forward and inverse kinematics. Forward kinematics problem is not difficult to perform and there is no complexity in deriving the equations in contrast to the 1. Introduction inverse kinematics. Especially nonlinear equations make the inverse kinematics problem complex.
[Show full text]
Kinematics, Kinematics Chains Previously
Kinematics, Kinematics Chains Previously • Representation of rigid body motion • Two different interpretations - as transformations between different coordinate frames - as operators acting on a rigid body • Representation in terms of homogeneous coordinates • Composition of rigid body motions • Inverse of rigid body motion Rigid Body Transform Translation only is the origin of the frame B expressed in the Frame A Composite transformation: {B} Transformation: Homogeneous coordinates The points from frame A to frame B are {A} transformed by the inverse of (see example next slide) Kinematic Chains • We will focus on mobile robots (brief digression) • In general robotics - study of multiple rigid bodies lined together (e.g. robot manipulator) • Kinematics – study of position, orientation, velocity, acceleration regardless of the forces • Simple examples of kinematic model of robot manipulator and mobile robot • Components – links, connected by joints Various joints Kinematic Chains Tool frame Base frame • Given determine what is • Given determine what is • We can control , want to understand how it affects position of the tool frame • How does the position of the tool frame change as the manipulator articulates • Actuators change the joint angles Forward kinematics for a 2D arm • Find position of the end effector as a function of the joint angles • Blackboard example Kinematic Chains in 3D • More joints possible (spherical, screw) • Additional offset parameters, more complicated • Same idea: set up frame with each link • Define relationship
[Show full text]
Learning Tracking Control with Forward Models
Learning Tracking Control with Forward Models Botond Bocsi´ Philipp Hennig Lehel Csato´ Jan Peters Abstract— Performing task-space tracking control on redun- or not directly actuated elements are hard to model accurately dant robot manipulators is a difficult problem. When the with physics and even harder to control. In our experiments, physical model of the robot is too complex or not available, a ball has been attached to the end-effector of a Barrett standard methods fail and machine learning algorithms can have advantages. We propose an adaptive learning algorithm WAM [9] robot arm using a string, and the ball had to for tracking control of underactuated or non-rigid robots where follow a desired trajectory instead of the end-effector of the the physical model of the robot is unavailable. The control robot. For this system even the forward kinematics model method is based on the fact that forward models are relatively is undefined since the same joint configuration can result in straightforward to learn and local inversions can be obtained different ball positions (due to the swinging of the ball). This via local optimization. We use sparse online Gaussian process inference to obtain a flexible probabilistic forward model and underactuated robot is not straightforward to control since the second order optimization to find the inverse mapping. Physical swinging produces non-linear effects that lead to undefinable experiments indicate that this approach can outperform state- forward or inverse models. We were able to control such of-the-art tracking control algorithms in this context. a robot and show that our method is capable of capturing non-linear effects, like the centrifugal force, into the control I.
[Show full text]
Planar Kinematics
Introduction to Robotics, H. Harry Asada 1 Chapter 4 Planar Kinematics Kinematics is Geometry of Motion. It is one of the most fundamental disciplines in robotics, providing tools for describing the structure and behavior of robot mechanisms. In this chapter, we will discuss how the motion of a robot mechanism is described, how it responds to actuator movements, and how the individual actuators should be coordinated to obtain desired motion at the robot end-effecter. These are questions central to the design and control of robot mechanisms. To begin with, we will restrict ourselves to a class of robot mechanisms that work within a plane, i.e. Planar Kinematics. Planar kinematics is much more tractable mathematically, compared to general three-dimensional kinematics. Nonetheless, most of the robot mechanisms of practical importance can be treated as planar mechanisms, or can be reduced to planar problems. General three-dimensional kinematics, on the other hand, needs special mathematical tools, which will be discussed in later chapters. 4.1 Planar Kinematics of Serial Link Mechanisms Example 4.1 Consider the three degree-of-freedom planar robot arm shown in Figure 4.1.1. The arm consists of one fixed link and three movable links that move within the plane. All the links are connected by revolute joints whose joint axes are all perpendicular to the plane of the links. There is no closed-loop kinematic chain; hence, it is a serial link mechanism. y A 3 E End Effecter ⎛ xe ⎞ Link 3 ⎜ ⎟ ⎝ ye ⎠ A 2 θ3 φ B e A1 Link 2 Joint 3 θ 2 Link 1 A Joint 2 Joint 1 θ O 1 x Link 0 Figure 4.1.1 Three dof planar robot with three revolute joints To describe this robot arm, a few geometric parameters are needed.
