A novel wavelet transform based voltage sag/swell detection algorithm

This paper proposes a novel sag/swell detection algorithm based on wavelet transform (WT) operating even in the presence of flicker and harmonics in source voltage. The developed algorithm is the hybrid of Daubechies wavelets of order 2 (db2) and order 8 (db8) to detect voltage sag/swell with and without positive/negative phase jumps. The hybrid detection algorithm can detect the start and end times of voltage sag/swell with and without phase jumps within 0.5 ms and 1.15 ms, respectively. The performance of the proposed voltage sag/swell detection method is compared with the results of dq-transformation, Fast Fourier Transform (FFT) and Enhanced Phase Locked Loop (EPLL) based voltage sag/swell detection methods. The good robustness and faster processing time to detect balanced and unbalanced voltage sag/swell are provided using proposed method. With the proposed hybrid detection algorithm consisting of db2 and db8 wavelet functions, a robust sag/swell detection is achieved which can give precise and quick response. The performance of proposed hybrid algorithm is validated and confirmed through simulation studies using the PSCAD/EMTDC analysis program.     

A novel approach to fuzzy wavelet neural network modeling and optimization

In this paper, an efficient approach of combining Takagi–Sugeno–Kang fuzzy system with wavelet based neural network is presented. The model replaces the constant or a linear function of inputs in conclusion part of traditional TSK fuzzy model with wavelet neural network (WNN), thus each rule uses fuzzy set to separate the input space into subspaces spanned by different wavelet functions. For finding the optimal values for parameters of our proposed fuzzy wavelet neural network (proposed-FWNN), a hybrid learning algorithm integrating an improved particle swarm optimization (PSO) and gradient descent algorithm is employed. The two-layer inline-PSO process is proposed in this paper, whose adjustment scheme is more fitting the consequent pattern learning based gradient descent optimization and will locate a good region in the search space. Simulation examples are given to test the efficiency of proposed-FWNN model for identification of the dynamic plants. It is seen that our modeling and optimization approach results in a better performance.     

Voice conversion using General Regression Neural Network

The objective of voice conversion system is to formulate the mapping function which can transform the source speaker characteristics to that of the target speaker. In this paper, we propose the General Regression Neural Network (GRNN) based model for voice conversion. It is a single pass learning network that makes the training procedure fast and comparatively less time consuming. The proposed system uses the shape of the vocal tract, the shape of the glottal pulse (excitation signal) and long term prosodic features to carry out the voice conversion task. In this paper, the shape of the vocal tract and the shape of source excitation of a particular speaker are represented using Line Spectral Frequencies (LSFs) and Linear Prediction (LP) residual respectively. GRNN is used to obtain the mapping function between the source and target speakers. The direct transformation of the time domain residual using Artificial Neural Network (ANN) causes phase change and generates artifacts in consecutive frames. In order to alleviate it, wavelet packet decomposed coefficients are used to characterize the excitation of the speech signal. The long term prosodic parameters namely, pitch contour (intonation) and the energy profile of the test signal are also modified in relation to that of the target (desired) speaker using the baseline method. The relative performances of the proposed model are compared to voice conversion system based on the state of the art RBF and GMM models using objective and subjective evaluation measures. The evaluation measures show that the proposed GRNN based voice conversion system performs slightly better than the state of the art models.     

A wavelet-based autoregressive fuzzy model for forecasting algal blooms

This paper proposes fuzzy models for forecasting the complex behavior of algal blooms. The models are developed through the integration of autoregressive models, the Takagi-Sugeno fuzzy model, and discrete wavelet transform algorithms. The premise parts of the proposed models are determined using the subtractive clustering technique and the consequent parts are optimized using weighted least squares. To train and validate the proposed fuzzy models, a large number of data sets were collected from Daecheong reservoir in Geum River in the Republic of Korea. The data include both water quality and hydrological variables. Total nitrogen, total phosphorous, dissolved oxygen, chemical oxygen demand, biochemical oxygen demand, pH, air temperature, water temperature and outflow water were evaluated as input signals while chlorophyll-a was used as an output. It is demonstrated from the simulation that the proposed fuzzy models are effective in forecasting algal blooms.     

