Second Generation Wavelet Transform Research Articles

Deep second generation wavelet autoencoders based on curvelet pooling for brain pathology classification

Neurological disorders represent anomalies relevant to the human nervous system. They also contain biochemical, anatomical and electrical modifications in the central nervous system, the spinal cord and the brain. These disorders provoke different symptoms. Early diagnosis of such changes is necessary for treatment, to limit disease progression. An accurate and CAD system is introduced in this paper to classify brain magnetic resonance images, which overcomes crucial problems in pattern classification, such as extracting certain features in the training phase. Our contribution is to merge a Second-Generation Wavelet Transform Network (SGWTN) and deep learning architectures, hence suggesting novel supervised feature extraction approaches for pattern classification. Our novel architecture allows the classification of all classes of our dataset by the reconstruction of a deep stacked SGWTN-Autoencoder (SGWTN-AE). Combining Curvelet Pooling (CP) with the Adam gradient calculation method can improve their accuracy. We implement CP with Adam in this work using the Haar curvelet (CurvPool-AH) and the Shannon Curvelet (CurvPool-AS). This Network is obtained after a series of SGWTN-AEs followed by a Softmax classifier at the last layer. We find that CurvPool performs fairly well on all datasets. However, overfitting is an issue with all the approaches considered. CP has the potential to outperform current approaches, particularly when merged with an adaptive gradient and curvelets chosen specifically for the data. Our architecture is tested with various image datasets. These latter are as follows: DS-66, DS-90, DS-160, and DS-255. Based on 5×5 cross-validation, our suggested approach can outperform the state-of-the-art methods in terms of classification accuracy.

Leak Localization Method of Jet Fuel Pipeline Based on Second-Generation Wavelet Transform and Short-Time Energy Time Delay Estimation

There is a large error in the acoustic propagation speed of jet fuel pipeline based on cross correlation analysis, which leads to poor positioning accuracy of the leak location. Therefore, a new leak localization method of jet fuel pipeline based on second-generation wavelet (SGW) transform and short-time energy time delay estimation is proposed. Using SGW transform to extract the leak acoustic signal to reduce the interference of external noise in this method, the short-time energy function is estimated to calculate the sudden change time range of leak acoustic signal in the time window, so as to find the sudden change time range containing the sudden change point information of leak acoustic signal. According to the threshold of short-time energy function, the accurate sudden change time of leak acoustic signal is determined. When the accurate sudden change times of three acoustic sensors are obtained, the leak position of pipeline is calculated by a leak localization formula. The proposed method does not need to estimate the acoustic propagation speed, and it can obtain more accurate leak localization accuracy of pipeline leak. The field experimental pipeline results show that the average error rate of jet fuel pipeline leak localization is reduced from 1.25% of the cross correlation analysis method to 0.1% of the proposed method.

A Robust Passive Islanding Detection Technique With Zero-Non-Detection Zone for Inverter-Interfaced Distributed Generation

In this paper, a new and efficient passive islanding detection technique for a grid-connected hybrid distributed generating system has been proposed using Second Generation Wavelet Transform (SGWT) and Maximum Overlap Discrete Wavelet Transform (MODWT). In this technique, the ripple content of the voltage signal is measured at the Point of Common Coupling (PCC) and it is monitored through the decomposition of the signal by applying SGWT and MODWT technique up to suitable finer levels. This algorithm able to detects all kinds of islanding at nearly 0.3 sec, even with zero power mismatch. It is verified on a wide range of operational conditions and is also validated in various non-islanding conditions. Furthermore, the proposed technique is simple, there is no need for a classifier requirement, not depends on system parameters, zero non-detection zone (NDZ), and its operation is independent of the capacity and nature of distributed generation (DG) connected to the utility grid.

A novel genetic algorithm with CDF5/3 filter-based lifting scheme for optimal sensor placement

The generic algorithm has been receiving significant attention due to the node placement problem in the field of sensor application in terms of machine learning. Sensor deployment is able to provide maximum coverage and maximum connectivity with less energy consumption to sustain the network lifetime. The maximum quality coverage problem has been solved successfully by an evolutionary algorithm while placing nodes in optimal position. In evolutionary algorithms, genetic algorithm (GA) plays an important technique for deploying the sensor in the form of population matrix. However, the existing techniques are unable to place sensor position perfectly. In this paper, a novel genetic algorithm with second generation wavelet transform (SGWT) is proposed for identifying optimal potential position for node placement. In order to improve the quality of population matrix, bi-orthogonal Cohen-Daubechies-Feauveau wavelet (CDF 5/3) has been employed. The proposed method is performed primarily to generate sensor position with different populations. Subsequently, it can extend to CDF5/3 filter-based lifting scheme to adjust the sensor position. The proposed method has been compared with random deployment, genetic algorithm and GA with CDF5/3 wavelets in terms of target to cover by the sensor. The result of the proposed method affirms better optimisation as compared to the state-of-art techniques.

