No description available.

This research presents a complete review of signal processing techniques used, today, in vibration based industrial condition monitoring and diagnostics. It also introduces two novel techniques to this field, namely: the Kolmogorov-Smirnov test and Volterra series, which have not yet been applied to vibration based condition monitoring. The first technique, the Kolmogorov-Smirnov test, relies on a statistical comparison of the cumulative probability distribution functions (CDF) from two time series. It must be emphasised that this is not a moment technique, and it uses the whole CDF, in the comparison process. The second tool suggested in this research is the Volterra series. This is a non-linear signal processing technique, which can be used to model a time series. The parameters of this model are used for condition monitoring applications. Finally, this work also presents a comprehensive comparative study between these new methods and the existing techniques. This study is based on results from numerical and experimental applications of each technique here discussed. The concluding remarks include suggestions on how the novel techniques proposed here can be improved.

No description available.

Tire-pavement interaction noise is one of the dominant sources of vehicle noise, and one of the most significant sources of urban noise pollution. One critical generation mechanism of tire-pavement interaction noise is tire tread excitation. The tire tread contributes to the tire-pavement interaction noise mainly through two mechanisms: (1) tread block impact, and (2) the compression and expansion of the air in the tread groove at the contact patch. The tread pattern is the critical part of the tire design since it can be easily modified. Hence, the main focus of this study is to quantify the tread pattern contribution in total tire-pavement interaction noise. To achieve this goal, the noise produced by the tread pattern is separated from the total tire-pavement interaction noise. Since the tread pattern excitation is periodic with tire rotation, the noise produced by the tread is assumed to be related to the tire rotation. Hence, the order domain synchronous averaging method is used in this study to separate and quantify the tread pattern contribution to the total tire-pavement interaction noise. The experiment has been carried out using an On-Board-Sound-Intensity (OBSI) system. Five tires were tested including the Standard Reference Test Tire (SRTT). Compared to the conventional OBSI system, an optical sensor was added to the system to monitor the tire rotation. The once per revolution signal provided by the optical sensor is used to identify the noise signals associate to each revolution. In addition to the averaging method using optical signals, other data processing techniques have been investigated for separating the tread-pattern noise without utilizing the once per revolution signal. These techniques are autocorrelation analysis, a frequency domain filter, principal component analysis, and independent component analysis. In the tread-pattern noise generation, the tread profile is the most important input parameter. To characterize the tread profile, the tread pattern spectral content and air volume velocity spectral content for all the five tires are computed. Then, the tread pattern spectrum and the air volume velocity spectrum are both correlated with the separated tread-pattern noise by visual inspection of the spectra shape. / Master of Science

