STUDY ON THE PATH TRACKING AND POSITIONING METHOD OF WHEELED MOBILE ROBOT



Similar documents
Chapter 2 Solutions. 4. We find the average velocity from

Path Tracking for a Miniature Robot

OPTIMIZING RAILCAR MANAGEMENT

Optimizing Tension Control in Center-Driven Winders

CE801: Intelligent Systems and Robotics Lecture 3: Actuators and Localisation. Prof. Dr. Hani Hagras

Lecture 39: Intro to Differential Amplifiers. Context

prestack depth migration application with Kircohhff

Answer, Key Homework 3 David McIntyre 1

4 Impulse and Impact. Table of contents:

Signal Processing for Synthetic Aperture Radars

Basic Linear Algebra

Collision Prevention and Area Monitoring with the LMS Laser Measurement System

2. TEST PITCH REQUIREMENT

Dev eloping a General Postprocessor for Multi-Axis CNC Milling Centers

Vectors. Vector Multiplication

Agilent EEsof EDA.

Multi-Robot Tracking of a Moving Object Using Directional Sensors

Power Electronics. Prof. K. Gopakumar. Centre for Electronics Design and Technology. Indian Institute of Science, Bangalore.

Solutions to Exercises, Section 5.1

Vector Algebra. Addition: (A + B) + C = A + (B + C) (associative) Subtraction: A B = A + (-B)

3D Vision An enabling Technology for Advanced Driver Assistance and Autonomous Offroad Driving

13.4 THE CROSS PRODUCT

PY106 Class13. Permanent Magnets. Magnetic Fields and Forces on Moving Charges. Interactions between magnetic north and south poles.

KINEMATICS OF PARTICLES RELATIVE MOTION WITH RESPECT TO TRANSLATING AXES

77 GHz TRAFFIC RADAR SENSOR FOR SPEED ENFORCEMENT MEASURING SPEED AND DISTANCE

Spacetime Diagrams (1D in space)

Key Stage 3 Mathematics Programme of Study

Robot Sensors. Outline. The Robot Structure. Robots and Sensors. Henrik I Christensen

GANTRY ROBOTIC CELL FOR AUTOMATIC STORAGE AND RETREIVAL SYSTEM

Development of an Automotive Active Safety System Using a 24 GHz-band High Resolution Multi-Mode Radar

Midterm Exam 1 October 2, 2012

Vehicle-based Sensor Networks for Traffic Monitoring

MOBILE ROBOT TRACKING OF PRE-PLANNED PATHS. Department of Computer Science, York University, Heslington, York, Y010 5DD, UK

ON POINT. The School Improvement Process

ZEISS T-SCAN Automated / COMET Automated 3D Digitizing - Laserscanning / Fringe Projection Automated solutions for efficient 3D data capture

Synthetic Sensing: Proximity / Distance Sensors

ERserver. iseries. Backup, Recovery and Media Services (BRMS)

Probabilistic Modeling of Acceleration in Traffic Networks as a Function of Speed and Road Type

Alignment Laser System.

Closed form spread option valuation

ROBOTICS 01PEEQW. Basilio Bona DAUIN Politecnico di Torino

Robot Perception Continued

Grazing incidence wavefront sensing and verification of X-ray optics performance

On Motion of Robot End-Effector using the Curvature Theory of Timelike Ruled Surfaces with Timelike Directrix

SPINDLE ERROR MOVEMENTS MEASUREMENT ALGORITHM AND A NEW METHOD OF RESULTS ANALYSIS 1. INTRODUCTION

Autonomous Advertising Mobile Robot for Exhibitions, Developed at BMF

Adaptive Cruise Control of a Passenger Car Using Hybrid of Sliding Mode Control and Fuzzy Logic Control

Intelligent Submersible Manipulator-Robot, Design, Modeling, Simulation and Motion Optimization for Maritime Robotic Research

Improved Mecanum Wheel Design for Omni-directional Robots

UAV Pose Estimation using POSIT Algorithm

SOLID MECHANICS TUTORIAL MECHANISMS KINEMATICS - VELOCITY AND ACCELERATION DIAGRAMS

How To Fuse A Point Cloud With A Laser And Image Data From A Pointcloud

Automated Container Handling in Port Terminals

CNC Machine Control Unit

Polarization of Light

measurement of force and moment mobilities using a finite difference technique

Solar Tracking Application

1. SYSTEM OVERVIEW. 1) Basic Theory of ABS Function

Original Research Articles

The RTS-STILL Robotic Fork-Lift

Robotics. Lecture 3: Sensors. See course website for up to date information.

