Simple set-membership methods and control algorithms applied to robots for exploration Fabrice LE BARS.

Slides:



Advertisements
Similar presentations
Bayesian Belief Propagation
Advertisements

Robotics at ENSTA Bretagne Fabrice LE BARS. Robotics at ENSTA Bretagne 17/05/ Robotics at ENSTA Bretagne Main research areas : Autonomous marine.
Mobile Robot Localization and Mapping using the Kalman Filter
Reactive and Potential Field Planners
Probabilistic Inverse Dynamics for Nonlinear Blood Pattern Reconstruction Benjamin Cecchetto, Wolfgang Heidrich University of British Columbia.
StatisticalDesign&ModelsValidation. Introduction.
Welcome to PHYS 225a Lab Introduction, class rules, error analysis Julia Velkovska.
Yasuhiro Fujiwara (NTT Cyber Space Labs)
SLAM with fleeting detections Fabrice LE BARS. SLAM with fleeting detections 30/04/ Outline  Introduction  Problem formalization  CSP on tubes.
Sonar and Localization LMICSE Workshop June , 2005 Alma College.
Pathfinding Basic Methods.
Zonotopes Techniques for Reachability Analysis Antoine Girard Workshop “Topics in Computation and Control” March 27 th 2006, Santa Barbara, CA, USA
Dynamic Bayesian Networks (DBNs)
CIS 540 Principles of Embedded Computation Spring Instructor: Rajeev Alur
DARPA Mobile Autonomous Robot SoftwareMay Adaptive Intelligent Mobile Robotics William D. Smart, Presenter Leslie Pack Kaelbling, PI Artificial.
Robot Localization Using Bayesian Methods
Using Probabilistic Methods for Localization in Wireless Networks Presented by Adam Kariv May, 2005.
Uncertainty Representation. Gaussian Distribution variance Standard deviation.
Design of Attitude and Path Tracking Controllers for Quad-Rotor Robots using Reinforcement Learning Sérgio Ronaldo Barros dos Santos Cairo Lúcio Nascimento.
Path Planning with the humanoid robot iCub Semester Project 2008 Pantelis Zotos Supervisor: Sarah Degallier Biologically Inspired Robotics Group (BIRG)
Zachary Wilson Computer Science Department University of Nebraska, Omaha Advisor: Dr. Raj Dasgupta.
Active SLAM in Structured Environments Cindy Leung, Shoudong Huang and Gamini Dissanayake Presented by: Arvind Pereira for the CS-599 – Sequential Decision.
HMM-BASED PATTERN DETECTION. Outline  Markov Process  Hidden Markov Models Elements Basic Problems Evaluation Optimization Training Implementation 2-D.
1 Robust Video Stabilization Based on Particle Filter Tracking of Projected Camera Motion (IEEE 2009) Junlan Yang University of Illinois,Chicago.
Reliable Range based Localization and SLAM Joseph Djugash Masters Student Presenting work done by: Sanjiv Singh, George Kantor, Peter Corke and Derek Kurth.
ENSIETA Autonomous vehicle SAUC’ISSE. 09/07/ > Content 1.Introduction 2.Technical design 1.Mechanical description 2.Embedded electronic architecture.
Design of Autonomous Navigation Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow March 19, 2004.
Efficient Simulation of Physical System Models Using Inlined Implicit Runge-Kutta Algorithms Vicha Treeaporn Department of Electrical & Computer Engineering.
Particle Filtering for Non- Linear/Non-Gaussian System Bohyung Han
Geometric Probing with Light Beacons on Multiple Mobile Robots Sarah Bergbreiter CS287 Project Presentation May 1, 2002.
Real-time Video Streaming from Mobile Underwater Sensors 1 Seongwon Han (UCLA) Roy Chen (UCLA) Youngtae Noh (Cisco Systems Inc.) Mario Gerla (UCLA)
CISE-301: Numerical Methods Topic 1: Introduction to Numerical Methods and Taylor Series Lectures 1-4: KFUPM.
כמה מהתעשייה? מבנה הקורס השתנה Computer vision.
Speed and Direction Prediction- based localization for Mobile Wireless Sensor Networks Imane BENKHELIFA and Samira MOUSSAOUI Computer Science Department.
Muhammad Moeen YaqoobPage 1 Moment-Matching Trackers for Difficult Targets Muhammad Moeen Yaqoob Supervisor: Professor Richard Vinter.