In the field of ergonomics, biomechanics, sports and rehabilitation muscle fatigue is regarded as an important aspect since muscle fatigue is considered to be one of the main reasons for musculoskeletal disorders. Classical signal processing techniques used to understand muscle behavior are mainly based on spectral based parameters estimation, and mostly applied during static contraction and the signal must be stationary within the analysis window; otherwise, the resulting spectrum will make little physical sense. Furthermore, the shape and size of the analysis window also directly affect the spectral estimation. But fatigue analysis in dynamic conditions is of utmost requirement because of its daily life applicability. It is really difficult to consistently find the muscle fatigue during dynamic contraction due to the inherent non-stationary nature and associated noise in the signal along with complex physiological changes in muscles. Nowadays, in addition to linear signal processing, different non-linear signal processing techniques are adopted to find out the consistent and robust indicator for muscle fatigue under dynamic condition considering the high degree of non-linearity (caused by functional interference between different muscles, changes of signal sources and paths to recording electrodes, variable electrode interface etc.) in the signal. In this work, various linear and nonlinear-non-stationary signal processing methods, applied on surface EMG signal for muscular fatigue analysis under dynamic contraction are studied. In present study, surface EMG (sEMG) signals are recorded from Biceps Brachii muscles from eight (N=8) physically active college students during dynamic lifting 7 kg load at the rate of 20 lifts/min till they become fatigue. EMG data is processed in two ways -1. taking the whole EMG response and 2. breaking into three ranges of contraction (0-45)o, (45-90)o and >90o, to study better response region. It is observed that in spectral estimation techniques auto-regressive (AR) based spectral estimation technique gives better frequency resolution than periodogram for small epochs, as AR is based on parametric estimation. Both the previous methods provide only the frequency information in the signal. In order to estimate the time varying nature of frequency content in a signal various time-frequency signal processing techniques are used like – Short Time-Fourier Transform (STFT), Smoothed pseudo Wigner-Ville (SPWD), Choi-William distribution (CWD), Continuous Wavelet Transform (CWT), Huang-Hilbert Transform (HHT) and Recurrence Quantification Analysis (RQA) are used. The last two techniques are used by considering the EMG signal as non-linear and non-stationary signals. Among these techniques, STFT is the simplest time-frequency analysis technique. But tradeoff between time and frequency resolution is the major constraint in STFT, therefore, a window length of 256 samples are considered in this study. In order to tackle time-frequency resolution problem different Cohen-class distribution techniques are used like SPWD and CWD, where the result is severely affected by the presence of interference terms which make its interpretation really difficult. Different adaptive filters are used in SPWD and CWD to suppress these interference terms during analysis. Among these time-frequency analysis techniques continuous wavelet transform provides the most accurate results in comparison to other time-frequency analysis techniques. Similar result is obtained in present study. This fatigue response is further improved using non-linear and non-stationary techniques like HHT and RQA. HHT shows less variation in frequency response than CWT analysis result. Percentage of determinism calculated using recurrence quantification analysis method is found to be more sensitive than mean frequency estimation. Therefore, non-linear and non-stationary signal processing techniques are to be better indicator of muscle fatigue during dynamic contraction.

Protein sequencing has produced overwhelming amount of protein sequences, especially in the last decade. Nevertheless, the majority of the proteins' functional and structural classes are still unknown, and experimental methods currently used to determine these properties are very expensive, laborious and time consuming. Therefore, automated computational methods are urgently required to accurately and reliably predict functional and structural classes of the proteins. Several bioinformatics methods have been developed to determine such properties of the proteins directly from their sequence information. Such methods that involve signal processing methods have recently become popular in the bioinformatics area and been investigated for the analysis of DNA and protein sequences and shown to be useful and generally help better characterise the sequences. However, there are various technical issues that need to be addressed in order to overcome problems associated with the signal processing methods for the analysis of the proteins sequences. Amino acid indices that are used to transform the protein sequences into signals have various applications and can represent diverse features of the protein sequences and amino acids. As the majority of indices have similar features, this project proposes a new set of computationally derived indices that better represent the original group of indices. A study is also carried out that resulted in finding a unique and universal set of best discriminating amino acid indices for the characterisation of allergenic proteins. This analysis extracts features directly from the protein sequences by using Discrete Fourier Transform (DFT) to build a classification model based on Support Vector Machines (SVM) for the allergenic proteins. The proposed predictive model yields a higher and more reliable accuracy than those of the existing methods. A new method is proposed for performing a multiple sequence alignment. For this method, DFT-based method is used to construct a new distance matrix in combination with multiple amino acid indices that were used to encode protein sequences into numerical sequences. Additionally, a new type of substitution matrix is proposed where the physicochemical similarities between any given amino acids is calculated. These similarities were calculated based on the 25 amino acids indices selected, where each one represents a unique biological protein feature. The proposed multiple sequence alignment method yields a better and more reliable alignment than the existing methods. In order to evaluate complex information that is generated as a result of DFT, Complex Informational Spectrum Analysis (CISA) is developed and presented. As the results show, when protein classes present similarities or differences according to the Common Frequency Peak (CFP) in specific amino acid indices, then it is probable that these classes are related to the protein feature that the specific amino acid represents. By using only the absolute spectrum in the analysis of protein sequences using the informational spectrum analysis is proven to be insufficient, as biologically related features can appear individually either in the real or the imaginary spectrum. This is successfully demonstrated over the analysis of influenza neuraminidase protein sequences. Upon identification of a new protein, it is important to single out amino acid responsible for the structural and functional classification of the protein, as well as the amino acids contributing to the protein's specific biological characterisation. In this work, a novel approach is presented to identify and quantify the relationship between individual amino acids and the protein. This is successfully demonstrated over the analysis of influenza neuraminidase protein sequences. Characterisation and identification problem of the Influenza A virus protein sequences is tackled through a Subgroup Discovery (SD) algorithm, which can provide ancillary knowledge to the experts. The main objective of the case study was to derive interpretable knowledge for the influenza A virus problem and to consequently better describe the relationships between subtypes of this virus. Finally, by using DFT-based sequence-driven features a Support Vector Machine (SVM)-based classification model was built and tested, that yields higher predictive accuracy than that of SD. The methods developed and presented in this study yield promising results and can be easily applied to proteomic fields.