Sterling Store Inventory Management. Concepts Guide. Release 9.2

FRC WPI Robotics Library Overview

CALIBRATION OF A ROBUST 2 DOF PATH MONITORING TOOL FOR INDUSTRIAL ROBOTS AND MACHINE TOOLS BASED ON PARALLEL KINEMATICS

Prentice Hall Connected Mathematics 2, 7th Grade Units 2009

E190Q Lecture 5 Autonomous Robot Navigation

Dock Scheduling Guide

INSTRUCTOR WORKBOOK Quanser Robotics Package for Education for MATLAB /Simulink Users

Robot Task-Level Programming Language and Simulation

The Influence of EDM Parameters in Finishing Stage on Material Removal Rate of Hot Work Steel Using Artificial Neural Network

A System for Capturing High Resolution Images

Lecture 7 Force and Motion. Practice with Free-body Diagrams and Newton s Laws

The Math in Laser Light Math

11.1. Objectives. Component Form of a Vector. Component Form of a Vector. Component Form of a Vector. Vectors and the Geometry of Space

Unit 1 - Radian and Degree Measure Classwork

FURTHER VECTORS (MEI)

3D Vision An enabling Technology for Advanced Driver Assistance and Autonomous Offroad Driving

Force/position control of a robotic system for transcranial magnetic stimulation

E70 Rear-view Camera (RFK)

An Energy-Based Vehicle Tracking System using Principal Component Analysis and Unsupervised ART Network

Onboard electronics of UAVs

Target Tracking with Monitor and Backup Sensors in Wireless Sensor Networks

DINAMIC AND STATIC CENTRE OF PRESSURE MEASUREMENT ON THE FORCEPLATE. F. R. Soha, I. A. Szabó, M. Budai. Abstract

EDUMECH Mechatronic Instructional Systems. Ball on Beam System

Physics 202 Problems - Week 8 Worked Problems Chapter 25: 7, 23, 36, 62, 72

TRIMBLE ATS TOTAL STATION ADVANCED TRACKING SYSTEMS FOR HIGH-PRECISION CONSTRUCTION APPLICATIONS

Design of a Universal Robot End-effector for Straight-line Pick-up Motion

Sensor-Based Robotic Model for Vehicle Accident Avoidance

Problem 6.40 and 6.41 Kleppner and Kolenkow Notes by: Rishikesh Vaidya, Physics Group, BITS-Pilani

Curriculum Map by Block Geometry Mapping for Math Block Testing August 20 to August 24 Review concepts from previous grades.

Some Comments on the Derivative of a Vector with applications to angular momentum and curvature. E. L. Lady (October 18, 2000)

Visual Servoing Methodology for Selective Tree Pruning by Human-Robot Collaborative System

Static Environment Recognition Using Omni-camera from a Moving Vehicle

Transcription:

International Journal of Computer Science & Engineering Surey (IJCSES) Vol.6 No.3 June 05 STUDY ON THE PATH TRACKING AND POSITIONING METHOD OF WHEELED MOBILE ROBOT Cheng Jing and Cheng Wushan College of Mechanical Engineering Shanghai Uniersity of Engineering Science Shanghai 060 China ABSTRACT As a kind of wheeled moile root used in intelligent logistics system AGV is mainly used for automatic material transportation the precise positioning and path tracking is the assu- rance of accurate material transportation. In this article the laser coordinate positioning technology is used to realize accurate positioning for AGV a new method of target reference point selection is put forward and path tracking is implemented in comination with the kinematics model of single steering wheel AGV the ojectie function that AGV successfully reaches the destination accurately according to the preset trajectory is completed finally. The study is in trial stage and otains good operation effectieness. KEY WORDS Wheeled Moile Root Laser Positioning Path Tracking Kinematics Model.. INTRODUCTION Wheeled moile root of Logistics system is also known as automatic guided ehicle (AGV) mainly used for the transportation of all kinds of material in warehouse it is an important equipment in logistics system of modern manufacturing enterprise it proides an important guarantee for integration flexile and efficient operation of the system. The accurate positioning of AGV comined with its effectie track to the desired trajectory can greatly improe the accuracy of material transportation in the three-dimensional warehouse. At present according to the particularity of working enironment a lot of positioning methods are aailale; they are classified into the orientation of inertial and eacon according to the positioning ased on sensor information types []. Independence of inertial positioning is strong ut the positioning error has a cumulatie effect it should e periodically reised []; Laser positioning and isual positioning are two of the most commonly techniques in eacon positioning. Visual position including image processing is greatly influenced y light conditions such as shadow with poor real-time performance and high system cost. Laser positioning is of high precision and good flexiility and there is no need to deal with the ground which is conenient for path transformation and suitale for field enironment meanwhile it is conenient to modify the motion parameters and driing paths. Laser positioning is one of the most adanced guidance in the word [3]. The laser coordinate positioning technology is used to realize accurate positioning of AGV in this paper. We can get the angle deiation to e corrected in real-time during the path tracking process through the analysis of the kinematics model of single steering wheel AGV and DOI:0.5/ijcses.05.630

