No description available.

A range of signal processing techniques have been adopted and developed as a methodology which can be used in developing an intelligent surface electromyography (SEMG) signal classifier. An intelligent SEMG signal classifier would be used for recognising and treatment of musculoskeletal pain and some neurological disorders by physiotherapists and occupational therapists. SEMG signals displays the electrical activity from a skeletal muscle which is detected by placing surface electrodes placed on the skin over the muscle. The key factors of this research were the investigation into digital signal processing using various analysis schemes and the use of the Artificial Neural Network (ANN) for signal classification of normal muscle activity. The analysis schemes explored for the feature extraction of the signals were the Fast Fourier Transform (FFT), Short Time Fourier Transform (STFT), Continuous Wavelet Transform (CWT), Discrete Wavelet Transform (DWT) and Discrete Wavelet Packet Transform (DWPT).Traditional analysis methods such as FFT could not be used alone, because muscle diagnosis requires time-based information. CWT, which was selected as the most suitable for this research, includes time-based information as well as scales, and can be converted into frequencies, making muscle diagnosis easier. CWT produces a scalogram plot along with its corresponding frequency-time based spectrum plot. Using both of these plots, overviewed extracted features of the dominant frequencies and the related scales can be selected for inputs to train and validate an ANN. The purpose of this research is to classify (SEMG) signals for normal muscle activity using different extracted features in an ANN. The extracted features of the SEMG signals used in this research using CWT were the mean and median frequencies of the average power spectrum and the RMS values at scales 8, 16, 32, 64 and 128. SEMG signals were obtained for a 10 second period, sampled at 2048 Hz and digitally filtered using a Butterworth band pass filter (5 to 500 Hz, 4th order). They were collected from normal vastus lateralis and vastus medialis muscles of both legs from 45 male subjects at 25%, 50%, and 75% of their Maximum Voluntary Isometric Contraction (MVIC) force of the quadriceps. The ANN is a computer program which acts like brain neurons, recognises, learns data and produces a model of that data. The model of that data becomes the target output of an ANN. Using the first 35 male subjects' data sets of extracted features, the ANN was trained and then validated with the last 10 male subjects' data sets of the untrained extracted features. The results showed how accurate the untrained data were classified as normal muscle activity. This methodology of using CWT for extracting features for analysing and classifying by an ANN for SEMG signals has shown to be sound and successful for the basis implementation in developing an intelligent SEMG signal classifier.

Thesis made openly available per email from author, August 2015.

published_or_final_version / Electrical and Electronic Engineering / Doctoral / Doctor of Philosophy

published_or_final_version / Electrical and Electronic Engineering / Master / Master of Philosophy

published_or_final_version / abstract / Electrical and Electronic Engineering / Master / Master of Philosophy

published_or_final_version / Electrical and Electronic Engineering / Doctoral / Doctor of Philosophy

No description available.

A thesis submitted to the Faculty of Engineering, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Doctor of Philosophy Johannesburg, October 1997 / This thesis develops and investigates signal processing models that are useful both for interpreting and implementing certain types of Model Predictive Control. Two types of Model Predictive Control are investigated, namely, techniques based on Internal Model Control and Long Range Predictive Control. [Abbreviated abstract. Open document to view full version].

A dissertation submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in fulfilment of requirements for the degree of Master of Science. Johannesburg, 2016. / The l1-norm solution to an under-determined system of linear equations y = Ax is the sparsest solution to the system. In digital signal processing this mathematical problem is known as compressive sensing. Compressive sensing provides a mathematical framework for sampling and reconstructing an analogue signal at a rate far lower than the rate provided by the standard information theory. The reconstruction from few samples is possible using non-linear optimization algorithms provided that the signal is sparse and the sensing matrix in incoherent. The major algorithmic challenge in compressive sensing is to efficiently and effectively find sparse solutions from minimal measurements. General purpose optimization algorithms are not suitable for solving non-differentiable l1-minimization problem. In this dissertation, we survey the major practical algorithms for nding l1-norm solution of under-determined linear system of equations. Specific attention is paid to computational issues, in which individual methods tends to perform well. We propose a hybrid algorithm that combines complementary strengths of the fixed-point method and the interior-point method. The strong feature of the xed-point method is its speed, while the strength of the interior-point method is accuracy. The hybrid algorithm combine the two methods in a probabilistic manner. The algorithm tends to prioritise a method that is efficient and robust. The computational performance of the hybrid algorithm is tested on simple signal reconstruction problems. The hybrid algorithm is shown to produce similar recoverability of sparse solution as that of the xed-point method and the interior-point method. Furthermore the proposed hybrid algorithm is comparative in terms of speed and accuracy with existing methods. / LG2017