Advanced Modeling & Simulation Techniques for Multibody Robotic Systems

Advanced Modeling & Simulation Techniques for Multibody Robotic Systems

This webinar will be available afterwards at

designworldonline.com & email

Q&A at the end of the presentation

Hashtag for this webinar: #DWwebinar

Before We Start

http://www.designworldonline.com/

Moderator

Laura Carrabine Design World

Presenters

Dr. Amir Khajepour University of Waterloo

Paul Goossens Maplesoft

© 2012 Maplesoft, a division of Waterloo Maple Inc.

Paul Goossens, VP of Applications Engineering, Maplesoft Dr. Amir Khajepour, President, AEMK Systems, and Professor, Mechanical Engineering

© 2012 Maplesoft, a division of Waterloo Maple Inc. A CYBERNET group company

• Introduction – Challenges in Model-based design and development

• Case Studies:

Space Applications oPlanetary Rovers

Automotive Applications oPulsed Active Steering

Robotics Applications oCable-based Parallel Robot

• Summary - Maplesoft Engineering Solutions

• Q&A


“Virtual” Prototyping through Model-based Design and Development plays an increasingly key role in system design, commissioning and testing.

•Increasing adoption of MBD and simulation

• Reduce prototyping cycles and costs

• Increase end-user functionality, quality, safety and reliability

• Deterministic, repeatable testing platform

Connection to real components with virtual subsystems through Hardware-in-the-Loop (HIL) Testing is critical to this strategy

• Validation of subcomponents and/or controllers before integrating into the vehicle reduces errors and costs

• Validation of the model against the real thing improves the whole process, dramatically reducing development cycles and time-to-market in the future

Greater demand for greater model fidelity…


Scalability

Task

s Capacity

Number of functions (Complexity)


Scalability

Multi-domain Modeling

Engine/ Powertrain

Torque/Speed Inputs

Chassis/Tire Torque/Speed

Outputs

Apply Load??? Driveline

Task

s Capacity



Scalability

Multi-domain Modeling

Real-time Performance

Engine/ Powertrain

Torque/Speed Inputs

Chassis/Tire Torque/Speed

Outputs

Apply Load??? Driveline

Task

s Capacity



Space Applications o Planetary Rovers

Automotive Applications o Pulsed Active Steering

Robotics Applications o Cable-based Parallel Robot


Wheels

Solar cells

Wheel motors

Battery

Power electronics

Heaters

Robotic arms, other peripherals

System Components

Terrain

Environment

Analysis

Rover simulation and animation


Six-wheeled Rocker-Bogie Rover

Angular velocity input

Stee

rin

g an

gle

inp

ut

Modeling Environment

Visualization Environment


Dynamic Model Component Library

Component Library in MapleSim

Planetary Rovers: Dynamic Modeling in MapleSim


Custom Components



Movie No. 1



Irradiation on Mars - MapleSim Model

Planetary Rovers: Component Modeling and HIL


Solar Cells Model in MapleSim



LabVIEW™ Model for Hardware/Software MapleSim Connector for LabVIEW™



LabVIEW™ Model for Hardware/Software



Input Software

Path Planning Optimization

Rover Kinematics

Hardware (Test Bench)

Halogen Lamps

Solar Panels

Battery

Motor

Charge Controller

Inverter

Load Simulator

CVT

Flywheel

Power Consumption (Driving)

Vehicle Speed

Vehicle - position - orientation - tilt

Component Modeling

Solar Panels

Battery

Motor

Irradiation Model NI® PXI

Light Intensity

LabVIEW™ 2009

HIL Graphical User Interface

Temperature

Voltage

Current

Angular Position

Measurements & Data Logging

Angular Velocity

Hardware in the Loop Overview



HIL Graphical User Interface



δcActuator

N δ

δ

Controller

Vehicle

Dynamic

Sensors

Control System that adds/subtracts a steering angle to the drivers steering input

Has Two Effects: 1. Rollover Prevention 2. Lower Path Following Deviation

What is Active Steering?

Pulsed Active Steering: Introduction


Basic Vehicle Model with input/output signals for simulation

Vehicle Model in MapleSim

Pulsed Active Steering: HIL Experiment


Real-time Simulink® Program

*MATLAB and Simulink are registered trademarks of The Mathworks, Inc. All other trademarks are the property of their respective owners.





