DSP Based Voice Operated Wheelchair


ABSTRACT'The proposed idea is to implement voice operated wheelchair for physically disabled people. Commercially available powered wheelchairs are developed for people who have reliable mobility in their arms and body. These drawback overcome in voice operated wheelchair. It mainly consist voice recognition module, controller logic and wheelchair. For speech recognition zero crossing rate (ZCR), standard deviation (SD) and autocorrelation (AC) features are used. Using SIMULINK speech recognition model is developed. For standalone working of system Digital Signal Processor Kit (DSK TMS320C6713) is used. Simulink model is downloaded into DSP with the help of Code Composer Studio (CCS). Controller is used for actuating wheelchair motors. Wheelchair operate on Five Marathi commands 'pudhe', 'mage', 'davikade', 'ujvikade' and 'thamba'.

Keywords 'Zero crossing rate(ZCR), Standard deviation(SD,)Autocorrelation(AC), DSK, CCS, SIMULINK.

INTRODUCTION
A chair with wheels designed as a replacement for walking is known as wheel chair. This is used for movement of physically disabled, elder people, children who have difficulty and are unable to walk [1].
People use wheelchair as a consequence of a variety of disabilities, including spinal cord injury, multiple sclerosis, muscular dystrophy, arthritis, cerebral palsy, polio etc. Such persons face difficulties in their day to day functional activities, work and studies. These people depend on others for their living. But in today's fast world, everyone is busy and there are less people to care for the increasing number of elderly and the physically challenged people.
The traditional wheelchairs available in the market are manually propelled. However many individuals have weak upper limbs or find the manual mode of operating too tiring. Hence it is desirable to provide them with a motorized wheelchair that can be controlled by moving a joystick, keypad, chin-operated joysticks, steering, touchpad etc. Conventional electric powered wheelchairs operated using joystick, keypad and steering have several disadvantages. For persons with limited dexterity, or fine control of the fingers, access to mechanical hardware such as buttons and joysticks can be quite difficult and sometimes painful. For individuals with Traumatic Brain Injury (TBI), Multiple Sclerosis (MS) or Amyotrophic lateral sclerosis(ALS) voluntary control of limb movement maybe substantially limited or completely absent. There is error and variability in movement produced during joystick. With Prolonged use, the joystick and the buttons can become less responsive. Extensive use of joystick can cause repetitive stress injuries and Hand injuries which can lead to inflammation, muscle strain or tissue damage [2].
To overcome this kind of limitations, voice operated wheelchair come into picture. Many voice operated wheelchairs are bulky and need to interface laptop with wheelchair. This limitation of existing voice operated wheelchair is overcome in the proposed system. For standalone working of wheelchair DSP TMS320C6723 kit is helpful. Five commands are recognized using DSK TMS320C6713.For each command different logic levels are generated by DSK. These analog signals are further processed by microcontroller & drives DC motors of wheelchair. Alternate provision for controlling wheelchair is done by keypad.
Most of voice operated wheelchairs available are developed for national/multinational languages but not in regional languages such as Marathi. Many Peoples are not able to speak national/multinational languages with clear pronunciation. So there is need of wheelchair which can be controlled by regional languages (Marathi). Thus proposed wheelchair design is even convenient to uneducated people, thereby minimizing language barrier [2][3][4][5][6][7].
The paper is divided into eight different sections. Section II presents system block diagram of its function. Section III shows system flowchart. Section IV presents hardware and software requirements of proposed work. Section V explains procedure for downloading Simulink model into DSK TMS320C6713. Section VI discusses the results. Section VII is conclusion and section VIII presents future work.

SYSTEM MODEL

A design of sophisticated wheelchair application for physically disabled persons most probably for the people those who have 'speech' as only way of communication.

Figure2. Architecture of grain storage system

Fig.1 System Block Diagram

The microphone, used to convert the voice into analog electrical signal, is directly connected to an analog input of the DSK. Voice recognition module is used for identifying speech commands uttered by user. It consists of SIMULINK model for voice recognition. This model is downloaded into DSP TMS320C6723 kit through Code Composer studio.CCS is used for converting SIMULINK model into 'C' code & this code is downloaded into DSK.DSK generates different analog signals according to input signal. This signal is applied to a controller & peripheral hardware section. In this block analog signal is converted into digital & generate different logic level for controlling wheelchair motors. If given command is 'pudhe' then system will generate respective voltage level according to given command& DC motor will drive wheelchair forward. Same is the case with 'mage', 'davikade', 'ujvikade'& 'thamba' followed by desired action.

