MOTION CONTROL USING VOICE FOR WHEELCHAIR APPLICATION
HASHIMAH BINTI ISMAIL
A project report submitted in partial fulfillment of the
requirements for the award of the degree of
Master of Engineering (Electrical – Mechatronic & Automatic Control)
Faculty of Electrical Engineering
Universiti Teknologi Malaysia
APRIL 2006
iii
To my dear family members and friends
iv
ACKNOWLEDGEMENT
Praised be to the Almighty Allah (SWT) who is the most Gracious and
Merciful. With the help and guidance endowed by Him, I able to finish this master
project.
I would like to express my sincere gratitude and respect towards my project
supervisor Professor Dr. Ruzairi for his kind encouragement and suggestions. Special
appreciation goes to Mr Siw and his friends, for helping me throughout the
development of the project. May Allah bless and reward them for their sincere
endeavor and contribution in the way of knowledge.
Thank you to all lecturers, staffs, friends and all who has directly and
indirectly involved in the production of this project. Your helps and cooperation will
never be forgotten.
Lastly, high gratitude and deep thank to my parents En. Ismail and Pn.
Sabariah and all the family members for their continued loves, supports and
motivation.
v
ABSTRACT
This paper describes the significant design to build a voice-controlled
wheelchair. This project is intended to increase the ease of mobility for
disabled/injured people. The design would allow these people to live more
independently. Presently, people use blow-tubes or chin-joysticks to control
motorized wheelchairs. Speech recognition is a prominent technology which can give
an alternative to people to interact with machines or devices especially to those who
are quadriplegics. We have resolved the disabled problems by implementing voice
control interfacing, over a microphone, for the wheelchair. In this project, the manual
wheelchair has been modified so that it can be actuated by two DC motors. The
motions of the wheelchair are then controlled by the verbal instructions of the user.
The results show that the design is applicable and feasible. The speech processing
can be done in real-time and is therefore deemed a viable alternative to present
methods of motorized wheelchair control. The design and the analysis of the project
are presented in this report.
vi
ABSTRAK
Laporan ini menerangkan tentang merekabentuk kerusi roda yang boleh
dikawal dengan menggunakan suara pengguna. Projek ini adalah bertujuan untuk
mempermudahkan pergerakan orang-orang cacat atau yang cedera anggota. Hasil
rekaan ini akan membolehkan orang-orang tertentu untuk menjalani kehidupan
dengan kurang bergantung kepada orang lain. Sekarang kerusi roda dilengkapi
dengan alat yang memerlukan pergerakan fizikal untuk digunakan. Pengenalan suara
menjadi satu teknologi penting yang mana boleh menyediakan suatu jalan yang baru
dalam interaksi manusia dengan mesin atau alat. Ini adalah penting bagi mereka yang
tidak boleh menggerakkan tangan dan kaki. Masalah mereka yang tidak
berkemampuan ini dapat diselesaikan dengan menggunakan teknologi pengenalan
suara bagi mengerakkan kerusi roda. Ini dapat direalisasikan dengan mennggunakan
mikrofon sebagai perantara. Projek ini menggunakan kerusi roda yang telah
diubahsuai dengan memasang dua DC motor sebagai penggerak. Pergerakan kerusi
roda tersebut akan dikawal dengan hanya menggunakan suara. Hasil dari projek yang
dijalankan ini, dapat dirumuskan rekabentuk yang telah digariskan adalah boleh
digunapakai dan. Kesemua hasil ciptaan dan analysis akan diterangkan dalam
laporan ini.
vii
TABLE OF CONTENTS
CHAPTER TITLE PAGE
1 INTRODUCTION 1
1.1 Project Background 1
1.2 Problems Statement 2
1.3 Project Significances 3
1.4 Objectives 4
1.5 Scopes 4
1.6 Organization of report 5
2 LITERATURE REVIEW 6
2.1 Wheelchair Components 8
2.2 Related Researches 10
2.3 Speech Recognition 12
2.3.1 Training HM 2007 voice recognition processor 16
2.3.2 Interfacing Circuit 17
3 METHODOLOGY 19
3.1 System Flowchart 21
3.2 Speech Recognition Board 23
3.3 Product Development 25
3.3.1 Electronic Circuit Development 26
3.3.2 Interfacing Circuit 27
3.3.3 Circuit To Control The Wheelchair Direction 30
3.3.4 Circuit To Control The Speed 30
3.3.5 Wheelchair Development 31
4 RESULT AND ANALYSIS 40
4.1 Accuracy for voice recognition circuit 40
viii
4.2 Velocity 42
5 CONCLUSION AND RECOMMENDATIONS FOR
FUTURE WORKS
44
5.1 Conclusion 44
6.2 Recommendations 45
REFERENCES 46
APPENDICES 47
ix
LIST OF TABLES
TABLE NO. TITLE PAGE
1.1 Power Wheelchair Control Interface 3
2.1 Speech Recognition Techniques 14
2.2 Speech Recognition Program 14
2.3 List of Voice Recognition Processors 15
3.1 Voice Command 22
3.2 The Memory Used For Storing Commands 25
3.3 The Binary Codes and Commands 29
4.1 The Result In Silent Area 40
4.2 The Result In Noisy Area 41
x
LIST OF FIGURES
FIGURE NO. TITLE PAGE
2.1 Joystick Control Block Diagram 8
2.2 Joystick Axis Interpretation 8
2.3 Basic Components of Electrical Power Wheelchair 10
2.4 Voice Controlled Wheelchair User 11
2.5 HM2007 Voice Recognition Circuit 16
2.6 Interface Circuit 18
3.1 Flowchart For The Project 20
3.2 Flowchart For The Motion Controlled Wheelchair Using
Voice
21
3.3 SR-07 Speech Recognition Kit Circuit 23
3.4 Speech Recognition Kit 26
3.5 Electronic Circuit System 27
3.6 Interfacing Circuit To The Motors 28
3.7 Circuit To Control The Motors Speed 31
3.8 Manual Wheelchair Drawing 31
3.9 The Power Window Motor 32
3.10 The Driver Pulley 33
3.11 Power Window Cover 33
3.12 The Gear Welded On Driver Pulley 34
3.13 The Finished Part of Motor With Pulley 34
3.14 The Bracket for Motor 35
3.15 The Motor Assembly 35
3.16 The CATIA Drawing For The Bush 36
3.17 The Finished Part Using Lathe Machine 36
3.18 The Welded Bush 37
3.19 The Plate Assembly With Big Pulley 37
xi
3.20 The Electrical Wheelchair 38
3.21 The Completed Voice Controlled Wheelchair 38
4.1 The Graph of Accuracy of The SR-07 in Two Conditions 42
xii
LIST OF SYMBOLS
DC - Direct Current
SR - Speech Recognition
RAM - Random Access Memory
SRAM - Static Random Access Memory
LED - Light Emitting Diode
R - Relay
V - Velocity
ECU - Environmental Control Unit
ANN - Artificial Neural Network
BPA - Back Propagation Algorithm
FFT - Fast Fourier Transform
LVQ - Learn Vector Quantization
DTW - Dynamic Time Warping
DSP - Digital Signal Processing
HMM - Hidden Markov Model
AC - Alternate Current
CLR - Clear
TRN - Train
PIC - Peripheral Interface Controller
xiii
LIST OF APPENDICES
APPENDIX TITLE PAGE
A Interfacing Circuit Schematic Diagram 47
B PCB layout of Interfacing Circuit 48
C HM 2007 Voice Recognition Description 49
CHAPTER 1
INTRODUCTION
While the needs of many individuals with disabilities can be satisfied with
power wheelchairs, some members of the disabled community find it is difficult or
impossible to operate a standard power wheelchair. This project could be part of an
assistive technology. It is for more independent, productive and enjoyable living. The
background, objectives, significance and scopes of the project will be discussed in
this chapter.