The rate at which a railway track deteriorates depends on the response of the track under different static and repeated dynamic forces. These wheel/rail forces lead to imperfections in the rail surface and deviation in track geometry alignment. The wheel/rail forces are dependent upon the quality of maintenance of the characteristics of both train and track. If train components such as wheelsets and suspensions are maintained to a high standard, less dynamic forces are generated at the wheel/rail interface and less damage is caused over time. Therefore, the amount and cost of maintenance of track are reduced. However, there is little known about how the characteristics of train components affect time-dependent track degradation. Track degradation through deviation of track from its ideal position has the most effect on maintenance costs. Therefore, the present research aims to investigate this track degradation and improve understanding of the effects of train characteristics (such as train mass and speed, suspension stiffness and damping) on railway tracks. The research is conducted by looking into the relationship between wheel/rail forces and track degradation on one hand and between wheel/rail forces and train characteristics on the other hand, with the objective of assisting in managing vehicle/track interaction in order to minimise track degradation. This aim is achieved by investigating the above two relationships to attain the desired relationship between track degradation and train characteristics. The research focuses on wheel/rail vertical forces (both amplitudes and frequencies), vertical track alignment (longitudinal vertical profile), and rail head defects. The study started by collecting wheel/rail vertical forces data in addition to data on vertical track degradation under sustained traffic loads on a heavy haul railway section of track in Central Queensland. Also, five years of degradation and maintenance history data were collected on three other test sections of railway track under variety of traffic conditions and loads in Central Queensland. There were four main analyses of this data employed to probe the study. The first analysis was performed by examining the track degradation history data. The standard deviation method was used in this first analysis to acquire the rate of deterioration in terms of its relationship to track profile (roughness). The second analysis was accomplished by correlating the vertical wheel/rail forces to both vertical track profile and rail roughness using signal processing principles and a function know as coherence. The third analysis was carried out by using the computer simulation software NUCARS to obtain the link between wheel/rail forces and the deterioration of the vertical track profile. The fourth analysis was achieved by combining the results obtained from the above three analyses to acquire the rate of track deterioration in terms of its relationship to varying train characteristics. The first analysis mentioned above quantified the relationship between the level of roughness of the track and rate at which that roughness deteriorated. An important outcome of this relationship is that there is a threshold of roughness below which track deterioration is minimal. The track maintenance planners can now use that threshold for cost effective targeting of tamping activities. The correlation study between track roughness and wheel/rail forces using the coherence function found, surprisingly, that the overall deterioration of the track roughness, in the absence of frequencies of forces above 30 Hz, is due to the so-called quasi-static lower frequency oscillations of dynamic forces. This conclusion together with the relationship between vehicle characteristics and track forces, established in the analyses above, has significant implications for the design of wagon bogies and for charges track owners might levy on trains using their tracks. This research is part of a larger Rail CRC project 11/4 called 'Enhancing the Optimisation of Maintenance/Renewal' being carried out in the School of Urban Development in Queensland University of Technology.