Production and Analytics of a Multi-Linked Robotic System

Author(s):  
Thorstein R. Rykkje ◽  
Eystein Gulbrandsen ◽  
Andreas Fosså Hettervik ◽  
Morten Kvalvik ◽  
Daniel Gangstad ◽  
...  

Abstract This paper extends research into flexible robotics through a collaborative, interdisciplinary senior design project. This paper deploys the Moving Frame Method (MFM) to analyze the motion of a relatively high multi-link system, driven by internal servo engines. The MFM describes the dynamics of the system and enables the construction of a general algorithm for the equations of motion. Lie group theory and Cartan’s moving frames are the foundation of this new approach to engineering dynamics. This, together with a restriction on the variation of the angular velocity used in Hamilton’s principle, enables an effective way of extracting the equations of motion. The result is a dynamic 3D analytical model for the motion of a snake-like robotic system, that can take the physical sizes of the system and return the dynamic behavior. Furthermore, this project builds a snake-like robot driven by internal servo engines. The multi-linked robot will have a servo in each joint, enabling a three-dimensional movement. Finally, a test is performed to compare if the theory and the measurable real-time results match.

Author(s):  
Torstein R. Storaas ◽  
Kasper Virkesdal ◽  
Gitle S. Brekke ◽  
Thorstein Rykkje ◽  
Thomas Impelluso

Abstract Norwegian industries are constantly assessing new technologies and methods for more efficient and safer maintenance in the aqua cultural, renewable energy, and oil and gas industries. These Norwegian offshore industries share a common challenge: to install new equipment and transport personnel in a safe and controllable way between ships, farms and platforms. This paper deploys the Moving Frame Method (MFM) to analyze ship stability moderated by a dual gyroscopic inertial device. The MFM describes the dynamics of the system using modern mathematics. Lie group theory and Cartan’s moving frames are the foundation of this new approach to engineering dynamics. This, together with a restriction on the variation of the angular velocity used in Hamilton’s principle, enables an effective way of extracting the equations of motion. This project extends previous work. It accounts for the dual effect of two inertial disk devices, it accounts for the prescribed spin of the disks. It separates out the prescribed variables. This work displays the results in 3D on cell phones. It represents a prelude to testing in a wave tank.


Author(s):  
Thorstein R. Rykkje ◽  
Tord Tørressen ◽  
Håvard Løkkebø

Abstract This project creates a model to assess the motion induced on a buoy at sea, under wave conditions. We use the Moving Frame Method (MFM) to conduct the analysis. The MFM draws upon concepts and mathematics from Lie group theory — SO(3) and SE(3) — and Cartan’s notion of Moving Frames. This, together with a compact notation from geometrical physics, makes it possible to extract the equations of motion, expeditiously. This work accounts for the masses and geometry of all components and for buoyancy forces and added mass. The resulting movement will be displayed on 3D web pages using WebGL. Finally, the theoretical results will be compared with experimental data obtained from a previous project done in the wave tank at HVL.


2019 ◽  
Vol 141 (5) ◽  
Author(s):  
Paulo Alexander Jacobsen Jardim ◽  
Jan Tore Rein ◽  
Øystein Haveland ◽  
Thorstein R. Rykkje ◽  
Thomas J. Impelluso

A decline in oil-related revenues challenges Norway to focus on new types of offshore installations. Often, ship-mounted crane systems transfer cargo or crew onto offshore installations such as floating windmills. This project analyzes the motion of a ship induced by an onboard crane in operation using a new theoretical approach to dynamics: the moving frame method (MFM). The MFM draws upon Lie group theory and Cartan's moving frames. This, together with a compact notation from geometrical physics, makes it possible to extract the equations of motion, expeditiously. While others have applied aspects of these mathematical tools, the notation presented here brings these methods together; it is accessible, programmable, and simple. In the MFM, the notation for multibody dynamics and single body dynamics is the same; for two-dimensional (2D) and three-dimensional (3D), the same. Most importantly, this paper presents a restricted variation of the angular velocity to use in Hamilton's principle. This work accounts for the masses and geometry of all components, interactive motor couples and prepares for buoyancy forces and added mass. This research solves the equations numerically using a relatively simple numerical integration scheme. Then, the Cayley–Hamilton theorem and Rodriguez's formula reconstruct the rotation matrix for the ship. Furthermore, this work displays the rotating ship in 3D, viewable on mobile devices. This paper presents the results qualitatively as a 3D simulation. This research demonstrates that the MFM is suitable for the analysis of “smart ships,” as the next step in this work.