Regression Analysis Regression analysis is a statistical technique that is very useful for exploring the relationships between two or more variables (one.
July 11, 2001Daniel Whiteson Support Vector Machines: Get more Higgs out of your data Daniel Whiteson UC Berkeley.
Efficient Model Selection for Support Vector Machines
1 Constant Following Distance Simulations CS547 Final Project December 6, 1999 Jeremy Elson.
Behavior Based Robotics: A Wall Following Behavior Arun Mahendra - Dept. of Math, Physics & Engineering, Tarleton State University Mentor: Dr. Mircea Agapie.
Joint International Master Project Dennis Böck & Dirk C. Aumueller 1.
Target Tracking with Binary Proximity Sensors: Fundamental Limits, Minimal Descriptions, and Algorithms N. Shrivastava, R. Mudumbai, U. Madhow, and S.
CISE-301: Numerical Methods Topic 1: Introduction to Numerical Methods and Taylor Series Lectures 1-4: KFUPM CISE301_Topic1.
CISE301_Topic11 CISE-301: Numerical Methods Topic 1: Introduction to Numerical Methods and Taylor Series Lectures 1-4:
Simultaneous Localization and Mapping Presented by Lihan He Apr. 21, 2006.
Particle Filters for Shape Correspondence Presenter: Jingting Zeng.
Name : Arum Tri Iswari Purwanti NPM :
Forward-Scan Sonar Tomographic Reconstruction PHD Filter Multiple Target Tracking Bayesian Multiple Target Tracking in Forward Scan Sonar.
Wandering Standpoint Algorithm. Wandering Standpoint Algorithm for local path planning Description: –Local path planning algorithm. Required: –Local distance.
Low Level Control. Control System Components The main components of a control system are The plant, or the process that is being controlled The controller,
State estimation with fleeting data. 05/12/20152 > Plan 1.Introduction 2.Approach used: fleeting data and tubes 3.Demo 4.Conclusion.
Control of an autonomous sailboat : application to the VAIMOS robot Fabrice LE BARS.
Laboratory of mechatronics and robotics Institute of solid mechanics, mechatronics and biomechanics, BUT & Institute of Thermomechanics, CAS Mechatronics,
Monte-Carlo based Expertise A powerful Tool for System Evaluation & Optimization  Introduction  Features  System Performance.
1/17/2016CS225B Kurt Konolige Probabilistic Models of Sensing and Movement Move to probability models of sensing and movement Project 2 is about complex.
Computer Science 1 Using Clustering Information for Sensor Network Localization Haowen Chan, Mark Luk, and Adrian Perrig Carnegie Mellon University
Randomized Kinodynamics Planning Steven M. LaVelle and James J
Regression Analysis1. 2 INTRODUCTION TO EMPIRICAL MODELS LEAST SQUARES ESTIMATION OF THE PARAMETERS PROPERTIES OF THE LEAST SQUARES ESTIMATORS AND ESTIMATION.
Paging Area Optimization Based on Interval Estimation in Wireless Personal Communication Networks By Z. Lei, C. U. Saraydar and N. B. Mandayam.
A track fitting method for multiple scattering Peter Kvasnička, 5th SILC meeting, Prague 2007.
3D Perception and Environment Map Generation for Humanoid Robot Navigation A DISCUSSION OF: -BY ANGELA FILLEY.
Introduction to Machine Learning, its potential usage in network area,
Introduction to Quantitative Analysis
Introduction, class rules, error analysis Julia Velkovska
Machine Learning for dotNET Developer Bahrudin Hrnjica, MVP
CS 416 Artificial Intelligence
SENSOR BASED CONTROL OF AUTONOMOUS ROBOTS
Reinforcement Learning (2)
CISE-301: Numerical Methods Topic 1: Introduction to Numerical Methods and Taylor Series Lectures 1-4: KFUPM CISE301_Topic1.
Reinforcement Learning (2)
Presentation transcript:

Simple set-membership methods and control algorithms applied to robots for exploration Fabrice LE BARS

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ Outline  Introduction  A simple localization example  Interval analysis  Other localization scenarios  Line following  Additional common problems and possible methods  Conclusion