1.1 Project Background
The idea of using voice activated technology for controlling the motion of the
wheelchair is to prove that it can be a unique concept that would stand apart from the
rest of the average projects. The use of this new technology in conjunction with a
mechanical system in order to simplify everyday life would spark interest in an ever
growing modern society. Many people with disabilities do not have the dexterity
necessary to control a joystick on an electrical wheelchair. This can be a great for
the quadriplegics who is permanently unable to move any of the arms or legs. They
can use their wheelchair easier only using voice commands. The aim of this study is
to implement an interesting application using small vocabulary word recognition
system. The methodology adopted is based on grouping a microprocessor with a
speech recognition development kit for isolated word from a dependent speaker. The
resulting design is used to control a wheelchair for a handicapped person based on
the vocal command. It therefore involves the recognition of isolated words from a
2
limited vocabulary. In order to gain in time design, tests have shown that it would be
better to choose a speech recognition kit and to adapt it to the application.
There are five options for basic motions of a wheelchair to be applied by the
user. The five conditions of the wheelchair can be described as the following:
i. Moving forward to the front of the user
ii. Moving backward to the back of the user
iii. Turning to the right
iv. Turning to the left
v. Static or stop condition
In this project the extra options are designed so that the user can choose the
speed. The speed is divided into two parts. The user can select either slow or fast
speed to move. This speed selection is in important for safety and extra
maneuverability of the user. For example if the user need only to move in a short
distance or to approach object, he should use the slow speed. This paper describes
the design and development of the motion control using voice recognition for a
wheelchair application.This design then been tested and analyzed.
1.2 Problem Statements
Research from University of Notre Dame, 2000, suggests that the current
power wheelchair control interfaces used may not, be adequate to provide truly
independent mobility for substantial number of person with disabilities. The
Respondents to the survey reported on average that approximately ten percent of the
patients trained to operate a power wheelchair cannot use the chair upon completion
of training for activities of daily living or can do so only with extreme difficulty
(Linda Fehr,2000). The data of the patients is as the table 1.1 below. From the table,
we can see the list of the main types of control interfaces employed by power
wheelchair users and the adequacy of these controls.
3
Percent of patients using Simple Average *Weighted Average
Joystick 81 81
Head or chin control 9 9
Sip and puff 6 9
Others- eye gaze; tongue pad;
head, hand, foot
switch controls
4 1
TOTAL 100 100
* weighted by total number of power wheelchair users reported in survey
Table 1.1: Power wheelchair control interfaces used
The challenge for engineering is to provide safe and effective mobility in a
dynamic environment. Through thoughtful research and design, power wheelchair
control will progress along safe and effective pathways towards providing users
independent and self-guided mobility. This project will give the severely disabled
people an innovative solution to control their wheelchair using voice interfacing.
1.3 Project Significances
User interface is an important component of any product handle by the
human user. The concept of the design is to make a voice activated wheelchair,
which can replace the use of a joystick. In the past decades GUI (Graphical User
Interface), Keyboard, Keypad, Joystick is the dominating tools for Interaction with
machine. Now from them SR system is one of the interesting tool to the researchers
for interaction with machine. The reason draws attention to the researcher, because
people are used to communicate with a natural language in the social context. So this
technology can be widely-accepted to the human user fairly and easily. But for the
wheelchair application more researches and more analysis have to be done. This is
because this will include the human safety and more over this kind of application is
very new especially in Malaysia. Thus the project is significant because:
i. Speech processing can be done in real time and has long been considered
as a natural to assist powered wheelchair user.
4
ii. Many disabled people exist in today’s world and require help in order to
overcome physical challenges. Thus this project will provide an
alternative to the disabled in controlling the motion of the wheelchair
using their voices.
iii. The efficiency of using voice controlled wheelchair can be identified.
1.4 Objectives
i. To implement voice of the user as an input to control the speed of a
wheelchair.
ii. To develop a voice interface system for wheelchair control.
iii. To construct an effective algorithm for voice recognition.
iv. To provide an extra alternative to the wheelchair users so that this can
increase the ease of mobility for severely disabled/injured people.
1.5 Scopes
i. To study the currents systems, researches, components used for
wheelchairs.
ii. To design and develop a wheelchair system which can be controlled using
voice.
iii. To build up the speech recognition interfacing using voice recognition
processor and to train the system accordingly.
iv. To design switching modes for controlling the motion of the wheelchair.
v. To build up the interfacing between the hardware and software to realize
the real time application.
vi. Integration of all of the components needed and testing.
5
1.6 Organisation of the report
In the following chapter we are going to discuss more about the literature
review in chapter 2, the methodology in chapter 3, result and analysis of the system
in chapter 4, and final chapter is the conclusion plus the recommendations. At the
end of the report the list of references and related appendices are attached.
We start with the literature review about the wheelchair evolution and voice
recognition system (Chapter 2 on page 6). Then we discuss about the flow of the
project and the important components of the project development in Chapter 3 (on
page 19). This includes speech recognition Chip, circuit interfacing and the hardware
development. In the result and analysis chapter, Chapter 4 (on page 40) contains the
description about the implementation part of our project. There, we discuss about the
result of the system, and also we have presented our test result in that chapter. We
conclude in the project as well as suggestions for future works in Chapter 5 (on page
44).
CHAPTER 2
LITERATURE REVIEW
2.0 Introduction
Wheelchairs have evolved very little over the past 1000 years. Most of the
design changes have occurred within recent decades as shown in the following
outline of wheelchair history.
6th Century A.D. - Earliest recording of a wheelchair; a Chinese engraving picturing
a man in a chair with three wheels (Kamenetz, 1969).