Author(s):  
Maren Eriksen Eia ◽  
Elise Mari Vigre ◽  
Thorstein Ravneberg Rykkje

Abstract A Knuckle Boom Crane is a pedestal-mounted, slew-bearing crane with a joint in the middle of the distal arm; i.e. boom. This distal boom articulates at the ‘knuckle (i.e.: joint)’ and that allows it to fold back like a finger. This is an ideal configuration for a crane on a ship where storage space is a premium. This project researches the motion and control of a ship mounted knuckle boom crane to minimize the pendulum motion of a hanging load. To do this, the project leverages the Moving Frame Method (MFM). The MFM draws upon Lie group theory — SO(3) and SE(3) — and Cartan’s Moving Frames. This, together with a compact notation from geometrical physics, makes it possible to extract the equations of motion, expeditiously. The work reported here accounts for the masses and geometry of all components, interactive motor couples and prepares for buoyancy forces and added mass on the ship. The equations of motion are solved numerically using a 4th order Runge Kutta (RK4), while solving for the rotation matrix for the ship using the Cayley-Hamilton theorem and Rodriguez’s formula for each timestep. This work displays the motion on 3D web pages, viewable on mobile devices.


Author(s):  
Josef Flatlandsmo ◽  
Torbjørn Smith ◽  
Ørjan O. Halvorsen ◽  
Johnny Vinje ◽  
Thomas J. Impelluso

Norwegian industries are constantly assessing new technologies and methods for more efficient and safer production in the aqua cultural, renewable energy, and oil and gas industries. These Norwegian offshore industries share a common challenge: to install new equipment and transport personnel in a safe and controllable way between ships, farms and platforms. This paper deploys the Moving Frame Method (MFM) to analyze the motion induced by a crane and controlled by a gyroscopic inertial device mounted on a ship. The crane is a simple two-link system that transfers produce and equipment to and from barges. An inertial flywheel — a gyroscope — is used to stabilize the barge during transfer. The MFM describes the dynamics of the system using modern mathematics. Lie group theory and Cartan’s moving frames are the foundation of this new approach to engineering dynamics. This, together with a restriction on the variation of the angular velocity used in Hamilton’s principle, enables an effective way of extracting the equations of motion. This project extends previous work. It accounts for the dual effect of both the crane and the stabilizing inertial device. Furthermore, this work allows for buoyancy and motor induced torques. Furthermore, this work displays the results in 3D on cell phones. The long-term results of this work leads to a robust 3D active compensation method for loading/unloading operations offshore. Finally, the interactivity between the crane and the stabilizing gyro anticipates the impending time of artificial intelligence when machines, equipped with on-board CPU’s and IP addresses, are empowered with learning modules to conduct their operations.


Author(s):  
Paulo Alexander Jacobsen Jardim ◽  
Jan Tore Rein ◽  
Øystein Haveland ◽  
Thomas J. Impelluso

A decline in oil-related revenues challenges Norway to focus on new types of offshore installations and their maintenance. Often, ship-mounted crane systems transfer cargo or crew onto marine structures such as floating windmills. This project analyzes the motion of a ship induced by an onboard crane in operation. It analyzes the motion of a crane mounted on a ship using The Moving Frame Method (MFM). The MFM draws upon Lie group theory and Cartan’s Moving Frames. This, together with a compact notation from geometrical physics, makes it possible to extract the equations of motion, expeditiously. This work extends a previous project that assumed many simplifications. It accounts for the masses and geometry of all components. This current approach also accounts interactive motor couples and prepares for buoyancy forces and added mass. The previous work used a symbolic manipulator, resulting in unwieldy equations. In this current phase, this research solves the equations numerically using a relatively simple numerical integration scheme. Then, the Cayley-Hamilton theorem and Rodriguez’s formula reconstructs the rotation matrix for the ship. Furthermore, this work displays the rotating ship in 3D, viewable on mobile devices. WebGL is a JavaScript API for rendering interactive 3D and 2D graphics within any compatible web browser without the use of plug-ins. This paper presents the results qualitatively as a 3D simulation. This research proves that the MFM is suitable for the analysis of “smart ships,” as the next step in this work.


