Synchronization and System Identification Using Symbolic Dynamics
Date
2018-01-22
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Research on communication, signal processing and robotics have benefited from chaos theory. In robotics, chaos is applied in path planning and coverage, which can be improved by employing multiple robots working concurrently. However, collaborating multiple chaotic robots is difficult since chaos synchronization is difficult to realize in the presence of noise. In signal processing, chaos is applied in system identification problem. But existing chaos-based identification techniques impose limitations caused by not exploiting characteristics unique to the input information.
The first problem considered in this thesis is collaborative exploration among autonomous mobile robots using chaos. To achieve this, the chaotic mobile robots are synchronized by chaos synchronization that can control multiple robots. However, the critical issue of noise introduces instabilities in motion thereby hindering collaboration. This work proposes a novel technique
which uses symbols to achieve collaboration between chaotic robots in presence of noise.
The second problem considered is system identification without any knowledge of input, known as blind identification. Performance of existing blind identification methods degrades at strong noise. This problem is addressed in this work by using a chaos representation of the random symbolic input. The blind identification approach using chaos in this work is analytically proved to achieve the optimal performance bound of non-blind using non-chaotic input. Results show superior performance in comparison to existing methods.
The third problem considered is blind system identification when the input is a chaotic signal. An estimator for a chaotic property is derived which is used as the optimization criteria for identification. Results show the merit of chaos-based system identification over popular techniques. The outcome of the work presented in this thesis is shown to work well when applied to symbolic signal processing using chaos and can be used for harnessing noise for useful engineering purposes such as multi-robot collaborative exploration using autonomous chaotic robots.
Description
Keywords
symbolic dynamics, system identification, chaos, chaotic system, chaotic signal, communication, parameter estimation
Citation
Mukhopadhyay, S. (2018). Synchronization and System Identification Using Symbolic Dynamics (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca.