Multidirectional electrical wheelchair controller design using electrical signals from shoulder muscles.

摘要

A system was designed to demonstrate the feasibility of a portable, low cost, multidirectional electric wheelchair control unit using electromyographic (EMG) signals collected from an array of arm muscles. Signal collection was achieved using non-invasive capacitive sensors attached to the skin directly above muscles of interest. The system was first implemented for proof of concept using an IBM PC compatible computer in conjunction with a 12bit A/D internal board, and then transferred over to a dual PIC 16C74A MCU processing unit (PU). The signals are conditioned through differential amplification and switched capacitor filtering and then made available to the PU through digitization by an analog to digital conversion block. The wheelchair is ultimately controlled by CMOS logic level signals generated by the PU. A signal processing and analysis algorithm was implemented in software using 'C++' for the PC and Microchip{dollar}sp{lcub}rm TM{rcub}{dollar} Assembly Language for the PIC based PU. This algorithm was used to pair up opposite force generating muscles and compare their average energy content over a predetermined sampling period. Based on the comparison results a motion right, left, back or forth/no motion decision was made. Muscle signal energy content use proved to be an effective and efficient method of obtaining the desired results.