Author(s):  
H. Murakami

Élie Cartan’s moving frame method, developed in differential geometry, has been applied to multi-body dynamics to derive equations of motion. The explicit representation of a body-attached orthonormal coordinate basis and its origin, referred to as a moving frame, enables the usage of the special orthogonal group, SO(3), and the special Euclidean group, SE(3), to describe kinematics and kinetics of interconnected bodies by joints and force elements. The moving frame representation using Theodore Frankel’s compact notation is adopted to alleviate theoretical complexities of the Lie group theory to which SO(3) and SE(3) belong. For the variational formulation, the restricted variation of angular velocity is derived for the moving frame method. Starting from two connected rigid bodies, it will be demonstrated that the explicit representation of moving frames renders straight-forward symbolic computations of three-dimensional kinematics and kinetics. This simplicity eliminates errors in computing analytical expressions for kinematic and kinetic variables and streamlines the coding effort for numerical solution. For controller design, if the degrees-of-freedom is small, the moving frame method allows a straight-forward derivation of equations of motion in analytical form.


Author(s):  
Hidenori Murakami

To describe the configuration of a multi-body system, Cartesian coordinate systems are attached to all bodies comprising the system. Their connections through joints and force elements are efficiently expressed by using 4×4 matrices of the homogeneous transformation, presented by Denavit and Hartenberg in 1955. However, at this time, there is no systematic method to compute velocities and angular velocities using the matrices of such homogeneous transformations. In this paper, homogeneous transformation matrices are identified as a subset of a Lie group, called the special Euclidean group denoted by SE(3). This observation enables the usage of the Lie group theory in multibody kinematics. The effective use of the theory is built upon a platform of a moving frame method as presented in this paper. In this method, for each body-attached Cartesian coordinate system, the coordinate vector basis is written explicitly following Élie Cartan. This moving frame notation enables us to use the Lie algebra of SE(3), denoted by se(3), to compute velocities and angular velocities by minimizing the complexities of the Lie group theory. For kinetics, a variational method is established in se(3) by deriving a relationship between a virtual angular velocities and the corresponding virtual rotational displacements. This constrained variation of virtual angular velocities allows the derivation of the d’Alembert principle of virtual work from Hamilton’s principle for multibody systems. Utilizing this variational tool, we present a systematic computation of equations of motion from Hamilton’s principle. Finally, we reduce the spatial dynamics to planar dynamics and list the simplifications achieved in the two-dimensional problems using SE(2). Then, for a two-degree-of-freedom manipulator the analytical equations of motion are obtained to demonstrate the power of the moving frame method.


Author(s):  
Andreas Nordvik ◽  
Natalia Khan ◽  
Roberto Andrei Burcă ◽  
Thomas J. Impelluso

This paper presents a new method in dynamics — The Moving Frame Method (MFM) — and applies it to analyze the roll, yaw and pitch of a ship at sea, as induced by an onboard moving crane. The MFM, founded on Lie Group Theory, Cartan’s Moving Frames and a compact notation from geometrical physics, enables this expedited extraction of the equations of motion. Next, the method deploys the power of the special Euclidean Group SE(3) and a restricted variation to be used in Hamilton’s Principle, to extract the equations of motion. The mathematical model is then simplified to get a clearer picture of the parameters that impact the motion of the crane. The equations of interest are numerically solved by using fourth order Runge-Kutta method to obtain the specific data for the motion induced by the crane. Then, The Cayley-Hamilton theorem is used to reconstruct the rotation matrix. To supplement the paper, a webpage is coded with a model of the crane and ship, to graphically visualize the motion in 3D. It is imperative to note that while there are many approaches to dynamics, the MFM presents a consistent method, from 2D to 3D, and across sub-disciplines. The simplification is what has enabled undergraduate students to undertake this project.


Author(s):  
Hidenori Murakami

In order to develop an active large-deformation beam model for slender, flexible or soft robots, the d’Alembert principle of virtual work is derived for three-dimensional elastic solids from Hamilton’s principle. This derivation is accomplished by refining the definition of the Cauchy stress tensor as a vector-valued 2-form to exploit advanced geometrical operations available for differential forms. From the three-dimensional principle of virtual work, both the beam principle of virtual work and beam equations of motion with consistent boundary conditions are derived, adopting the kinematic assumption of rigid cross-sections of a deforming beam. In the derivation of the beam model, Élie Cartan’s moving frame method is utilized. The resulting large-deformation beam equations apply to both passive and active beams. The beam equations are validated with the previously reported results expressed in vector form. To transform passive beams to active beams, constitutive relations for internal actuation are presented in rate-form. Then, the resulting three-dimensional beam models are reduced to an active planar beam model. Finally, to illustrate the deformation due to internal actuation, an active Timoshenko-beam model is derived by linearizing the nonlinear planar equations. For an active, simply-supported Timoshenko-beam, the analytical solution is presented.


Sign in / Sign up

Export Citation Format

Share Document