Identification of Mobile Robot Using Ultrasonic Waves

Robots play a major role in present world to minimize human labour and to accomplish complicated/dangerous tasks. The pre-requisite for a motilerobot to perform its activities in a given environment is for it to know its current location, which would help it to know the degree and distance to move to perform its tasks., Different techniques have been suggested over the years to solve the problem of the Robot knowing its own location. This document presents a survey on the different approaches and implementations proposed by various research scientists. In this paper, the Ultrasonic sensors are used for distance measurement and the Speedometer for knowing the speed in which the Robot travels. Low Pass Filter is used for error reduction in measurements and to provide accurate localization of posture and orientation. The overall process of measuring based on the sensor inputs and guiding the Robot to the desired position is performed by a PIC microcontroller (Model: PIC16F877A). The Ultrasonic Hybrid Localization algorithm proposed in an earlier work, allows for up to 500 measurements beyond which the error level is significant. The oversampling algorithm proposed in this work allows much more measurements to be taken to locate a robot in motion more accurately. Extended Kalman Filter is based on time domain, while Low Pass Filter is based on Frequency domain, which helps in covering greater distance and measurements to accurately track the posture of the motile robot. The proposed solution is designed to be more cost effective but also provides reliable solution to the problem of Mobile Robot localization.