16th Century A.D. - Wheelchairs were well-developed in Europe and commonly
found in drawings and literature (Kamenetz, 1969).
1869 - The first wheelchair patent was issued in the United States (Hotchkiss,1993).
1903 - An electrically-driven wheelchair operating on a 12-volt battery and a 3/8
horsepower motor was used to give people rides. At the time it was not used
for handicapped mobility but it did pave the way for future developments
(Kamenetz, 1969).
1909 - Compact wheelchairs were developed using metal tubing instead of the
traditional bulky wood components (Kamenetz, 1969).
World War I - The first electric wheelchairs were used for the handicapped. A
battery and motor were applied to existing wheelchairs with a simple
one-speed on/off switch (Kamenetz, 1969).
1937 - The patent for a wheelchair with a folding X-brace frame was issued to two
engineers named Everest and Jennings. Though previous chairs had been
foldable top-to-bottom, the side-to-side folding position of the cross frame
7
allowed the drive wheels to remain in place. This basic concept is still the
standard for manual wheelchairs today (Hobson, 1990).
1940 - The first patent was issued for an electric wheelchair (Hobson, 1990).
1950 - Sam Duke received a patent for a releasable add-on power drive applied to
manual wheelchair (the unit was actually permanently fitted to the chair with
Ubolts) (Kamenetz, 1969).
1960’s - Folding wheelchairs were commonly fitted with electric drives. The drive
units were still very heavy and quite difficult to put on and take off. At that
point both joystick and steering column mechanisms were available
(Kamenetz, 1969).
1970’s - Wheelchair frames made of aircraft quality aluminum were introduced to
the market and started a revolution of ultralight wheelchairs. The
technology has aided in the reduction of the overall weight of many
types of wheelchairs (Hobson, 1990).
1980’s - Most electric wheelchairs on the market were still bulky, heavy, and
required a special vehicle for transportation. The power components of
the chair were integrated into the frame which has been strengthened to
support them (Hobson, 1990).
1990’s - The popular electric wheelchairs on the market are foldable though they
require removal of at least the leg rests and batteries. The Katalavox
speech-recognition control system can be used by quadriplegics to control
their power wheelchair. The commands are combined to emulate the
movements of a joystick. This voice controlled wheelchair was not been
commercialized but it is customized for individual used.
2000’s – The use of joystick, head or chin control and sip and puff control for
severely disabled people are recognized. There are also other
interfacing used like eye gaze; tongue pad; head, hand, foot switch
controls.
8
The block diagram of the current system using joystick is as the figure 1 below:
Figure 2.1: Joystick control block diagram
Figure 2.2: Joystick axis interpretation
In this configuration, the human operator applies a force on the input joystick
in order to drive the wheelchair to the desired position. The position of the joystick is
interpreted as desired speed and direction according to the Figure 2. The control
algorithm calculates the appropriate commands for the right and the left wheel
motors in order to drive the wheelchair in the desired speed and direction. The
human operator observes the present position of the wheelchair and modifies the
applied force such that the present position approaches the desired one.
In this paper, the use of the joystick in controlling the speed in this system is
replaced to the voice recognition processor so that it can be wireless speed control
using voice wheelchair.
Human
operator
Joystick
Control
algorithm
Wheel-
chair
dynamic
Desired path
Force applied
speed
direction Applied
commands
Actual
path
9
2.1 Wheelchair Components
There are a number of possible driving wheel configurations (front wheel
drive, rear wheel drive and mid wheel drive) which affect the characteristics of the
chair in different situations, with turning while driving being the most complex.
Further features can be added to assist the user such as lights, actuators and wireless
links.
The heart and brains of the powered wheelchair is in the controller as it
provides both a conduit for the power to the motors and controls the overall system.
The typical powered wheelchair user is disabled in a way that means they rely almost
totally on the software contained within this controller to provide safe and reliable
performance. This reliance is the same for every user, no matter the ability,
preference, or operating environment.
The general features of the wheelchair are as the following.
i. The user interface. It can be a programmable joystick, however many
other methods of control are possible (sip and puff, scanning, head
movement, etc.) as power wheelchairs have sophisticated electronics to
control their motors.
ii. The seating and postural support. Some power wheelchair models have
features like power stand, power recline, power tilt, and power elevation.
iii. Power wheelchairs come with more tire and powerbase options.
iv. Prices of the mobility rises depending on the features it has.
v. Power wheelchairs have a variable type speed control knob so we can set
the speed from 1 to 5+ mph. It can accept a 4 point tie-down in a motor
vehicle which could tend to make them safer as a seat in a motor vehicle.
vi. Batteries for most power wheelchairs are gel cell sealed batteries which
are approved for transport. Most electric wheelchairs are equipped with 2
gel cell 12 volt batteries capable of going 15 to 20 miles on a full charge
over level terrain. These sealed batteries are approved for transport.
10
vii. Very few power wheelchairs are breakdown for placement in the trunk of
a car. Vehicles most always need to be adapted with a lift or ramps
because they are too heavy.
Figure 2.3: Basic Components of Electrical Power Wheelchair
2.2 Related Researches
Several researchers have considered using human voice to control powered
wheelchairs, see, e.g., Simpson and Levine (2002) and the references listed therein.
Naturally, a wheelchair voice control system should operate reliably for a large
number of users, reduce the physical requirements; and if avoiding the need to move
on one or more road extremities, should assist a user in maintaining well the chair
position. However, the voice’s limited bandwidth makes it difficult to adjust
frequently the wheelchair’s velocity, and also a voice input system may fail to
identify a speaker. Thus, voices interface has yet to become commercially viable for
wheelchair control; rather its use is normally suggested in combination with a
navigation assistance system for obstacle identification and avoidance in the
wheelchair’s path (Q.P. Ha, T.H. Tran and G. Dissanayake, 2005).
The example of other research doing the Voice Recognition is done by
acquiring the Microsoft SDK 5.1 software development kit, which has the necessary
Input Device / Microphone
Controller
Battery
Motors Power Controller
11
capabilities. The team will be using the Speech Recognition (SR) engines provided
by the SDK to interpret voice commands. The control algorithms for smooth
movement are an incremental system, thus the current state of the velocity
parameters (V, ?) will be updated based on the specific voice commands. This allows
the developer to use a small number of commands to create a fluid and flexible
control motion, while also maintaining a short training period and ease of control for
the user.
Speech recognition systems were first used by severely disabled individuals
with normal speech. The goal was to promote independence whereby SR was used to
convert human speech signals into effective actions. Frequently, speech is the only
remaining means of communication left for these individuals. The first voice
activated wheelchair with an environmental control unit (ECU) was developed in the
late 1970s at Rehabilitation Medicine in New York (Youdin, et al., 1980). The user
could operate multiple items including the telephone, radio, fans, curtains, intercom,
page-turner and more. A group of individuals with cerebral palsy rated the
wheelchair as superior to breath control systems because it eliminated the need for
scanning, allowing the user quicker access by directly selecting the desired function
with voice. (Nancy Manasse,1999)
The first voice activated power wheelchair was used by a young Norwegian
law-student in 1984. It enabled him to attend his classes without the help of an
attendant. The wheelchair was customized for him using Katalavox speech-
recognition system.
