S. Venkatesh

Bio: S. Venkatesh is an academic researcher from Shanmugha Arts, Science, Technology & Research Academy. The author has contributed to research in topics: Artificial neural network & Probabilistic neural network. The author has an hindex of 6, co-authored 17 publications receiving 149 citations. Previous affiliations of S. Venkatesh include Yahoo!.

Partial discharge pattern classification using composite versions of probabilistic neural network inference engine

Abstract: A major requirement of any power apparatus is the reliable performance of its insulation. The incidence of minor flaws and irregularities such as voids, surface imperfections, in the electrical insulation is however inevitable and lead to partial discharge (PD). Since each defect has a unique degradation mechanism, it is imperative to ascertain the correlation between the discharge patterns and the type of defect in order to evaluate the quality of the insulation. Efforts to correlate discharge patterns with the type of defects have been undertaken by several researchers. Though encouraging attempts to recognize and classify simple PD defect sources have been reported, misclassifications still occur, which affect the assessment of the index of the insulating degradation. A Composite Probabilistic Neural Network Inference System has been devised and elucidated in this research using two versions of Probabilistic Neural Network. The inference is obtained based on the outcome to innovatively conceived fourteen unique characteristic vector inputs to enable an accurate and reliable decision in the classification of complex stochastic PD patterns thus obviating the necessity of skilled operators. Validation of the fingerprints of PD patterns has also been carried out using well-established techniques.

ART 2—an unsupervised neural network for PD pattern recognition and classification

Abstract: This paper introduces a method of classifying partial discharges of unknown origin. The innovative trend of using Artificial Neural Network (ANN) towards classification of Partial Discharge (PD) patterns is cogent and discernible. The Adaptive Resonance Theory (ART), a type of neural network which is suitable for PD pattern recognition is explained here. To ensure the suitability and reliability of chosen network for PD pattern recognition, the network is tested with the well known Iris plant database and alphabet character for recognition & classification. Further more the network is trained with various combinations off–q–n distributions of PD patterns and tested. It is shown that the ART 2 network is able to classify the PD patterns. The paper ends with analyzing the efficacy of multifarious features selected in the measurement space. Also the validation of input features is done using ‘Hold-One-Out’ method and partial set training technique q 2005 Elsevier Ltd. All rights reserved.

Robust Heteroscedastic Probabilistic Neural Network for multiple source partial discharge pattern recognition - Significance of outliers on classification capability

Abstract: Among various insulation diagnostic techniques utilized by researchers and personnel handling power equipment, partial discharge (PD) recognition and analysis has emerged as a vital methodology since it is inherently a non-intrusive testing strategy. Of late, the focus of researchers has shifted to the identification and classification of multiple sources of PD since it is most often encountered in practical insulation systems of power apparatus. Researchers have carried out studies to recognize multi-source PD and expounded the difficulties experienced in discriminating such discharge patterns. It has also been observed that identification of such patterns becomes increasingly difficult with the degree of overlap. Review of recent research studies indicates that classification of fully overlapped patterns is yet an unresolved issue and that techniques such as Mixed Weibull Functions, neural networks (NN) and Wavelet Transformation have been attempted with reasonable degree of success for single source and partially overlapped PD patterns only. This research study focuses on extending the previous work attempted by the authors in utilizing the novel approach of Heteroscedastic Probabilistic Neural Network (HRPNN) for classification of single source PD patterns to that of multiple PD sources also. Further, a Robust Heteroscedastic Probabilistic Neural Network (RHRPNN) is implemented for the classification of multi-source PD patterns. The RHRPNN utilizes the jackknife procedure for handling problems associated with training the neural network due to the presence of outliers, thus providing a compact yet effective set of centers in terms of probability density functions. In addition to the previously utilized traditional statistical operators in the pre-processing phase, a Two Pass Split Window (TPSW) scheme has also been developed to study and compare the classification capability of the RHRPNN with that of HRPNN. Detailed analysis of the performance of RHRPNN is carried out to ascertain the influence of the smoothing parameter in classifying PD patterns, to determine the role played by the pre-processing techniques during classification and to find the significance of the parsimonious set of centers in eliminating the effect of outliers during classification. Finally, the ability of the HRPNN and the RHRPNN in classifying large dataset multiple source PD patterns obtained from varying applied voltages is analyzed for its further applicability in real-time PD pattern recognition studies.