International Journal of Computer Science & Engineering Surey (IJCSES) Vol.6 No.3 June 05 constantly modify the primary drie angle deiation. At the same time on the choice of reference target a new method is adopted in this method target reference points moe coherently with the AGV making that the tracking accuracy would not reduce for AGV getting close to the reference point in the moement aoiding the redundancy of system caused y setting too much reference points. Which is different from general path tracking methods? We deelop a method of locating and tracking of wheeled moile root this method can meet the requirements of intelligent warehouse for AGV deliery location accuracy.. LASER COORDINATE POSITIONING TECHNOLOGY.. Classification of Laser Positioning Algorithms For AGV naigation module the application of laser positioning algorithms includes the triangulation method and the three-point positioning method. The triangulation method is known y calculating the position and orientation of AGV through the position of the reflecting plate and the angle etween the AGV and them. This method requires the measurement signal with ery high angular resolution and narrow eam it is generally used in the occasion with laser positioning. The disadantage of the three-point positioning method is that the cost of location deice is high since measuring asolute distance etween the receier to the reflectie eacon such as reflector column is necessary and positioning system must hae a synchronization time as a enchmark which is ery complex [4]... Initialization of Coordinate System For otaining AGV location data in the actual operation process initialize the coordinate system efore the car moement determine the coordinate information of each laser reflector and see them as the asolute coordinate system [5]. Since cylindrical reflector can e measured at any angle we design reflector into a cylinder and initialization process is as follows. Choose the origin position of asolute coordinate system and the position must ensure that the laser scanning head can scan more than 3 pieces of reflectie pillars and the AGV keeps still at the origin; Open laser scanning head y scanning the distance and angel etween reflectie pillar and the scan head the position information of each reflectie pillar in the asolute coordinate system can e determined [6] ; In the actual moement process of AGV The coordinate information is coordinate reference point for AGV pose..3. Laser Positioning Principle For realizing the laser guidance of AGV place a certain numer of reflectie pillars around fixed location in its running enironment the exact alues of their coordinates are known and stored on control computer of AGV onoard [7]. Install a laser emission and detection deice on the top of AGV when laser eam fires into reflectie pillars most of the energy is reflected and detected y photosensitie diode which is mounted near the emission source at the same time the angle made y reflected light and adancing direction of AGV is determined. The AGV in action continually receies reflected laser eam from 3 (or more) known positions and we can determine the exact location and orientation of AGV through the calculation of geometric on azimuth relatie to the direction of the AGV [8]. As shown in Figure AGV has measured reflected laser eam from 3 (or more) known reference points( R R R3 ) their coordinates ( x y ) ( x y ) and ( x y 3 3) are known in the coordinate system xoy.

International Journal of Computer Science & Engineering Surey (IJCSES) Vol.6 No.3 June 05 Figure. Relation of ehiclethree reference points and coordinate system When the angles θ θ θ and 3 etween the running direction of AGV and three reference points x and direction of AGV can e otained according to the are measured then the coordinate ( ) y following methods. Assume that α = θ θ β = θ θ3 circleo represents a circle which goes through three points ehiclev R and R and circle O 3 represents a circle which goes through three points ehiclev R andr 3.Then the coordinates O ( x y ) and O 3 ( x3 y3 ) of the centers are as follows. In formula () and formula () ( Q ) ( x y ) ( x + P y + Q ) = () ( x y ) ( x + P y + Q ) 3 3 ( P Q ) = () ( α ε ) L cos( α ε ) L sin = sinα sinα ( α + ε + ε ) L cos( α + ε ε ) L3 sin 3 + P =. sin β sin β The coordinates ( x y ) ( x y ) and ( ) x y are known then we can calculate alues of 3 3 L L 3 ε and ε. Among them L indicates the distance etween R and R L3 indicates the distance etween R and R 3 εrepresents the angle etween L and the x axis ε represents the angle etween L and L 3. Via the equation () and () circle O ( x ) and O ( x ) y 3 3 y3 can e represented: 3