Figure 2.4: Voice controlled wheelchair user
There are some other researches have been done to upgrade the wheelchair
using assistive technology. A good research example is done by S. Rao and R. Kuc
Figure 2.4: Mr. Robert Kotz has used his voice-activated wheelchair from
1986 to the end of 1992 when he passed away. (katalavox.com)
12
from Yale University. They built and analysis a prototype of an intelligent
wheelchair which is equipped with ultrasonic range sensors and a Motorola 68HC11.
The prototype was designed to detect fall-offs and objects in its path, thereby
addressing the needs of visually impaired persons who has been confined to
wheelchairs.
2.3 Speech Recognition
Speech recognition is the process of converting an acoustic signal, captured
by micro- phone or a telephone, to a recognized command or word. There two
important part of in Speech Recognition - i) Recognize the series of sound and ii)
Identified the word from the sound. This recognition technique depends also on
many parameters - Speaking Mode, Speaking Style, Speaker Enrollment, Size of the
Vocabulary, Language Model, Perplexity, Transducer etc. There are two types of
Speak Mode for speech recognition system - one word at a time (isolated-word
speech) and continuous speech. Depending on the speaker enrolment, the speech
recognition system can also divide - Speaker dependent and Speaker independent
system. In Speaker dependent systems user need to be train the systems before using
them, on the other hand Speaker independent system can identify any speaker’s
speech. Vocabulary size and the language model also important factors in a Speech
recognition system. Language model or artificial grammars are used to confine word
combination in a series of word or sound. The size of the vocabulary also should be
in a suitable number. Large numbers of vocabularies or many similar sounding words
make recognition difficult for the system.
The most popular and dominated technique in last two decade is Hidden
Markov Models. There are other techniques also use for SR system - Artificial
Neural Network (ANN), Back Propagation Algorithm (BPA), Fast Fourier
Transform (FFT), Learn Vector Quantization (LVQ), Neural Network (NN).
(Shafkat Kibria, 2005)
13
Two basic choices are available in recognition algorithms: Dynamic Time
Warping and Hidden Markov Models. DTW has lower requirements on hardware.
HMMs are more complex; they yield better recognition scores but also require more
speech data in the training phase. Even with a limited command set and speaker
dependence, memory remains the most important limiting factor. Using common
speech parameterization, approximately 1k x 16 per command (assuming about a 1-
second length) is usually needed. The speech data is created by the user during the
training of the recognition system.
The input buffers will use RAM (with a size equal to the speech-frame size),
as will the buffer for the recognition algorithm, or DTW matrix. Assuming the
system will have 30 commands, one user and a sampling frequency of 8 kHz, a rough
estimation of data memory consumption would be 32k x 16 of flash and 8k x 16 of
RAM. A minimalist solution would require the minimum consumption of program
memory (about 32k x 16), but the value ultimately depends strongly on the processor
instruction set and compiler efficiency. (Richard Mensik,2001)
The digital processing capabilities of microcontrollers have enabled voice
control to penetrate embedded systems. These new microcontrollers, sometimes
called embedded digital signal processors or DSP controllers, have sufficient
performance for real-time speech processing, and they integrate almost all needed
control peripherals on one piece of silicon.
Voice control implies that the system will recognize only a limited command
set, not fluent speech. That limitation markedly decreases memory and performance
requirements compared with those needed for fluent-speech recognition.
"Embedded," meanwhile, implies a single-chip solution. The technique of voice
recognition techniques available are as the table 1. (Shafkat Kibria, 2005)
Technique Sub Tech-
nique
Relevant
Variable(s)/Data
Structures
Input Output
Sound
sampling
ALL
Analog
Sound
Signal Analog Sound Signal
Feature
Extraction
Dynamic
Time
Warping
Statistical Features
(e.g.
LPC coefficients)
Digital Sound
Samples
Acoustic
Sequence
Templates
14
(DTW)
Hidden
Markov
Models
(HMM)
Subword Features
(e.g.
phonemes)
Digital Sound
Samples
Subword
Features
(e.g.phonemes)
Artificial
Neural
Networks
(ANN)
Statistical Features
(e.g.
LPC coefficients)
Digital Sound
Samples
Statistical
Features (e.g.
LPC coeffi-
cients)
Dynamic
Time
Warping
(DTW)
Reference Model
Database
Acoustic
Sequence
Templates
Comparison
Score
Hidden
Markov
Models
(HMM)
Markov Chain Subword
Features
(e.g.phonemes)
Comparison
Score
Training&
Testing
Artificial
Neural
Networks
(ANN)
Neural Network
with
Weights
Statistical
Features (e.g.
LPC)
Positive/Negative
Output
Table 2.1: Speech Recognition Techniques
There are both Speech Recognition Software Program and Speech
Recognition Hardware is available now in the market. See table 2.2 for the available
SR programs for developer and their vendors. (Shafkat Kibria, 2005)
SR programs for
developer
Vendors
IBM Via Voice IBM http://www306.ibm.com/software/voice/
viavoice/
Dragon Naturally Speaking
8 SDK
Nuance http://www.nuance.com/naturallyspeaking
/sdk/
Voxit http://www.voxit.se/ (Swedish)
VOICEBOX: Speech
Processing
Toolbox for MATLAB
http://www.ee.ic.ac.uk/hp/staff/dmb/voicebox
voicebox.html
Java Speech APIa Sun Microsystems, Inc http://java.sun.com/products
/javamedia/ speech/index.jsp
The CMU Sphinx Group
Open Source Speech
Recognition Engines
http://cmusphinx.sourceforge.net/html/cmusphinx.php
SpeechStudio Suitec SpeechStudio Inc. http://www.speechstudio.com/
Table 2.2 Speech Recognition Program
15
There are many voice recognition processors available in the market. In table
2.3, there are some of the of voice recognition module or processor.
SR Module/Processor Manufacturer
Voice ExtremeTM Module Sensory,Inc.
http://www.sensoryinc.com/
VR StampTM module Sensory,Inc.
http://www.sensoryinc.com/
HM2007 - Speech Recognition Chip HUALON Microelectronic Corp. USA
OKI VRP6679 – Voice Recognition
Processor
OKI Semiconductor and OKI Distributors
Corporate Headquarters 785 North Mary
Avenue, Sunnyvale, CA, 94086 2909
Speech Commander – Verbex Voice
Systems
Verbex Voice Systems 1090 King
Georges Post Rd., Bldg 107, Edison NJ
08837, USA
Voice Control Systems Voice Control Systems, Inc.