Graphic User Interface



Movie No. 2



• Parallel Robot

– Low Inertia: Actuating motors are located at base and motor inertia is not part of the system

– High Speed: Less inertia means less required torque, and more power available for speed

• Cable Based Robot

– Use cables under tension in place of solid links

– Apply spine force on end-effector to keep cables under tension

– Less inertia compared to solid links

– Achieve even higher speed

Cable Based Parallel Robots: DeltaBot™ Cable Robot


Cable Based Parallel Robots: DeltaBot™ Cable Robots, 2 and 3 translational and 1 rotational DOF


MapleSim Model



Construction Steps

• Define ground points

Cable Based Parallel Robots: DeltaBot™ Cable Robot (2-Axis)


Construction Steps

• Create model of arm using Multibody library components

• Rigid body center of mass

• Rigid body frame (links)

• Visualization component



Construction Steps

• Define parameters for the arm

• Define default values of parameters

• Parameters are unique to each instance of shared subcomponent



Construction Steps

• Attach arms to grounds using revolute joints

• Define initial conditions for revolute joints



Construction Steps

• Construct model of triangle assembly

• Convert it to subcomponent and connect it to the main model



Construction Steps

• Create model of cable

• Use a custom spring with a slider



Cable Model using Custom Spring

• Very high spring constant under tension

• No spring constant under compression

• Cannot stretch but able to collapse

0.00E+00

1.00E+05

2.00E+05

3.00E+05

4.00E+05

5.00E+05

6.00E+05

-8 -6 -4 -2 0 2 4 6

Spri

ng

Co

nst

ant

K (

N/m

)

Displacement (m)

spring constant vs. displacement

Compression Tension



Construction Steps

• Create subcomponent consisting of pair of cables and spherical joints

• Connect cables to the main model



Construction Steps

• Create model of end-effector subcomponent



Construction Steps

• Add cylinder to the main model



Final Model for Simulation



Sample Generated Plots

• Red: Cable Tension

• Blue: Arm Torque

• Tension is becoming negative in this particular motion

• Try increasing spine force



Sample Generated Plots (Revised Simulation)

• Red: Cable Tension

• Blue: Arm Torque

• Increased spine force

• Tension is now always positive for this particular motion

• Drawback: Increased torque requirement



Movie No. 3



MapleSim is a truly unique physical

modeling tool:

• Built on a foundation of symbolic

computation technology

• Handles all of the complex mathematics

involved in the development of engineering

models

• Multi-domain systems, plant modeling,

control design

• Leverages the power of Maple to take

advantage of extensive analytical tools

• Reduces model development time from

months to days while producing high-

fidelity, high-performance models


Driveline Component Library More Libraries

-dSPACE® -LabVIEW™ -NI® VeriStand™ -MATLAB® & Simulink® -B&R Automation Studio

*Simulink and MATLAB are registered trademark of The Mathworks, Inc. All other trademarks are property of their respective owners.


Automatic

Equation

Generation Symbolic model simplification

Optimized code generation

Best performance

~10x faster than similar tools

Advanced analysis

Parameter optimization

Sensitivity etc

Multibody kinematics and dynamics

Equation-based Model Creation

Enter system equations

Test/Validate model

Easy component block generation

Greater insight into

system behavior


Plant model Analysis Controller design

Equation and code generation

Controller implementation (and design) Real-time management

Embedded controller Data acquisition

System HIL Simulation

*Simulink and MATLAB are registered trademark of The Mathworks, Inc. All other trademarks are property of their respective owners.

-dSPACE® -LabVIEW™ -NI® VeriStand™ -MATLAB® & Simulink® -B&R Automation Studio


• Physical modeling: increasingly important – and increasingly complex – in systems design, testing and integration.

• Symbolic technology: proven engineering technology that significantly improves model fidelity without sacrificing real-time performance.

• MapleSim: ideal tool for rapid development of high-fidelity physical models of mechatronics systems to help engineers achieve their design goals.


Thank You!

Questions?

To stay connected: www.maplesoft.com/subscribe

Questions?

Design World Laura Carrabine [email protected] Phone: 440.234.4531 Twitter: @wtwh_laurac

University of Waterloo Dr. Amir Khajepour [email protected]

Maplesoft Paul Goossens [email protected] www.maplesoft.com/subscribe

mailto:[email protected]



Thank You

This webinar will be available at designworldonline.com & email

Tweet with hashtag #DWwebinar

Connect with

Twitter: @DesignWorld

Facebook: facebook.com/engineeringexchange

LinkedIn: Design World Group

YouTube: youtube.com/designworldvideo

Discuss this on EngineeringExchange.com

http://www.designworldonline.com/

http://www.engineeringexchange.com/

Technology

Advanced Modeling & Simulation Techniques for Multibody Robotic Systems