Application of Random Forest and Hidden Markov Models for Automatic and Fast Classification of Power Quality Signals

In this paper, wavelet transform, namely the maximal overlap discrete Wavelet Transform (MODWT) and the second generation Wavelet Transform (SGWT) have been implemented. These wavelet transforms are applied to get selected features of the signals. Features are used as inputs to two types of classifiers namely, Hidden Markov Model (HMM) classifiers and the Random Forest (RF) classifier in the both absence and presence of Noise to evaluate the efficiency. The classification accuracy (CA) calculated using these classifiers clearly shows that the RF classifiers is a better classifier then the HMM classifier as it possess higher recognition rate at all levels of noise along with the pure PQ signals. Another important property of RF classifier is the proper classification of large number of class of both slow and the fast disturbances.

Fast Methods for Power Quality Analysis

AbstractIn this paper, the lifting based second generation Wavelet Transform (SGWT) is implemented along with the widely used discrete Wavelet Transform (DWT) for the detection and localization of ten different types of power quality (PQ) disturbance signals. The SGWT provides the time domain interpretation which is an opposition to the frequency domain analysis of DWT. Further, the selected features are extracted from the detail coefficient of the variants of WT and given as inputs to the classifiers in order to characterize the signals. Moreover, a comparative assessment of the PQ signal carried out with different classifiers such as Support Vector Machine (SVM), Decision Tree (DT) and Random Forest (RF) have been presented along with aforementioned detection techniques. The RF is an ensemble decision tree and used for the classification of large number of data set. Hence, various single as well as combined power quality disturbance signals have been simulated in noisy and noise free environment in order to demonstrate the efficiency of the proposed techniques. Moreover, in order to represent in realistic environment, these proposed techniques are tested with both the single phase and tree phase signals captured from different transmission panels. Further, to aid this PQ disturbance detection, different types of real time fault signals are characterized with these aforementioned approaches.

Mono-modal feature extraction for bonding quality detection of explosive clad structure with optimized dual-tree complex wavelet transform

Bonding quality detection of explosive clad structure is significant to prevent catastrophic accidents. Multi-modal features related to bonding quality are contained in structural vibration response signal. Different modal feature has different sensitivity to the bonding quality. Extracting the desired mono-modal feature from the vibration response signal is necessary. Due to the mode aliasing easily appeared in the process of extracting the desired mono-modal feature, there is no effective method for this task. Dual-tree complex wavelet with attractive properties such as shift invariance and reduced spectral aliasing may provide a better way to extract the mono-modal feature. However, the fixed basis functions independent of the analyzed signal may weak the advantage of the method and even reduce the accuracy of detection result. To overcome this shortcoming, a technique called optimized dual-tree complex wavelet transform (ODTCWT) is proposed in this paper. Based on the analyzed signal, the optimized dual-tree complex wavelet basis function is constructed by searching for the proper parameters of vanishing moment K and the order of filter L. The optimized dual-tree complex wavelet with improved wavelet filters can best matched the modal frequencies of the analyzed signal. The ODTCWT can extract the mono-modal feature from vibration response signal with lower mode aliasing. The feasibility and effectiveness of the method of constructing ODTCWT is illustrated by the simulated signal. The proposed ODTCWT is combined with time entropy to detecting bonding quality of explosive clad pipes. For comparison, un-optimized dual-tree complex wavelet transform (UODTCWT), second-generation wavelet transform (SGWT) and band-pass filter (BPF) are also used for this task to demonstrate the validity of ODTCWT.

A hybrid fault diagnosis method based on second generation wavelet de-noising and local mean decomposition for rotating machinery

In order to extract fault features of large-scale power equipment from strong background noise, a hybrid fault diagnosis method based on the second generation wavelet de-noising (SGWD) and the local mean decomposition (LMD) is proposed in this paper. In this method, a de-noising algorithm of second generation wavelet transform (SGWT) using neighboring coefficients was employed as the pretreatment to remove noise in rotating machinery vibration signals by virtue of its good effect in enhancing the signal–noise ratio (SNR). Then, the LMD method is used to decompose the de-noised signals into several product functions (PFs). The PF corresponding to the faulty feature signal is selected according to the correlation coefficients criterion. Finally, the frequency spectrum is analyzed by applying the FFT to the selected PF. The proposed method is applied to analyze the vibration signals collected from an experimental gearbox and a real locomotive rolling bearing. The results demonstrate that the proposed method has better performances such as high SNR and fast convergence speed than the normal LMD method.