14140Midway Rd., Dallas, Tx. 75244,
USA http://www.voicecontrol.com/
VCS 2060 Voice Dialer Voice Control Systems 14140 Midway
Rd., Dallas, Tx. 75225, USA
http://www.voicecontrol.com/
DVC306 Processor DSP Communications,Inc. 20300 Stevens
Creek Blvd. suite 465 Cupertino, CA
95014 USA
D6106 Processor DSP Communications,Inc.) 20300
Stevens Creek Blvd. suite 465 Cupertino,
CA 95014 USA
TC8860F, 64F, 65F Processor Toshiba, 1-1, Shibaura 1-Chome, Minato-
ku, Tokyo, 105-01,JAPAN
5A128,custom Processor Ricoh, Electronic devices Division San
Jose Office 3001 Orchard Parkway, San
Jose, CA 95134-2088, USA
In this project HM 2007, Hualon voice recognition processor is chosen. It
can be programmed words for both speaker dependent and speaker independent.
Moreover it is easy to program and low cost. The specifications of the HM 2007 are
as the following:
• IWR and SI.It's a single chip CMOS LSI chip.
• 48 pin plastic DIP package,
• 5V single power supply,
• 6mA operating current (idle),
• 15mA operating current (max),
• response time less than 300 ms.
• It can work with an electret microphone,
16
• It needs an external 8Kbyte Static RAM,
• It needs an external micro-controller for specific application.
Figure 2.5: Voice recognition Circuit
2.3.1 Training the HM 2007 voice recognition processor
The chip is developed using Dynamic Time Warping. As all the components
are connected and soldered, we can use the microphone, keypad and digital display
to communicate with and program the HM 2007 chip. Speech recognition is
classified into two categories, speaker dependent and speaker independent. Both can
be done using this chip.
Speaker dependent systems are trained by the individual who will be using
the system. These systems are capable of achieving a high command count and better
17
than 95% accuracy for word recognition. The drawback to this approach is that the
system only responds accurately only to the individual who trained the system. This
is the most common approach employed in software for personal computers.
Speaker independent is a system trained to respond to a word regardless of
who speaks. Therefore the system must respond to a large variety of speech patterns,
inflections and enunciation's of the target word. The command word count is usually
lower than the speaker dependent however high accuracy can still be maintain within
processing limits. Industrial requirements more often need speaker independent voice
systems, such as the AT&T system used in the telephone systems. It is possible to
use a different person speaking the word. This will enable the system to recognize
different voices, inflections and enunciations of the target word. The more system
resources that are allocated for independent recognition the more robust the circuit
will become.
2.3.2 Interface Circuit
The schematic for the interface circuit is shown in the figure 5. The circuit
connects to the 10 pin Right Angle interface header on the circuit board. This header
is also used for the Digital Display board.
18
Figure 2.6: Interface circuit
The 4028 has ten output lines. Whatever number is displayed on the LCD the
corresponding line number off the 4028 will be brought high. The high signal from
the 4028 can be connected to a NPN transistor to control a DC load as shown figure
5 (A) or control an AC or DC load using a simple relay as shown in figure 5 (B). The
disadvantage in using a simple set up like these two is that only one switch out of ten
may be turned on at any given time. This doesn’t make for a very good system. To
configure in allowing one to turn on or off any line without changing the status of
any other line, inserting a flip-flop can be a solution. The flip-flop acts like a simple
memory. When the input is brought high, its output line goes high, turning on the
NPN transistor. When the output line is brought low, the output line still stays high.
When the flip-flop receives a second high signal on its output line it brings the output
low.
CHAPTER 3
METHODOLOGY
3.0 Introduction
In order to make this project successful, there are several methodologies have
been carried out. After doing the literature review, development and testing have
been done. Research methodology is the system of methods and rules for conducting
research. In order to achieve the objectives right methodologies have been chosen for
it. The correct flow will make the work become systematic and easy. The figure 3.1
the flow chart shows the sequence of the project that has been done.
The literature review is very useful for the future development of the project.
In literature review part, previous work by different researchers are analyzed and
compared. There is also need to study on the actual product for developing and
modifying on the project. The information that has been studied is well written in
chapter two.
20
Figure 3.1: Flow chart for the Project
Literature
review
stage
Design and
Development
stage
Experimental
stage
Experimental
stage
Modify part
and design
part
Testing Testing
Assembly
Testing the
completed
system
End
Prepare
material
Mechanical
part
Electronic
part
Start
Speech
Recognition
Wheelchair
Design
Build motor drive
circuit and interfacing
circuit
21
3.1 System Design
The main part of the design is to control the motion of the wheelchair. There
are four types of motions are considered, moving forward, moving in reverse
direction, moving to the left and moving to the right. For the speed, the user may use
slow or fast speed. Slow speed is important as the user want to move in short
distance or approaching an object. The system is designed as the following
flowchart.
Figure 3.2: Flowchart for the motion controlled wheelchair using voice
The system starts by applying the supply voltage to the speech recognition
circuit. The system will be in stand by condition which the LED on circuit
recognition board will be turned on. The system can be controlled in two speed
conditions which are fast and slow. For fast condition the system will supply higher
current to the motors. If the user does not want the wheelchair move in high speed,
the slow speed can be set by applying low current supply to the motors.
22
The direction and speed of wheelchair depend on the user. Forward command
the wheelchair move in forward direction. For the reverse direction the opposite
movement of wheel rotation will occur. The left command will make right wheel
moves forward and left wheel moves backward. The right command makes left
wheel moves forward and right wheel rotate backward.
In this system, by assigning the word command stop the rotation of both
motors will stop. The wheelchair system will go back to the stand by condition or
end the whole system by turning off the power supply of the speech recognition
board. The voice commands used are as the table 3.1 below.
Voice
Command Condition
forward Moving straight to the forward
reverse Moving straight in the backward
fast Setting the speed level high
slow Setting the speed level low
left Turning to the left
right Turning to the right
stop No motion/wheelchair stops
Table 3.1: Voice command
3.2 Speech recognition board
As all the voice recognition processor and software are distinguished, the HM
2007 is selected. The criteria of selection are as the following:
i. Availability
ii. Cheap
iii. Simple interfacing
iv. Applicable
23
v. Programmable
vi. Independent/portable
This processor is the best choice due to the above criteria. It is completely an
easy to build programmable speech recognition circuit. Programmable in the sense
that we train the words (or vocal utterances) that we want the circuit to recognize.
This kit allows us to experiment with many facets of speech recognition technology.