[Show full text]
Forward Kinematics of RRR Robot Manipulator (1/2)
1/39 Introduction Robotics dr Dragan Kosti ć WTB Dynamics and Control September-October 2009 Introduction Robotics, lecture 3 of 7 2/39 Outline • Recapitulation • Forward kinematics • Inverse kinematics problem Introduction Robotics, lecture 3 of 7 3/39 Recapitulation Introduction Robotics, lecture 3 of 7 4/39 Robot manipulators • Kinematic chain is series of links and joints. SCARA geometry • Types of joints: – rotary (revolute, θ) – prismatic (translational, d). Introduction Robotics, lecture 3 of 7 schematic representations of robot joints 5/39 Common geometries of robot manipulators Introduction Robotics, lecture 3 of 7 6/39 Forward kinematics problem • Determine position and orientation of the end-effector as function of displacements in robot joints. Introduction Robotics, lecture 3 of 7 7/39 DH convention for homogenous transformations (1/2) • An arbitrary homogeneous transformation is based on 6 independent variables: 3 for rotation + 3 for translation. • DH convention reduces 6 to 4, by specific choice of the coordinate frames. • In DH convention, each homogeneous transformation has the form: Introduction Robotics, lecture 3 of 7 8/39 DH convention for homogenous transformations (2/2) • Position and orientation of coordinate frame i with respect to frame i-1 is specified by homogenous transformation matrix: q xn z 0 0 q i+1 zn qi ‘0’ n y0 αi ‘ ’ yn zi ai x0 xi di z i-1 q where i xi-1 Introduction Robotics, lecture 3 of 7 9/39 Physical meaning of DH parameters • Link length ai is distance from zi-1 to zi measured along xi. q xn z 0 • Link twist α is angle between z 0 q i i-1 i+1 zn qi and z measured in plane normal ‘0’ n i y0 αi ‘ ’ yn zi ai to x (right -hand rule).
[Show full text]
Kinematic Modelling of the Human Hand for Robotics
Georg Stillfried Kinematic modelling of the human hand for robotics PhD Thesis TECHNISCHE UNIVERSITÄT MÜNCHEN Fakultät für Informatik Lehrstuhl für Echtzeitsysteme und Robotik Biomimetic Robotics and Machine Learning Kinematic modelling of the human hand for robotics Dipl.-Ing. Univ. Georg Norbert Christoph Dominik Graf von Stillfried-Rattonitz Vollständiger Abdruck der von der Fakultät für Informatik der Technischen Uni- versität München zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaften (Dr. rer. nat.) genehmigten Dissertation. Vorsitzender: Univ.-Prof. Dr.-Ing. Alin Albu-Schäﬀer Prüfer der Dissertation: 1. Univ-Prof. Dr. Patrick van der Smagt 2. Univ-Prof. Dr.-Ing. Tamim Asfour, Karlsruher Institut für Technologie Die Dissertation wurde am 30.03.2015 bei der Technischen Universität München eingereicht und durch die Fakultät für Informatik am 27.07.2015 angenommen. ii Abstract Human hand movement models are needed for the design of humanoid robotic hands. Humanoid hands are robotic hands that resemble human hands, espe- cially regarding their appearance and their ability to move. Possible applica- tion areas of humanoid hands are expected in human-inhabited environments, in teleoperation and in prosthetics. The goal for this thesis is to ﬁnd out which kinematic properties of the human hand are important and should be implemented in a humanoid hand. For this, a model is desired that matches the mobility of the human hand as closely as possible. It should be able to replicate the movements that the human hand is able to do and avoid movements that the human hand is unable to do. Such a kinematic model of the human hand that can be used for simulations has been lacking.
[Show full text]
Kinematics of Robots
Kinematics of Robots Alba Perez Gracia c Draft date September 9, 2011 24 Chapter 2 Robot Kinematics Using Matrix Algebra 2.1 Overview In Chapter 1 we characterized the motion using vector and matrix algebra. In this chapter we introduce the application of 4 × 4 homogeneous transforms in traditional robot analysis, where the motion of a robot is described as a 4 × 4 homogoeneous matrix representing the position of the end-effector. This matrix is constructed as the composition of a series of local displacements along the chain. In this chapter we will learn how to calculate the forward and inverse kinematics of serial robots by using the Denavit-Hartenberg parameters to create the 4 × 4 homogeneous matrices. 2.2 Robot Kinematics From a kinematics point of view, a robot can be defined as a mechanical system that can be programmed to perform a number of tasks involving movement under automatic control. The main characteristic of a robot is its capability of movement in a six-dimensional space that includes translational and rotational coordinates. We model any mechanical system, including robots, as a series of rigid links connected by joints. The joints restrict the relative movement of adjacent links, and are generally powered and equipped with systems to sense this movement. 25 26 CHAPTER 2. ROBOT KINEMATICS USING MATRIX ALGEBRA The degrees of freedom of the robot, also called mobility, are defined as the number of independent parameters needed to specify the positions of all members of the system relative to a base frame. The most commonly used criteria to find the mobility of mechanisms is the Kutzbach-Gruebler formula.