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ Introduction

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ Robotics at ENSTA Bretagne  Main research areas : Autonomous marine and submarine robotics using interval methods Swarm of robots

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/  Facts Autonomous robotics rise new and difficult problems There are few demonstrations of cheap autonomous robots able to do survey, cartography, localization tasks, especially in marine and submarine environments Current methods : mainly probabilistics Robotics at ENSTA Bretagne

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/  Goals Develop observation, control methods for submarine, marine, ground and aerial robotics Demonstrate the use of interval methods through new applications in autonomous robotics Build real and convincing demonstrators Copyright Ifremer Robotics at ENSTA Bretagne

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ A simple localization example

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ A simple localization example  Context : An underwater robot must follow a wall (not necessarily straight) The environment (walls to follow) is known The robot has a good compass and a sonar

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ A simple localization example  Assumptions : The wall to follow is on the right of the robot We can approximate it as a line y=ax+b

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ A simple localization example  Assumptions : The wall to follow is on the right of the robot We can approximate it as a line y=ax+b

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ A simple localization example  Assumptions : The wall to follow is on the right of the robot We can approximate it as a line y=ax+b

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ A simple localization example  Assumptions : The wall to follow is on the right of the robot We can approximate it as a line y=ax+b

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ A simple localization example  Assumptions : The wall to follow is on the right of the robot We can approximate it as a line y=ax+b

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ A simple localization example  The robot needs to find : Its distance to the line Its angle to the line (-> angle of the line)

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ A simple localization example  Available data : The angle of the robot thanks to the compass Sonar data

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ A simple localization example  Methods : Linear regression using least squares What if we want to localize in a pool, a lake…?

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/  Uncertainty representations : Probabilistics Gaussians Particles => We try to get a probability density Sets Zonotopes Ellipsoids Intervals => We try to enclose all possible solutions A simple localization example

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ Interval analysis

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ ,, are examples of intervals  Operations smallest interval containing the set of possible values for  Multiplication by a number, intersection, union  Image by a function Interval analysis

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ Interval analysis  Real intervals can be generalized Vectors intervals (boxes) Sets intervals Functions intervals (tubes) Any set with a lattice structure

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/  Contraction If and,,, then Interval analysis

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/  Contraction and propagation We call contractor an operator that reduce the domain of variables A propagation is a repeated call to contractors We can repeat contractions until a fix point Interval analysis

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/  Other techniques such as bisections can also be done Interval analysis

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/  Handling outliers : relaxed intersection (q-intersection) Interval analysis

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ Other localization scenarios

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ State Measurements For each k Other localization scenarios  Dynamic localization

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/  EURATHLON 2013 Task "Search and rescue in a smoke-filled underground structure“ Other localization scenarios Controller Robot Observer (implemented as an interval simulator)

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/  Localization of a swarm of robots with acoustic communication Other localization scenarios

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ Line following

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/  Purpose Cover autonomously an area as accurately as possible Line following

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/  From waypoints following to line following Primitive heading control loop Existing approaches : basic waypoint following The robot follows a heading in direction of its waypoint Waypoint reached when in a predefined radius Problem : nothing prevent the drift between waypoints (because of currents...) => Line following (use the distance to the line) Line following

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ Line following Primitive heading regulator Robot

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ Init GetSensorsData NavigationManager Supervisor PrimitiveHeadingController SetActuators Line following

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/  Line following Primitive controller stage for heading control Rudder control Navigation manager sends lines to supervisor and validates lines Validation condition Line following

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/  Line following Desired heading is the line made by the 2 current waypoints with an attractiveness angle to the line depending on the distance to the line Line following

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ Additional common problems and possible methods

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ Additional common problems and possible methods  Evaluate the guaranteed covered area depending on localization accuracy  Obstacle avoidance using vector fields

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ Conclusion

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ Conclusion Interval methods can be efficient for parameters and state estimation problems Advantages : can give an estimation of the error together with the state, parameters Can also be used when there are outliers Line following can improve easily the accuracy of a trajectory following especially in marine environment Vector fields are interesting for obstacle avoidance or other higher level trajectory planning

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/ Questions?  Contact

Simple set-membership methods and control algorithms applied to robots for exploration 12/08/