Detection of the Dynamic Unbalance with Cardan Shaft Applying the Second Wavelet Transform and Singular Value Decomposition

A new method of detecting cardan shaft dynamic unbalance in high-speed train is proposed applying the combination between the second generation wavelet transform(SGWT) and singular value decomposition(SVD) contrary to the basic defect of overlap frequency existing in the SGWT. The vibration acceleration signals installed on gimbal next to gearbox are decomposed through the SGWT to get the different scale decomposition signal. The Hankel matrix, which is constructed throng the single scale approximate decomposition signal, is orthogonally decomposed by SVD, the key singular values are selected to reconstruct vibration sign on the base of key stacks of singular values. Fourier spectrum of the reconstructed signal through SVD is applied to detect cardan shaft dynamic unbalance in high-speed train, eliminate the effect of overlap frequency on submerging unbalance fault features and highlight the failure characteristics. The method is verified by test data in the condition of dynamic unbalance, the results show that the method can effectively seize the fault features caused by cardan shaft dynamic unbalance, such as fundamental, multiplier, divider frequency. With comparison to the traditional wavelet decomposition, the simple SGWT, the clarity and failure characterization force of Fourier spectrum are significantly improved.

Composite Damage Detection Based on Redundant Second-Generation Wavelet Transform and Fractal Dimension Tomography Algorithm of Lamb Wave

In the purpose of achieving composite structure damage identification and localization of structural health monitoring, a denoising algorithm of redundant second-generation wavelet transform considering neighboring coefficients is selected as the best solution from 18 denoising schemes performed in this paper. Through introducing fractal dimension as a damage-sensitive feature and adopting the probabilistic reconstruction algorithm, the damage status on the composite panel could be identified and located as tomography maps. The practicability of the presented approach is validated in the experiment operating in the composite damage monitoring system.

On-line partial discharge monitoring in medium voltage underground cables

For many years incipient partial discharge (PD) faults in power cables have been identified through off-line investigation techniques. More recently, in an effort to allow pro-active asset management of the medium-voltage (MV) cable network to be carried out, continuous on-line monitoring systems are being installed with the aim of reducing unexpected failures. This study presents work on the analysis and handling of acquired data, from the point of view of asset management and the PD activities observed in an on-line cable monitoring systems. Initially, a review of on-line against off-line cable PD monitoring is presented, in terms of their setups and their respective advantages and disadvantages. The study then presents the authors’ experience of applying wavelet-based denoising techniques [both the discrete wavelet transform (DWT) and the second generation wavelet transform (SGWT)] to PD data denoising. Results of a study on the on-line PD-based monitoring of MV underground cables that are presented in the following section demonstrate that PD activity which is observed in on-line monitoring also has an associated high level of electrical noise that must be removed to allow proper identification of the PD signals and the denoised PD activity is seen to vary with time. Finally, the necessity of developing a means by which knowledge rules can automatically be acquired from on-line condition monitoring data, to reduce reliance on human expertise, is discussed.

Enhancement of signal denoising and multiple fault signatures detecting in rotating machinery using dual-tree complex wavelet transform

In order to enhance the desired features related to some special type of machine fault, a technique based on the dual-tree complex wavelet transform (DTCWT) is proposed in this paper. It is demonstrated that DTCWT enjoys better shift invariance and reduced spectral aliasing than second-generation wavelet transform (SGWT) and empirical mode decomposition by means of numerical simulations. These advantages of the DTCWT arise from the relationship between the two dual-tree wavelet basis functions, instead of the matching of the used single wavelet basis function to the signal being analyzed. Since noise inevitably exists in the measured signals, an enhanced vibration signals denoising algorithm incorporating DTCWT with NeighCoeff shrinkage is also developed. Denoising results of vibration signals resulting from a crack gear indicate the proposed denoising method can effectively remove noise and retain the valuable information as much as possible compared to those DWT- and SGWT-based NeighCoeff shrinkage denoising methods. As is well known, excavation of comprehensive signatures embedded in the vibration signals is of practical importance to clearly clarify the roots of the fault, especially the combined faults. In the case of multiple features detection, diagnosis results of rolling element bearings with combined faults and an actual industrial equipment confirm that the proposed DTCWT-based method is a powerful and versatile tool and consistently outperforms SGWT and fast kurtogram, which are widely used recently. Moreover, it must be noted, the proposed method is completely suitable for on-line surveillance and diagnosis due to its good robustness and efficient algorithm.

Second generation wavelet transform for data denoising in PD measurement

Detection and diagnosis of partial discharge (PD) activity has been widely adopted in electrical plant condition monitoring. Analysis and detection of PD in practical applications is often hampered by noise in the signal. Recent research has shown that the discrete wavelet transform (DWT) is effective in extracting PD pulses from severe noise. One disadvantage, however, is that DWT does not reproduce accurate PD pulse magnitude and pulse shape after thresholding in the presence of strong noise. This paper presents the application of the second generation wavelet transform (SGWT), as an improved algorithm, to extraction of PD pulse from electrical noise. The paper begins with the description of the fundamental theory and structure of SGWT analysis and comparisons with DWT. The method is then applied to both simulated and real world PD data. Results prove that SGWT can significantly improve the effectiveness of PD denoising.