Scribd
Upload
43 views21 pages

L6-TSP Sensing and Perception

TexPoint fonts used in EMF. Read the TexPoint manual before you delete this box. Axon Learning Lab is a free, web-accessible, searchable, yet another login. Each week the tutors will select the "best" student answer They will award a gold and silver star.

Uploaded by

Ivan Avramov
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
43 views21 pages

L6-TSP Sensing and Perception

TexPoint fonts used in EMF. Read the TexPoint manual before you delete this box. Axon Learning Lab is a free, web-accessible, searchable, yet another login. Each week the tutors will select the "best" student answer They will award a gold and silver star.

Uploaded by

Ivan Avramov
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 21

RoboticsCourseWare Contributor 2012 School of Information Technology and Electrical Engineering at the University of Queensland

TexPoint fonts used in EMF. Read the TexPoint manual before you delete this box.: AAAAA

Schedule
Week 1 2 3 4 5 Date 23-Jul 30-Jul 6-Aug 13-Aug 20-Aug Lecture (M: 12-1:30, 43-102) Introduction Representing Position & Orientation & State (Frames, Transformation Matrices & Affine Transformations) Robot Kinematics and Dynamics Robot Dynamics & Control Obstacle Avoidance & Motion Planning

6
7 8 9 10 11 12 13

27-Aug Sensors, Measurement Perception


3-Sep 10-Sep 17-Sep 24-Sep 1-Oct 8-Oct 15-Oct 22-Oct

and

Localization and Navigation (GPS, INS, & SLAM) State-space modelling & Controller Design Vision-based control Study break Public Holiday Robot Machine Learning Guest Lecture Wrap-up & Course Review

Quick Outline
1. Discussion Forum & Sign-up Updates 2. Path Planning Recap 3. Sample-based Path Planning & Sequencing 4. Perception
1. 2. 3. 4. Sensing Sensors (Laser, Vision) Calibration Feature extraction

Updates: Discussion Forum & Signup Sheet


Two Options:
Blackboard: Not-free, access controls already in place Shapado: Free, Web-accessible, searchable, yet another login

Proposal:
Each week the tutors will select the best student answer
They will award a gold and silver star. Gold star entitles a free 6 hour extension for the team Silver star entitles a 1 hour extension The tutors can issue multiple (or no) stars

Please try the forums before emailing metr4202@

Sign-up Sheet is available (after class) and will be posted outside the Axon Learning Lab

Geometric Planning Methods


Several Geometric Methods:
Goal

Start

Geometric Planning Methods


Several Geometric Methods:
Goal Vertical (Trapezoidal) Cell Decomposition

Start

Geometric Planning Methods


Several Geometric Methods:
Goal Vertical (Trapezoidal) Cell Decomposition

Start

Geometric Planning Methods


Several Geometric Methods:
Goal Vertical (Trapezoidal) Cell Decomposition Roadmap Methods

Start

Geometric Planning Methods


Several Geometric Methods:
Start Vertical (Trapezoidal) Cell Decomposition

Roadmap Methods Goal


Cell (Triangular) Decomposition Visibility Graphs Veroni Graphs

Sample-Based Motion Planning

PRMs

RRTs

Rapidly Exploring Random Trees (RRT)

r(m)

xrand

xgoal

xinit s(m) q(m)

Sampling and the Bug Trap Problem

q(m/s)

Multiple Points & Sequencing


Sequencing
Determining the best order to go in Start

Travelling Salesman Problem


Goal Goal Goal
A salesman has to visit each city on a given list exactly once. In doing this, he starts from his home city and in the end he has to return to his home city. It is plausible for him to select the order in which he visits the cities so that the total of the distances travelled in his tour is as small as possible.

Goal Goal

Travelling Salesman Problem


Given a matrix C=(cij)
Start Goal Goal
Note that this problem is NP-Hard

distance

Minimize:

Goal Goal

Goal

BUT, Special Cases are Well-Solvable!

