Abstract: This paper presents an automated voice recognized wheelchair (AVRW) based on speech recognition. It’s common that physically challenged people find it hard to move the wheelchair without external aid so this AVRW can be used by them to move around without any difficulty. The voice of the person is detected by voice capture module which will be compared by the voice recognition module with predefined voices loaded into the system. Vector Quantization technique is used to recognize the command. A microphone is used for taking code words from the patient to the PC. The signals from the PC are processed into code words. These code words cannot be directly sent to the PIC Microcontroller, so a ZigBee module is used for transmission and reception. According to the codes, the motors are rotated. These codes are compared with the MATLAB codes, if they match with each other, then the movement of motor occurs. There is also a motor driver for the control of the motor. It is also equipped with obstacle avoidance technique to help when the person may not be able to provide proper voices at the right time.

Keywords: Speech recognition, MATLAB, Vector Quantization, ZigBee, PIC Microcontroller.