[Show full text]
Robotics & Automation Lecture 23 Closed Kinematic Chain And
Robotics & Automation Lecture 23 Closed Kinematic Chain and Parallel Mechanisms John T. Wen November 10, 2008 JTW-RA23 RPI ECSE/CSCI 4480 Robotics I Examples of closed kinematic chains ² Constrained robot, e.g., robot in contact with environment (polishing, assembly, etc.). ² Multiple interacting robots, e.g., multi-finger grasp, multiple cooperative robots. ² Parallel mechanism, e.g., 4-bar linkage, slider-crank, Stewart-Gough Platform, Delta robot. qq 2 qq 1 qq 3 November 10, 2008Copyrighted by John T. Wen Page 1 JTW-RA23 RPI ECSE/CSCI 4480 Robotics I Kinematics Forward kinematics and inverse kinematics are the same as before: given joint positions, find the task frame, and vice versa – subject to the closed chain constraints. Usually: Serial mechanism has complicated geometry, inverse kinematics is more dif- ficult than the forward kinematics. Parallel mechanisms usually are based on imposing kinematic constraints to simple serial mechanisms. Therefore, inverse kinematics is usu- ally easy. The forward kinematics needs to take into account of constraints, and is more difficult to solve. November 10, 2008Copyrighted by John T. Wen Page 2 JTW-RA23 RPI ECSE/CSCI 4480 Robotics I Constrained mechanisms Consider a serial kinematic chain with n joints (q 2 Rn). Suppose the chain is constrained: f(q) = 0; f : Rn ! Rk: Then the mechanism has n ¡ k (unconstrained) DOF. Example: ² 3-DOF Planar arm with tip constrained to move along a line (2 dof) ² 4-bar linkage (1 dof) November 10, 2008Copyrighted by John T. Wen Page 3 JTW-RA23 RPI ECSE/CSCI 4480 Robotics I Greubler’s Formula Consider a planar mechanism.
[Show full text]
Animation, Motion Capture, & Inverse Kinematics Last Time? Today
Last Time? • Navier-Stokes Equations Animation, Motion Capture, • Conservation of & Inverse Kinematics Momentum & Mass • Incompressible Flow Today Keyframing • How do we animate? • Use spline curves to automate the in betweening – Keyframing – Good control – Less tedious than drawing every frame – Procedural Animation • Creating a good animation still requires considerable – Physically-Based Animation skill and talent – Motion Capture – Forward and Inverse Kinematics • Rigid Body Dynamics • Finite Element Method ACM © 1987 “Principles of traditional animation applied to 3D computer animation” Procedural Animation Physically-Based Animation • Describes the motion algorithmically, • Assign physical properties to objects as a function of small number of (masses, forces, inertial properties) parameters • Example: a clock with second, minute • Simulate physics by solving equations and hour hands • Realistic but difficult to control – express the clock motions in terms of a “seconds” variable – the clock is animated by varying the seconds parameter g • Example: A bouncing ball -kt – Abs(sin(ωt+θ0))*e “Interactive Manipulation of Rigid Body Simulations” SIGGRAPH 2000, Popović, Seitz, Erdmann, Popović & Witkin 1 Motion Capture Today • Optical markers, high-speed • How do we animate? cameras, triangulation – Keyframing → 3D position – Procedural Animation • Captures style, subtle nuances – Physically-Based Animation and realism at high-resolution – Motion Capture • You must observe someone do something – Forward and • Difficult (or impossible?) to edit mo-cap data Inverse Kinematics • Rigid Body Dynamics • Finite Element Method Articulated Models Skeleton Hierarchy • Articulated models: • Each bone transformation – rigid parts described relative – connected by joints to the parent in hips the hierarchy: • They can be animated by specifying the joint left-leg ... angles as functions of time.
[Show full text]
Robotic Motion Planning: Configuration Space
Robotic Motion Planning: Configuration Space Robotics Institute 16-735 http://www.cs.cmu.edu/~motionplanning Howie Choset http://www.cs.cmu.edu/~choset 16-735, Howie Choset with slides from G.D. Hager, Z. Dodds, and Dinesh Mocha What if the robot is not a point? The Scout should probably not be modeled as a point... β α Nor should robots with extended linkages that may contact obstacles... 16-735, Howie Choset with slides from G.D. Hager, Z. Dodds, and Dinesh Mocha What is the position of the robot? Expand obstacle(s) Reduce robot 16-735, Howie Choset with slides fromnot quite G.D. rightHager, ... Z. Dodds, and Dinesh Mocha Configuration Space • A key concept for motion planning is a configuration: – a complete specification of the position of every point in the system • A simple example: a robot that translates but does not rotate in the plane: – what is a sufficient representation of its configuration? • The space of all configurations is the configuration space or C- space. C-space formalism: Lozano-Perez ‘79 16-735, Howie Choset with slides from G.D. Hager, Z. Dodds, and Dinesh Mocha Robot Manipulators What are this arm’s forward kinematics? (How does its position (x,y) β depend on its joint angles?) L2 y L 1 α x 16-735, Howie Choset with slides from G.D. Hager, Z. Dodds, and Dinesh Mocha Robot Manipulators What are this arm’s forward kinematics? (x,y) β Find (x,y) in terms of α and β ... L2 y L 1 α x Keeping it “simple” cα = cos(α) , sα = sin(α) cβ = cos(β) , sβ = sin(β) c+= cos(α+β) , s+= sin(α+β) 16-735, Howie Choset with slides from G.D.
[Show full text]