Unlike software based speech recognition systems like Dragon naturally speaking
(tm) and Via Voice (tm), it is stand alone circuit and works without a personal
computer. It can be reprogrammed as the datasheet for the chip is accessible. The
datasheet is attached in the appendix. To train the voice, the keypad of made up of 12
switches is used.
Figure 3.3: SR-07 speech recognition kit circuit
The SR-07 speech recognition circuit operates as the main part for storing the
command in to the HM2007 chip. This speech recognition system uses a simple
keypad and digital display to communicate with and program the HM2007 chip.
24
When the circuit was turned on, the HM2007 chip will check the static RAM. If
everything checks out the board displays "00" on the digital display and lights the red
LED (READY). It is in the "Ready" waiting for a command. In this project, the
display board will be taking off from the output of speech recognition circuit and
replace it by the interfacing circuit which will connect to the motor driven circuit.
The step begins by pressing the word number which wants to train on the
keypad. The circuit can be trained to recognize up to 40 words. It can be used any
numbers between 1 and 40. In another way if the 40 word vocabulary is not
desirable, it can configure the circuit for the 20 word vocabulary as this configuration
usually provides better recognition accuracy. In this project, second method has been
chosen because only have seven commands are needed to execute. It is started by
pressing the number "1" to train word. When the number is activated, on the keypad
the red LED will turn off. The chosen number is displayed on the digital display.
Before we replace the display board with interfacing circuit, the confirmation need to
take by looking at digital display from board for make sure that the right commands
was trained. Then we pressed the "#" key to start training the voice commands. When
the "#" key is pressed it sends the signals so that the chip could catch for the training
word and the red LED turns back on. Then, the user should say the specific
command clearly to the microphone. LED will blink off momentarily showing that
the word is trained. Finally, one by one of the commands which are list in the table
3.1 are speak out with different number. For each of them, the LED should blink off
momentarily; this is a sign that the word has been accepted.
The chip provides the following error codes: 55 for word too long, 66 for
word too short and 77 for word no match. If i want to retrain the speech recognition
kit, the number 99 is pressed at the keyboard and then button “CLR”. The numbers
will quickly scroll by on the digital display as the memory is erased.
There is also another way to change or erasing words. The words can easily
be changed by overwriting the original word. For instance suppose word two was the
word “back” and if we want to change it to the word “reverse”. By pressing “6” then
25
the TRN key and saying the word “reverse” into the microphone then the word had
been trained. If we wish to erase the word without replacing it with another word
then we can just press the number of the word which is desired to erase and press
“CLR” key. So, all words in the memory are erased. In this project the memory used
be train to the speech recognition are as table 3.2.
Digital display
(Decimal)
Voices Conditions
06 Forward Move straight to the forward
04 Reverse Move straight in the backward
02 Slow Set the speed level low
01 Fast Set the speed level high
03 Left Turn to the left side
05 Right Turn to the right side
07 Stop Stop the system
Table 3.2: The memory used for storing commands
3.3 Product Development
The system or product development is separated into five main sections. First
one is to develop the electronics circuit for voice recognition circuit, second one is on
the developing the interfacing circuit, third one is designing the circuit for controlling
the direction of the motor, the fourth is designing the circuit to set the speed of the
motor and the last one is to develop modified the manual wheelchair and interface it
with the control circuit.
3.3.1 Electronic Circuit Development
Developing the circuit for the voice recognition circuit is easy as the speech
recognition kit is used. In this kit, the HM 2007 processor is already assembled with
26
the input and output port, memory chip and the digital display. The instructions given
by the supplier must be followed carefully so that the system can work properly.
Figure 3.4: Speech recognition kit
Below is the block diagram of interfacing circuit which is used in this system.
27
Figure 3.5: Electronic circuit system
3.2.2 Interfacing Circuit
The interfacing electronic circuit is used to connect the real system. This
design of circuit used 4028 decodes, the output can be up to ten units but since there
are only seven commands. So, in this part, only seven units are used. The effective
vocabulary was decreased from forty to ten words. This was to gain a more robust
and accurate system.
Microphone Speech
recognition
circuit
Interfacing
circuit
Speed control
circuit
Motor
circuit
Motor
circuit
Motor Motor
28
Figure 3.6: Interfacing circuit to the motors
Figure 3.6 shows the interfacing circuit. The circuit is connected from the 10
pin interface header of the speech recognition circuit board. This header is also used
as data cable in Digital Display board. The schematic diagram of the circuit is shown
in appendix 1.
The output signal of the speech recognition circuit was sent in binary form
and the 4082 chip converted the signal into the decimal form. For example, if
forward command was trained as 06. So, the speech recognition circuit will generate
a binary code 0110. The 4028 chip converted the signal and show an output of 1 at
leg seven of the chip. For other commands, the sequence to convert the signals was
the same as the one mentioned above. The difference was the binary codes. Table 3.3
shows the binary codes and the related commands.
The interfacing circuit has seven relays which were labeled as R1, R2, R3,
R4, R5, R6 and R7. When the coils of R1, R2, R3 and R4 are contacted, direction
signals will be sent. R1 relay controlled the forward movement and R3 controlled the
reverse movement of the motor. The left and right commands are controlled by R2
and R4.
Digital dispaly voices Binary codes
R1
R2
R3
R4
R5
R6
R7
Y1
G2
G1
Y2
4028
29
01 Fast 0001
02 Slow 0010
03 Left 0011
04 Reverse 0100
05 Right 0101
06 Forward 0110
07 Stop 0111
Table 3.3: The binary codes and commands
When controlling the speed of the motor, the R5 or R6 relays were contacted.
R5 controls the slow mode while R6 was for the fast mode. Finally, the R7 use to
stop the wheelchair when it is contacted.
The directions of motor are indicated by four LEDs on the interfacing circuit.
When forward command is given, both yellow LEDs, Y1 and Y2, will be on while
both green LEDs, G1 and G2, will be on when reverse command was given. For the
left command, Y1 and G1 LEDs will turned on. Y2 and G2 LEDs are on when the
right command is given.
The output signal voltage from the speech recognition circuit is 5V but the
relays operate on 9V supply. So, a jumper from the speech recognition circuit was
connected to the interfacing circuit. When 5V output signal from the speech
recognition is given, it activates the transistors which permit 9V supply from speech
recognition circuit to flow and operate the relays. Refer to Appendix for detail
electronic schematics drawing.
3.3.3 Circuit to Control the wheelchair direction
30
Figure 3.5: Circuit to control the motor
Figure 3.5 shows the circuit to control the motor direction. When the output
signal from the interfacing circuit was given, relays R8, R9, R10 and R11 will
contact base on the command given. For example, left command will connect R8 and
R10 and make the two motors rotate in clockwise direction. If right command is
receive, relays R9 and R11 will contact and the two motors rotate in counter
clockwise direction. For forward command, relay R8 will connect motor 2 to rotate
clockwise and R11 to rotate motor 1 counter clockwise. While for reverse command,
it was the opposite of forward command. Refer to Appendix for detail electronic
schematics drawing.