System model mainly consist of three block,
Voice recognition module:-

Fig.2 Voice recognition module

Voice recognition is done using SIMULINK. Voice commands are recorded using MATLAB's 'wavrecord' command.
System will operate in two phases-A) Training phase B) Testing phase.
Training phase:
In this phase Marathi commands are recorded through microphone. After recording features are extracted with Zero crossing rate (ZCR) [10][11], standard deviation (SD) [11]& autocorrelation (AC) [11] feature extraction methods. Calculation of threshold values for different pitches of user is done and these threshold values are stored in database. This system design is proposed for single user operation [7][8].
Testing phase:
In the testing phase, it takes real time commands from speaker through microphone. After this features are extracted and these features are compared with threshold values stored in the database. If matched, command is matched then that particular action is followed. [7][8].
The five commands are recognized using SIMULINK. After working on PC, the software is downloaded into DSK through Code Composer Studio.CCS is converts simulink model into 'C' code. When system is ON, the commands are processed in real time by DSP kit TMS320C6713.The Texas Instruments TMS320C6713 DSP kit is used in this work which has a 225 MHz processor with 128MB RAM and on board AIC23 codec allows DSP to transmit & receive analog signal [9]. DSK checks the presence of voice or noise. If the human voice is detected then DSK processes the input voice command and decides what is being uttered. After the decision DSK generates a specific analog signal for the spoken word at its analog output. There are five different analog outputs one each for 'pudhe', 'mage', 'ujvikade', 'davikade' &'thamba'.

Controller & Peripheral Hardware:-
Analog output of DSK is applied to ADC. This signal is further processed & used for driving DC motors of wheelchair. Controller generates different logic level for respective input command. According to logic level wheelchair motors are operating.Here keypad is used as alternate terminology for controlling wheelchair in emergency.

Wheelchair:-
Wheelchair model consist two BLDC motors. For driving this motor driver circuit is used. Four 12v,2.6Ah batteries are connected to supply desired power of motor as well as controlling circuitry.

III.SYSTEM FLOWCHART


Fig.3 System flowchart

IV.Hardware and Software Requirements: [9]

MATLAB R2007a with Embedded Target for TI C6000 and Signal Processing Block set.
Code Composer Studio (CCS) v3.1
Texas Instruments DSK6713 hardware.
Microphone and computer loudspeakers/headphones.

V. Procedure for Downloading Simulink Model into DSK TMS320c6713: [9]

After recognition of commands the main task is to download SIMULINK model into DSK TMS320C6713.

1. Connect the C6713 hardware to a USB port of the computer and turn on the supply to the board. If the board is powered, he green LEDs start flashing on the board during the self-test.
2. Start Code Composer Studio for DSK6713 and use Debug -> Connect
3. In MATLAB2007a develop SIMULINK model using 'Embedded Target for TI C6000' library.
4. Do proper configuration and build project using real time workshop utility of MATLAB.
5. CCS will generate 'c' code for this model.

VI. Results & discussion:
The feature extraction & voice recognition model is developed in SIMULINK. Five Marathi commands are recorded using 'wavrecord' command of MATLAB. From this recorded .wav files zero crossing rate, standard deviation & autocorrelation features are extracted. Threshold values are calculated for all features for all commands & saved in voice recognition model for comparing with newly calculated features of real time command. Following table shows the threshold value for each command.
Table 1 feature value for 'PUDHE', 'MAGE', 'DAVIKADE', 'UJVIKADE'& 'THAMBA'

Table 1: Features.
Features
Zero crossing rate Standard deviation Autocorrelation
Commands
Running sum
Max ZCR count
pudhe Upper threshold 969 75 53 1.4
Lower threshold 600 52 10 0.3
Mage Upper threshold 1200 145 50 1.4
Lower threshold 830 80 19 0.62
Davikade Upper threshold 1300 135 65 1.6
Lower threshold 950 75 40 0.6
Ujvikade Upper threshold 1050 110 25 0.7
Lower threshold 617 65 11 0.2
Thamba Upper threshold 1090 115 56 1.2
Lower threshold 680 65 26 0.4

The accuracy of system is calculated using following formula.
Accuracy(%)=(No.ofcorrectrecognizedcommands)/(Totalno.oftestcommands)*100

Table.2 shows the offline accuracy of each command. For training 15 utterances of each command are used. Five utterances of each command are used for testing.

