Acoustic Methods for Robot Communication in Pipes

This research intends to develop communication techniques of using acoustic wave which can be adopted on a moving robot. Communication is important as autonomous robots are being developed to work in partially filled pipes such as sewers and drainage pipes. These robots need to change messages and communicate with the above ground control. The major challenge of sending message acoustically in the pipe are background noise and wave dispersion in the channel. In order to help the communication system overcome these effects the acoustic characteristic of the pipe need to be known. In this work, the frequency response functions (FRFs) of an empty pipe, partially water-filled pipe and pipe with artefacts (lateral connection/blockage/manhole) has been studied analytically and numerically. In addition, a computationally efficient empirical model for predicting the FRF of partially water filled pipe has been proposed in this thesis. The proposed models for empty pipe and partially water-filled pipe have been validated against an experiment. It has been shown that sound attenuation is a key factor that determines the communication range. In this report, the sound attenuation in empty pipe has been studied and maximum data and communication distance for a given acoustic communication technique have been estimated.
Four modulation techniques, binary amplitude/frequency/phase shift keying (2ASK/2FSK/2PSK) and chirp linear frequency modulation (CLFM) have been adapted for in-pipe acoustic communication for robotic system. Their performance in an empty and partially filled pipes has been evaluated numerically and experimentally against signal noise ratio (PER-SNR) and career frequency. The package error rate has been used as the measure of communication quality. The influence of typical artefacts such as blockage, lateral connection and pipe end has been studied numerically. The results suggest the presence of uncertainties in the pipe is likely to have strong impact on some communication techniques. It has been found that CLFM technique has got a relative high noise resistance and suffers least influence from the in-channel uncertainties making it a very competitive solution of in-pipe communication for robotic systems.