3.3.4 Circuit to Control the speed
Figure 3.7 shows the circuit to control the motor speed. This circuit will only
be operated when the relay R6 is contacted. The input voltage of 12V from the car
battery will be supply to the wheelchair sytem, otherwise a input of 6V was supplied.
Refer to Appendix for detail electronic schematics drawing.
R8 R9
R10 R11
31
Figure 3.7: Circuit to control the motor speed
3.3.5 Wheelchair Development
Figure 3.8: Manual wheelchair drawing
In this section, the manual wheelchair is modified into an electrical
wheelchair which is controlled using voice command. In this project, a real
wheelchair is needed to perform the demonstration.
The important part is to upgrade the manual wheelchair into an electrical
wheelchair. Thus, the addition parts like motors, pulleys, belts and a battery are
needed. With the combination of these mechanical and electrical parts, the manual
wheelchair now is turned to be an electrical wheelchair.
32
Figure 3.9: The power window motor
In today’s market, the electrical powered wheelchair uses a wheelchair motor
which was specially designed for the purpose to move the wheelchair with a load.
The wheelchair motors normally have high torque and high revolution per minutes
(rpm). These criteria will make sure that the electrical wheelchair can move smoothly
when it is being used. In this project, the wheelchair motor is replaced with an
automotive power window motor. Figure 3.9 shows the power window motor used
by Proton Iswara, which is used in my wheelchair system. This power window motor
also has high torque but the revolution per minutes (rpm) is low. So, the capacity of
this modified wheelchair system has a difference compared to the ones in the market.
The motor is used to rotate the pulley system. So, two smaller pulleys are
connected to two motors. These driver pulleys (figure3.10) are taken from the
alternator used in vehicles.
33
Figure 3.10: The driver pulley
Figure 3.11: The cover of the power window to remove the gear for welding
Before welding the pulley onto the gear of power window, the power window
motor’s gear cover is remove temporary. Then, remove the gear for welding. Figure
3.11 shows the gear removed from the housing. This step is taken to avoid the rubber
parts being melted by the high temperature from the welding. Figure 3.12 shows the
gear welded on the pulley. It is important to make sure that the gear is in the center of
the pulley since this will affect balance of the rotation.
Using arc welding, the power window motor’s gear was welded at four
points. Figure 3.12 shows the finished part of the driver pulley joined with motor
gear.
34
Figure 3.12: The gear welded on driver pulley
Figure 3.13: The finished part of motor with pulley
A bracket is needed to place the motor on the wheelchair. The bracket must
be strong enough to support the motor when motor is operating to push the rear
wheel. The position for placing the bracket of the motor was previously the manual
brake system of the wheelchair. This position is chosen because it is the most
suitable location and required minimum modification. The design of the bracket was
originally a piece of metal from the brake system and it is welded with a steel bar to
support the motor. Figure 3.14 shows the bracket of power window motor.
35
Figure 3.14: The bracket for motor
Then the motor was screwed to the bracket and assembled onto the
wheelchair. Figure 3.15 shows the motor attached to the wheelchair.
Figure 3.15: The motor assembly
After finishing the driver motor parts of the mechanical system, I need to
design and modify two pulleys which are attached to the rear wheels. No waste of
energy is needed to move the voice command wheelchair. So, the hand rims were
removed from the rear wheel. The larger driven pulleys were taken from the
compressor of a vehicle air conditioner system, which is hollow and have a large
hole in the middle. A bush is needed to fill the hole and attach it onto the rear wheel.
The bush designed is shown in Figure 3.16, which is produced using lathe machine.
36
Figure 3.16: The CATIA drawing for the bush
The outer diameter was used to connect the driven pulley and for the smallest
diameter of the bush is attached to the rear wheel. The bush was welded onto the
wheel using arc welding. The hole in the middle is to place the shaft. The original
shaft was short and was unable to be used in this project. A new shaft was produced
by using lathe machine. The thickness of the driven pulley had to be reduced by
4mm to make it possible for it to rotate. The finished products are shown in figure
3.17.
Figure 3.17: The finished parts using lathe machine
The bush is welded on the outer part of the wheel using arc welding. This is
to avoid the bearing grease being spoiled by the high temperature during welding
operation. Figure 3.18 shows the welding process of the bush onto the wheel.
37
Figure 3.18: The bush was welded to wheel
Then the welded bush was assembled into the hole of the driven pulley. The
hole was fixed to the pulley and no welding is needed to join the part.
Figure 3.19: The plate assembly with big pulley
Figure 3.19 shows the assembly of the bush joining the wheel with the driven
pulley. Finally, a belt is used to connect the pulley system and adjustments are done
to make sure the belt was tight enough to move the rear wheel when the motor was
given a power supply. The electrical wheelchair system without electronic circuit is
shown in figure 3.20. The wood plate was used to place the battery and electronic
circuit box.
38
Figure 3.20: The electrical wheelchair
After assembly mechanical and electronic part, the completed voice
controlled electrical wheelchair was shown at figure 3.21.
Figure 3.21: The completed voice controlled wheelchair
When a user wants to use the wheelchair, voice of the user is needed to train
into the speech recognition circuit. The word needed to train must be according to the
table 3.0. This is because the wheelchair system now already designed by that
conditions. If not following the sequence, an error will be occurred. After training the
command, speak to the microphone for the command which was user wanted. The
Electronic
circuit box
Battery
Motor
Bracket,
motor and
drive
pulley
Bush and
Driven
pulley
Belt
Microphone
39
circuit will control the direction and speed of the motor. Motor will rotate and driver
pulley will pull the driven pulley. Finally, the rear wheel of the wheelchair will move
with the command that had been given. Each command has been tested and analyzed.
The results of the analysis are tabled in chapter five.
CHAPTER 4
RESULT AND ANALYSIS
After the design and development parts are completed, some testing and
analysis are done. This includes testing on the accuracy of the system and wheelchair
velocity.
4.1 Accuracy for Speech Recognition Circuit
Condition 1: silent area
This experiment was conducted in a room which is in quiet condition to affect
the result of the experiment. Experiment purpose is to find out the accuracy of the
HM 2007 speech recognition circuit in different conditions. The things that need to
ready are microphone, SR-07 speech recognition circuit and paper to write the result.
Five trials were done to the circuit base on the commands listed at the table 4.1.
1 2 3 4 5 Total
Fast 1 1 0 1 1 4
Slow 1 1 1 1 0 4
Forward 1 1 1 1 1 5
Reverse 1 1 1 1 0 4
Left 1 1 1 1 1 5
Right 1 0 1 1 1 4
Table 4.1: the result in silent area
Trial
Commands
41
From table 4.0, there are 26 over 30 commands recognized by the SR-07
speech recognition circuit. The percentage of the accuracy of SR-07 speech
recognition circuit in silent condition is 86.67%. Calculation for percentage is shown
as below.
