Human-Machine Interface-Based Robotic Wheel Chair Control

Human-Machine Interface-Based Robotic Wheel Chair Control

Deepti Kakkar, Ashish Raman
DOI: 10.4018/978-1-7998-7433-1.ch001
OnDemand:
(Individual Chapters)
Available
$33.75
List Price: $37.50
10% Discount:-$3.75
TOTAL SAVINGS: $3.75

Abstract

This chapter presents the P300-based human machine interference (HMI) systems control robotic wheel chair (RWC) prototype in right, left, forward, backward, and stop positions. Four different targets letters are used to elicit the P300 waves, flickering in the low frequency region, by using oddball paradigms and displayed on a liquid crystal display (LCD) screen by Lab-VIEW. After the pre-processing and taking one second time window, feature is extracted by using discrete wavelet transform (DWT). Three different classifiers—two based on ANNs pattern recognition neural network (PRNN) and feed forward neural network (FFNN) and the and other one based on support vector machine (SVM)—are used. Those three techniques are designed and compared with the different accuracies among them.
Chapter Preview
Top

Introduction

Human Machine Interface (HMI) system is unique of wildest rising field of investigation and growing rapidly now a day (Farwell & Donchin, 1988). HMI system is a normal form of communication, requires peripheral nerves and muscles.it is started by user intension (Wolpaw et al., 2002). This intent initiates a process in which certain brain areas are activated, and hence signal are directed via peripheral nerves system, to the corresponding muscles, which in turn executes the movements required for the communication or controlling task (Wolpaw & McFarland, 2003). HMI provide an another way to the natural communication as well as it helps disabled persons to control their activities (Mason & Birch, 2003). Generally two ways: invasive and non- intrusive HMI are commonly used for developing natural communication.

Invasive HMI: Obtrusive BCI gadgets are inserted inside the skull however rest outside the cerebrum as opposed to inside the gray matter (Zhu et al., 2010). Signal quality utilizing this sort of BCI is bit smaller when it looks at to non-obtrusive BCI. They deliver preferred determination signals over non-intrusive BCI. In partially invasive BCIs have less danger of scar tissue development when contrasted with invasive BCI (Zhu et al., 2010). Electrocochleography [ECoG] utilizes the same innovation as non-obtrusive EEG, yet the anodes are inserted in a dainty plastic pad that is set over the cortex, underneath the dura mater (Towle et al., 1993). ECoG advancements were first exchange human in a later trial; the scientists empowered a high school kid to play space intruders utilizing his ECoG insert (Towle et al., 1993). This examination shows that it is hard to create kinematics BCI gadgets with more than one measurement of control utilizing ECoG (Towle et al., 1993). Light responsive imaging BCI devices are still in the area of hypothesis. These would include embedding laser inside the skull. The laser would be prepared on a solitary neuron and the neuron's reflectance measured by a different sensor. At the point when neuron fires, the laser light example furthermore, wavelengths it reflects would change marginally (Towle et al., 1993).

Non-Intrusive HMI: Non- intrusive BCI has the minimum sign clarity with regards to speaking with the cerebrum but it is thought to be most secure at the point when contrasted with different sorts. This kind of gadget has been observed to be effective in giving a patient the capacity to move muscle embeds and reestablish fractional development (Nicolas-Alonso & Gomez-Gil, 2012). A non-obtrusive procedure is one in which therapeutic checking gadgets or sensors are mounted on tops or headbands read cerebrum signals. This methodology is less meddlesome additionally perused flags less viably in light of the fact that cathodes can't be set specifically on the sought part of the mind. A standout amongst the most well-known gadgets under this classification is the EEG equipped for giving a fine fleeting determination. It is anything but difficult to utilize, shabby and portable (Nicolas-Alonso & Gomez-Gil, 2012).

HMI is an un-natural system that by passes body usual affects paths which are the neuromuscular output channels (Cheng et al., 2002). Instead of depending up on peripheral nerves and muscles, HMI straightly processes mind actions associated to the users intensions and translates the brain actions which are recorded, into respective control signals for HMI applications (Erdoğan, 2009). A computer is used to translate signal processing and pattern recognition because the measured actions directly emerged from the brain not from peripheral nerves, this system is called as Human Computer Interface [HCI]. Some of the brain activities that can be successfully noted from the scalp by using EEG are Event Related Potentials (ERPs), Slow Cortical Potentials (SCPs), P300 potentials and Steady-State Visual Evoked Potentials (SSVEPs) (Thulasidas et al., 2006; Zhu et al., 2010). Among them P300 are attracted due to its advantages of requiring less or no training, high information training rate (Towle et al., 1993).

Complete Chapter List

Search this Book:
Reset