Travelling Salesman Problem [2]


This problem is NP-Hard
For the Euclidean case (where the points are on the 2D Euclidean plane) : The shortest TSP tour does not intersect itself, and thus geometry makes the problem somewhat easier. If all cities lie on the boundary of a convex polygon, the optimal tour is a cyclic walk along the boundary of the polygon (in clockwise or counterclockwise direction). The k-line TSP The a special case where the cities lie on k parallel (or almost parallel) lines in the Euclidean plane. EG: Fabrication of printed circuit boards Solvable in O(n3) time by Dynamic Programming (Rote's algorithm) The necklace TSP The special Euclidean TSP case where there exist n circles around the n cities such that every cycle intersects exactly two adjacent circles

BUT, Special Cases are Well-Solvable!

Sensing: The BIG Picture


Recall from First Year:

Sense

Control

Act

Sensing: quantifying a systems state Sensor: device that does sensing


Receives and responds to a signal/stimulus

What is an electric transducer?


Convert from the measurement space to an electrical signal Why? : Measuring voltage is easy!
Transducer Electrical Signal

ENERGY

Thermal Electrical Fluid Mechanical Chemical Rotary Linear

Voltage Resistance Inductance

Current

Capacitance

Example Sensor: GP2D Range Sensor


(distance to target) Sharp GP2D12 (voltage)

Spectrum

Noise

Source: Prof. M. Siegel, CMU

Note: this picture illustrates the concepts but it is not quantitatively precise

10

Inertial: Translation Accelerometer


General accelerometer:

Inertial: Rotation Gyroscopes


Structural arrangement of silicon which records centrifugal acceleration and thus angular speed Use strain-gauge bridges and/or piezo structure to record deformations

11

Accelerometer Acceleration

signal
(Accelerometer)

rotation

acceleration gravity

tangential

centripetal

Noise

Noise adds uncertainty Gravitational & inertial forces are inseparable


(Gyro)

Position and Velocity Sensors


Position and velocity measurement is often required in feedback loops for positioning, and velocity control Position measurement
Potentiometers Linear Variable Differential Transformer (LVDT) Encoders

Velocity Measurement
Tachometer

Sources: ADI

signal

12

Position measurement: Potentiometers


An analog sensor Works as a voltage divider

Vin

R Vout

Where have you seen one of these?

Example Applications
Joystick Electric car control

13

Position/Velocity: Optical Encoders


Encoders are digital Sensors commonly used to provide position feedback for actuators Consist of a glass or plastic disc that rotates between a light source (LED) and a pair of photo-detectors Disk is encoded with alternate light and dark sectors so pulses are produced as disk rotates

Incremental Encoders
Pulses from leds are counted to provide rotary position Two detectors are used to determine direction (quadrature) Index pulse used to denote start point Otherwise pulses are not unique

14

Absolute Encoders
Absolute encoders have a unique code that can be detected for every angular position Often in the form of a grey code; a binary code of minimal change Absolute encoders are much more complex and expensive than incremental encoders

Laser measurement system


The LMS system operates by measuring the time of flight of laser light pulses:
A pulsed laser beam is emitted and reflected if it meets an object. The reflection is registered by the scanners receiver. The time between transmission and reception of the impulse is proportional to the distance between the scanner and the object

15

LMS & Coal Loading Automation


2D Scanning laser sensor Automation of the Coal Loading process

-100

Laser HD

Infrastructure

-200

-300

Front wall ore /coal car

Back wall

-400

-500

-600

200

400

600

800

1000

1200

1400

1600

1800

Industrial Sensors: Force


Proximity
Mechanical Optical Inductive/Capacitive

Position/Velocity
Potentiometer LVDT Encoders Tachogenerator

Force/Pressure

16

Sensor Information

Laser

GPS Vision

Mapping: Indoor robots

17

Cameras

Image Formation

18

Stereopsis

Perception
Making Sense from Sensors

http://www.michaelbach.de/ot/mot_rotsnake/index.html

19

Features -- Colour Features

Bayer Patterns

Edge Detection
Canny edge detector:

20

Edge Detection
Canny edge detector:

Cool Robotics Share

21

You might also like