Accuracy = 26/30 X100%
= 86.67%
Condition 2: noisy area
The testing is done outside of the quiet room where it is considered as natural
environment. From this testing, the results are as table 4.1.
1 2 3 4 5 Total
Fast 0 1 1 1 0 3
Slow 0 0 1 1 0 2
Forward 1 1 1 0 1 4
Reverse 1 1 0 1 0 3
Left 1 0 1 0 0 2
Right 0 0 1 1 1 3
Table 4.2: the result in noisy area
From table 4.2, there are 17 over 30 commands recognized by the SR-07
speech recognition circuit. The percentage of the accuracy of SR-07 speech
recognition circuit in silent condition is 56.67%. Calculation for percentage is shown
as below.
Accuracy = 17/30 X100%
= 56.67%
Trial
Commands
42
Accuracy of the SR-07 circuit
0
1
2
3
4
5
6
Fast Slow Forward Reverse Left Right
Commands
Qu
an
tity
Silent condition Noisy condition
Figure 4.0:Graph of accuracy of the SR-07 in two conditions
From the graph result, we can find out that the SR-07 speech recognition
circuit accuracy is less when assign the commands in the noisy area. That means the
voice controlled wheelchair system has less control when in the noisy condition.
4.2 Velocity
There is important to find out the velocity of the wheelchair system. The
experiment conducted by using the ruler and time watch. Voice controlled
wheelchair moved in a straight line then the distance and time was taken. There are
two conditions of velocity need to take in the experiment. Firstly, the velocity of the
unload condition. The wheelchair will let it go in a straight line and the result was
taken. The distance has been measure was six meter and time is 6.34s. So, distance
over the time is 0.95m/s. Secondly, a person has weight around 40kg to 45kg was sat
43
at the wheelchair. The voice controlled wheelchair also let it move in a straight line.
The distance has been measure was four meter and time is 5.44s. Calculation of
distance over time is 0.74m/s.
Based on the result above, the velocity of voice controlled wheelchair is
affected by the load. That’s mean the velocity of wheelchair system will decrease
proportional to the load that is carry by the system.
CHAPTER 5
CONCLUSION AND RECOMMENDATIONS FOR
FUTURE WORKS
5.1 Conclusion
As a conclusion, the objectives for this project were covered and achieved.
This is done by implementing voice recognition processor HM2007 chip for
acquiring and distinguishing the command for controlling the motion of a
wheelchair. The speed and direction of the wheelchair now can be selected using the
specified commands. Thus the only thing needed to ride the wheelchair is to have
voice. Beside that, the development of this project is done with less cost and
affordable.
The design not only reduce the manufacture cost compare with present
market one but also will give great competitive with other types of electrical
wheelchair. However there are some improvements should be done to make it more
reliable. This is outlined in the recommendation part.
By improving this system, we directly enhance the life style of the disable
people in the community. Lastly, we hope that this kind of system could contribute to
the evolution of the wheelchair technology.
45
5.2 Recommendations
This project still has many improvements that should be done to improve its
accuracy and reliability. There are some suggestions for the future research and
development.
i. Adding the signal conditioning part which is consisting of a filter circuit.
In signal processing, the function of a filter is to remove unwanted parts
of the signal, such as random noise, or to extract useful parts of the signal,
such as the components lying within a certain frequency range
ii. To apply sensors for security purpose. There so many types of sensors
are available. However, many researches and testing with different
algorithms have to be done in order to make it successful.
iii. Designing a controller to control the front wheels so that they will be self
centered each time the wheelchair stops.
iv. Lastly, the circuit to control the motor can be replaced by a
microcontroller for example peripheral interface controller (PIC). By
using the microcontroller the speed and direction of the wheelchair can be
controlled and performed better. So, the speed can be varied
simultaneously without stopping the movement of the wheelchair
46
REFERENCES
A. Fattouh, M.Sahnoun and G. Bourhis. Force Feedback Joystick Control of a
powered Wheelchair: Preliminary Study. University of Metz, France
Ali Kaya, Sidar Ok and Metin Yorulmaz. (2002). Smart Wheelchair. Eylul
University: Final Report.
Cooper, R.A. (2002). Intelligent Control of Power Wheelchair. IEEE Engineering in
Medicine and Biology Magazine. 14: 423-431.
David Spencer Lees. (1994). A Graphical Programming Language For Service
Robots In Semi-Structured Environments: Stanford University: Ph.D. Thesis.
G. Davinder, A. Caleb, C. Michael, H. David, C. Randy.(2003). Remote Access
Trainable System. University Of British Columbia: Final Report.
G.Maiocchi. Driving DC Motor, ST AN 281 Application Note.
Holly A. Yanco and James Gips. Driver Performance Using Single Switch
Scanning With A Powered Wheelchair: Robotic Control Versus Traditional
Control. MIT
Linda Fehr, MS. W. Edwin Langbein, Steven B. Skaar. (2000). Journal of
Rehabilitation Research and Development. Adequacy of power wheelchair
control interfaces for persons with severe disabilities: A clinical survey. 37
(3):353-360.
Inderscience Enterprises Ltd.(2005) Intelligent Systems Technologies and
Applications. 1(1/2): 49.
Pan Min, Sun Yihe. ASIC Design of Gabor Transform for Speech Processing.
Tsinghua university.
Propenta (PTY) LTD. (1999). Wheelchair Survey Project For The National
Department Of Health. South Africa: Propenta (PTY) LTD.
Q.P.Ha,T.H.Tran and G. Dissanayake.(2005). Int. J .Intelligent System Technologies
and Applications, A Wavelet and Neural Network-based voice interface
47
system for wheelchair control. 1: 49 – 64.
Richard Mensik.(2001). Voice Control Enhances Application. EE Times. Czes
Republic.
Simon P. Levine, David A. Bell, Lincoln A. Jaros, Richard C. Simpson, Yoram
Koren, and Johann Borenstein. (1999). The NavChair Assistive Wheelchair
Navigation System. IEEE Transaction on rehabilition Engineering.7(4): 4.
Shafkat Kibriar. (2005). Speech Recognition For Robotic Control. Umea University
Sweeden: Master Thesis.
Richard Simpson, Edmund LoPresti, Steve Hayashi, PhD; Illah Nourbakhsh,David
Miller.(2004). Journal of Rehabilition Research and Development. The Smart
Wheelchair Component System. 41(3B):429 -442.
Peter R. Brown. (2003). Nightside Test Design Ltd. Automating Embedded System
Testing for SHARK Powerchair Controllers. New Zealand.