Modified dynamic surface approach with bias torque for multi-motor servomechanism

This paper presents a modified neural dynamic surface control (DSC) with an adaptive bias torque for the multi-motor servomechanism (MMS) with backlash, friction and other disturbances. By introducing a continuous hybrid differentiator to replace the first-order filter in each step, a modified DSC is developed to improve the load tracking precision of MMS. However, when the MMS enters the backlash band, only DSC cannot guarantee the load tracking performance. Thus, an adaptive bias torque is firstly proposed based on the prescribed performance function technique to compensate the backlash nonlinearity and guarantee the load tracking performance of MMS. In addition, the unknown dynamics including the friction and other disturbances are approximated by using wavelet echo state networks where the weights are all updated online. By means of Lyapunov stability theory, the semi-globally uniformly ultimately bounded (SGUUB) property of all signals in the closed-loop system is proved. Finally, simulations and experimental results based on a four-motor servomechanism are presented to show the effectiveness of the proposed approach.     

Generalized neural network and wavelet transform based approach for fault location estimation of a transmission line

To maintain the efficient and reliable operation of power systems, it is extremely important that the transmission line faults need to be detected and located in a reliable and accurate manner. A number of mathematical and intelligent techniques are available in the literature for estimating the fault location. However, the results are not satisfactory due to the wide variation in operating conditions such as system loading level, fault inception instance, fault resistance and dc offset and harmonics contents in the transient signal of the faulted transmission line. Keeping in view of aforesaid, a new approach based on generalized neural network (GNN) with wavelet transform is presented for fault location estimation. Wavelet transform is used to extract the features of faulty current signals in terms of standard deviation. Obtained features are used as an input to the GNN model for estimating the location of fault in a given transmission systems. Results obtained from GNN model are compared with ANN and well established mathematical models and found more accurate.     

A splitting method for total least squares color image restoration problem

Color image restoration is an important problem in image processing. Using the structured total least squares (STLS) for fidelity term of the restoration process gives better results in comparison with the least squares (LS) approach. The main drawback of the STLS approach is its complexity. To overcome this issue, in this paper by an appropriate transformation the color image restoration is substituted with two smaller subproblems corresponding to smooth and oscillatory parts of the image. The first and second subproblems are modeled via STLS and LS approaches, respectively. We show that the proposed method is faster than STLS and gives competitive solutions with it. Also, we demonstrate that Haar wavelet perseveres the structure of the blurring operator, which causes a considerable reduction in computational and storage complexity of the proposed method.     

Biogeography based hybrid scheme for automatic detection of epileptic seizures from EEG signatures

Biogeography Based Optimization (BBO) algorithm is one of the nature-inspired optimization methods, based on the study of geographical distribution of species on earth. The present research work is based upon decomposition of human electroencephalograms (EEG) signal by Wavelet Packet Transform, computation of a complete feature set using statistical and non-statistical properties followed by optimal selection of features by BBO; the optimality criterion being classification rate. The stopping criterion for BBO is set to 100% correct classification rate. The proposed algorithm is novel in terms of TWSVM computing the Habitat Suitability Index (HSI) values for BBO, perfect classification rate, low computation time, and feature selection mechanism. The proposed scheme outperforms several previous results reported in literature in that it gives a feature subset which gives 100% classification accuracy for different classification instances.     

Ground penetrating radar response from voids: A demonstration using a simple model

A simple Ricker wavelet model can be used to illustrate some fundamental properties of the ground penetrating radar (GPR) response from both air- and water-filled voids. Reflections from the top and the bottom of a void overlap significantly, and generate one common characteristic of a void response: a “bright spot”, analogous to the “bright spot” observed in seismic exploration. For time delays equal to about half the pulsewidth, the reflected wavelets superimpose to yield maximum reflection amplitudes. The top reflected wavelet becomes completely separated from the bottom reflected wavelet when the void time delay exceeds twice the wavelet pulsewidth. The two wavelet reflections can be individually identified at earlier time delays, approximately equal to the wavelet pulsewidth. This is still substantial and explains why it is difficult to use GPR to infer void thicknesses, especially for air-filled voids.