International Journal of Computer Science & Engineering Surey (IJCSES) Vol.6 No.3 June 05 ( x x P ) + ( y y Q ) = P O + (3) : Q ( x x P ) + ( y y Q ) = P O + (4) : Q 3 By soling equation (3) and (4) the location of the car can e otained. ( x y ) [ x + k( Q Q ) y k( P P )] = (5) In formula (5) ( PQ P Q ) ( P P ) + ( Q Q ) k =. After get the position the direction of AGV can e otained y formula (6). z y = π + tan θ x x y z (6) 3. TRACKING GUIDANCE Efficient way to guide AGV can reduce the track deiation in the process of transportation improe the accuracy of the position of the loading and unloading of goods and reduce the time of adjusting the ody position at the loading and unloading point improing transport efficiency. This article takes the path tracking method to complete tracking guidance. The so-called tracking guidance is that the preset trajectory of AGV is diided into N reference points and AGV always run to track the reference point not far ahead. This reference point is always in front of the AGV and not far away ut AGV can neer catch up it. Through the arrangement of these series of reference points on the path the running track of AGV is guided and an control ojectie of AGV is the expected reference point not far away in front of it. Running direction of AGV can e adjusted periodically y tracking guidance method and reduce errors generated during operation [9]. Kinematics model of single steering wheel AGV is analyzed as follows. Single steering wheel AGV is a common structure form of AGV its gear train part includes steering wheel and fixed drien wheels and it also can e attached to multiple uniersal wheels. Among them the layout mode and the relatie position of the steering wheel and fixed wheels determine the kinematics model of AGV and uniersal support wheels hae no effect on kinematics model [0]. For the single wheel AGV a kinematics model is uilt its task is soling the angle of AGV from the known position to next reference point in real-time according to known conditions such as sensor return alues in order to realize the AGV path tracking. 4

International Journal of Computer Science & Engineering Surey (IJCSES) Vol.6 No.3 June 05 Figure. Layout schematic diagram of gear train of single steering wheel AGV Layout schematic diagram of gear train of single wheel AGV is shown in Figure wherein M is the rotation center of the steering wheel M is the geometric center of the drien wheel is wheelase of AGV d is the offset of steering wheel in the direction of the center line(we call its left direction plus). When the rotation angle of steering wheel is θ (We call counterclockwise plus) O is the intersection of the axis of front and rear wheels and the transient rotation center of AGV and R is the instantaneous radius of gyration. Suppose that the current linear elocity of walking AGV is and the angular elocity is ω. Since in practice the situation of θ is asolutely equal to zero does not existwe just need consider the situation ofθ 0 and follows can e otained. ω = ( sin θ ) ( θ > 0) ( sinθ ) ( θ < 0) + d R = tanθ d tanθ ( θ > 0) ( θ < 0) Posture of AGV is defined as ( x y z) (7) (8) Wherein x y are coordinates of the geometric center of drien wheel(the M in Figure ) in the two-dimensional orthogonal coordinate of the work area and Z is the attitude of AGV (we stipulate that X-axis in the negatie direction is 0 and the counterclockwise direction is positie). Suppose for a minimum period of time t the center of drien wheel of AGV moes from A to B and posture of A is x y z ) posture of B is ( x y z) x x = x. So R + R sin z + R sin( z + ω t) ( θ > 0) sin z R sin( z ω t) ( θ < 0) ( y y + R = y R cos z R cos( z + ω t) ( θ > 0) cos z + R cos( z ω t) ( θ < 0) (9) 5