Table.2 system accuracy
Command Number
of Train utterances Number of Test utterances Correct recognized utterances Accuracy
(%)
Pudhe 15 5 5 100
Mage 15 5 4 80
Davikade 15 5 5 100
Ujvikade 15 5 5 100
Thamba 15 5 4 80

VII. Conclusion

The paper presents voice operated wheelchair using DSP. This system is developed for physically disabled people. Voice operated wheelchair understand five Marathi commands (pudhe, mage, davikade, ujvikade, thamba.).DSK TMS320C6713 recognizes speech commands. Speech recognition model is developed in SIMULINK & through CCS downloaded into DSK.ZCR, SD & AC features are used for differentiating speech commands. System working is standalone & speaker dependant.

VIII. Future Work
This work will extend for speaker independent system also can develop security system for collision detection.

Acknowledgment
I am highly indebted to my guide Dr.S.A.Pardeshi for his valuable guidance. Special thanks to member of Department of Electronics and Telecommunication, RIT, Islampur, Maharashtra, India for their kind co-operation and encouragement. I would like to express my heartfelt thanks to my beloved parents, for their blessings and constant support.

References

[1]Mohan Kumar R., Lohit H. S., Manas Ranjan Mishra, Md. Basheer Ahamed, 'Design of Multipurpose Wheel Chair for Physically Challenged and Elder People',SASTECH journal,volume-11,page-107-117,2012.

[2]Ashwin.S, Sandhya Devi.R.S 'Novel Alternative Approaches for Developing SmartWheelchairs: A Survey' International Journal of Advanced Information Science and Technology (IJAIST),Vol.6, No.6, October 2012.

[3] Mr. Sangmeshwar S. Kendre et al,'Voice Activated Multiprocessor Embedded System To Improve The Control Of A Motorized Wheelchair' International Journal of Engineering Science and Technology Vol. 2(11), 2010,paged-6812-6818.

[4] G. Pacnik, K. Benkic and B. Brecko, ' Voice Operated Intelligent Wheelchair ' VOIC', IEEE ISIE 2005, June 20-23, Pp1221- 1226.

[5] Omair Babri Et Al, 'Voice Controlled Motorized Wheelchair With Real Timeobstacle Avoidance'ICCIT ,2012 ,pages-724-728.

[6] Joelle PINEAU et al ' On the Feasibility of Using a Standardized Test for Evaluating a Speech-Controlled Smart Wheelchair' Manuscript received December 10, 2010; revised April 11, 2011.

[7] Muhammad Tahir Qadri and Syed Ashfaque Ahmed, 'Voice controlled wheelchair using DSK TMS320C6711',2009 International Conference on Signal Acquisition and Processing, Pp-217-220.

[8]Deepali Y Loni 'DSP Based Speech Operated Home Appliances Using Zero Crossing Features' Signal Processing: An International Journal (SPIJ), Volume (6), pages- 45-55, 2012.

[9]Texas instruments 'TMS320C6713 user manual'.

[10]R.G. Bachu, S. Kopparthi, B. Adapa, B.D. Barkana, "Voiced/Unvoiced Decision for Speech Signals Based on Zero-Crossing Rate and Energy," IEEE International Joint Conferences on Computer, Information, and Systems Sciences, and Engineering (CISSE'08) pages-1-7,2008.

[11] Madiha J alil, Faran Awais Butt & Ahmed Malik' 'Short-Time Energy, Magnitude, Zero Crossing Rate and Autocorrelation Measurement for Discriminating Voiced and Unvoiced segments of Speech Signals' IEEE International Conference, pages-208-212,2013.

Source: ChinaStones - http://china-stones.info/free-essays/engineering/dsp-based-voice-operated-wheelchair.php



About this resource

This Engineering essay was submitted to us by a student in order to help you with your studies.


Search our content:


  • Download this page
  • Print this page
  • Search again

  • Word count:

    This page has approximately words.


    Share:


    Cite:

    If you use part of this page in your own work, you need to provide a citation, as follows:

    ChinaStones, DSP Based Voice Operated Wheelchair. Available from: <https://china-stones.info/free-essays/engineering/dsp-based-voice-operated-wheelchair.php> [18-06-19].


    More information:

    If you are the original author of this content and no longer wish to have it published on our website then please click on the link below to request removal:


     
    там

    www.ry-diplomer.com/kupit-nastoyashchij-diplom

    ссылка