International Journal of Computer Science & Engineering Surey (IJCSES) Vol.6 No.3 June 05 z z = z + ω t ω t ( θ > 0) ( θ < 0) Plug formula (7) and formula (8) into (9)then θ = (0). x x d (sin z sin z ) Form formula (0) We can know that θ is independent of the speed and d are the structural parameters of AGV the current posture ( x y z) is measured y laser positioning technology posture of reference point ( x y z) is the known posture in the process of path tracking so they are all known thenθ can e otained easily. After a period of program running ( t ) the running direction of AGV turns the angleθ after adjustments AGV moes toward the direction of target point with small errors achieing tracking guidance. 4. EXPERIMENTAL VERIFICATION During the experiment the surroundings are knowntest positioning accuracy of AGV which uses laser positioning technology during the path tracing process. Due to large ody many parts and components(all parts will produce iration while working) and the laser positioning is more sensitie to changes in the position a slight iration will affect the positioning of the laser scanner making that data of AGV position shows continuous eating in mm ody which is inconenient for reading and affect the test results so we choose the method of aeraging the testing alues of repeated measurements in the experiment process to otain the actual coordinates of arrial point. XOY coordinate system is estalished within the AGV's work area for the experiment of location and tracking which is used to erify the feasiility and staility of the study. Figure 3. Experiments of tracking guidance As shown in Figure 3 starting pose of path of AGV is M (000-65000) and end position is N (0976-399590.3) during the experiment path sites of AGV has a total of 8 from to N and trael path of the AGV is M 3 4 5 6 7 N the 8 path sites are as shown in Tale. Otain the coordinates of AGV in the asolute coordinate system through the 6

International Journal of Computer Science & Engineering Surey (IJCSES) Vol.6 No.3 June 05 laser positioning technology and compare with the actual coordinates then we can get errors and the final results of the experiment are shown in tale. ' ' ' Tale shows theory posture ( x y z) of each point on the path and the actual pose ( x y z ) of AGV during tracking guidance process. Analysis data shows that the position error is less than ±0mm and the angle error is in±0.3. Experiments show that y the method of laser positioning and path tracing aoe AGV can successfully complete the gien task and erify the feasiility of the research as the error is in the allowale range and meet the performance requirements of AGV industry. Tale. experimental results of tracking guidance REFERENCES [] Shi EnxiuHuang Yumei.Study on Positioning Method of Autonomous Naigation Vehicle [] AGV.Journal of transducer technology0070:33-36. Fang Qiang Xie Cunxi.Autonomous Positioning Naigation of Moile Root Based on Vision[J].Machine Tool & Hydraulics004(7):40 一 56. [3] Ni Zhen. Research on the key technology of laser guidance of four wheel differential omnidirectional moile AGV [D].Chongqing Uniersity03. [4] Liu Yang. Discussion on laser positioning algorithm of laser guidance AGV[J].Logistics Technology and Application007:00-0. [5] Teng Yunlong Shi Yiing. A three-dimensional positioning method ased on three satellites[j].journal of Central South Uniersity0:3449-3453. [6] A New Method of Gloal Path Planning for AGV[J]. International Journal of Plant Engineering and Management0060:5-58. [7] Yuan Wei Sun Jie Cao Zuoliang Tian Jing Yang Ming. A comined oject-tracking algorithm for omni-directional ision-ased AGV naigation[j]. Optoelectronics Letters000:37-39. [8] S.ButdeeA.Suesomran F.Vignat P.K.D.V.Yarlagadda.Control and pathprediction of an Automate Guided Vehicle[J].Journal of Achieements in Materials and Manufacturing Engineering.008(3); P 70-75 [9] Mehdi Yahyaei J.E.Jam&R.Hosnai.Controlling the naigation of automatic guided ehiele (AGV) using integrated fuzzy logic controllerwith programmale logic controller (IFLPLC)IJ8. Int J Ady Manuf Technol.00.47:795-807. [0] Yao Jiajia. Vision-ased Road Detection and Naigation for AGV Systems[D].Zhejiang Uniersity 04. 7

International Journal of Computer Science & Engineering Surey (IJCSES) Vol.6 No.3 June 05 Author Cheng Jing was orn on March 5 990 in Xianning City Huei Proince China. She receied the college diploma in mechanical design manufacturing and automation from the China Three Gorges Uniersity Yichang City China in03.then She was admitted as a graduate student of Shanghai Uniersity of Engineering Science Shanghai China in03. She is now a second-year grad student and majors in rootics and intelligent control. Her interests focus mainly in the area of root emedded systems and intelligent control. 8