Orthogonal least square center selection technique - A robust scheme for multiple source Partial Discharge pattern recognition using Radial Basis Probabilistic Neural Network

Abstract: Partial Discharge (PD) pattern recognition has emerged as a subject of vital interest for the diagnosis of complex insulation system of power equipment to personnel handling power system utilities and researchers alike, since the phenomenon serves inherently as an excellent non-intrusive testing technique. Recently, the focus of researchers has shifted to the recognition of defects in insulation due to multiple PD sources, as it is often encountered during real-time PD measurements. A survey of research literature indicates clearly that the recognition of fully overlapped PD patterns is yet an unresolved issue and that techniques such as Mixed Weibull Function, Neural Network (NN), Wavelet Transformation, etc. have been attempted with only reasonable success. Since most digital PD online acquisition systems record data for a stipulated and considerable duration as mandated by international standards, the database is large. This poses substantial complexity in classification during the training phase of the NNs. These difficulties may be attributed to ill-conditioned data, non-Markovian nature of discharges, curse of dimensionality of the data, etc. Since training methods based on random selection of centers from a large training set of fixed size are found to be relatively insensitive and detrimental to classification in many cases, a Forward Orthogonal Least Square algorithm (FOLS) is utilized in order to reduce the number of hidden layer neurons and obtain a parsimonious yet optimal set of centers. This algorithm, in addition, obviates the need for a separate clustering method making the procedure inherently viable for on-line PD recognition. This research work proposes a novel approach of utilizing Radial Basis Probabilistic Neural Network (RBPNN) with FOLS center selection algorithm for classification of multiple PD sources. Exhaustive analysis is carried out to ascertain the efficacy of classification of the proposed RBPNN-FOLS algorithm to cater to large training data set. A detailed comparison of the performance of the proposed scheme with that of the standard version of Probabilistic Neural Network (PNN) and Heteroscedastic PNN (HRPNN) that was taken up for study by the authors in their previous work indicates firstly the effectiveness of FOLS algorithm in obtaining parsimonious centers, points out secondly the capability of the Radial Basis Probabilistic Neural Network (RBPNN) model to integrate the advantages of the Radial Basis Function Neural Network (RBFNN) and PNN in classifying multiple PD sources and finally throws light on the exceptional capability of the FOLS-RBPNN in discriminating the sources of PD due to varying applied voltages also.

A heuristic complex probabilistic neural network system for partial discharge pattern classification

Abstract: Partial discharge (PD) pattern classification has recently become popular since the automated acquisition of PD signals has become vital and cogent. A novel method for identification of defects due to partial discharge is described in this paper. Starting from different PD families of specimen, several sets of characteristic vectors are determined and then used as input variables to the proposed neural network. The innovative trend of using probabilistic neural network (PNN) towards classification of PD patterns is coherent and perceptible. The paper elucidates the structure of PNN, which has been appropriately customized for determining the optimum value of smoothing parameter. PD is measured using the conventional discharge detector and previously developed statistical tools that processed the PD patterns. Satisfactory results in the past have revealed that the analysis of the properties of the phase position distributions can be made using mathematical descriptors. The ability of PNN to classify these descriptors in addition to classifying the inputs derived from the measures based on central tendency, dispersion, and maximum and minimum values are investigated. The classification of single-type insulation defects has been envisaged. The paper also expounds a novel complex technique adopted for precise PD classification.

Shape recognition based on radial basis probabilistic neural network and application to plant species identification

Abstract: In this paper, a novel shape recognition method based on radial basis probabilistic neural network (RBPNN) is proposed. The orthogonal least square algorithm (OLSA) is used to train the RBPNN and the recursive OLSA is adopted to optimize the structure of the RBPNN. A leaf image database is used to test the proposed method. And a modified Fourier method is applied to descript the shape of the plant leaf. The experimental result shows that the RBPNN achieves higher recognition rate and better classification efficiency with respect to radial basis function neural network (RBFNN), BP neural network (BPNN) and multi-Layer perceptron network (MLPN) for the plant species identification.

A decision tree logic based recommendation system to select software fault prediction techniques

Abstract: Identifying a reliable fault prediction technique is the key requirement for building effective fault prediction model. It has been found that the performance of fault prediction techniques is highly dependent on the characteristics of the fault dataset. To mitigate this issue, researchers have evaluated and compared a plethora of fault prediction techniques by varying the context in terms of domain information, characteristics of input data, complexity, etc. However, the lack of an accepted benchmark makes it difficult to select fault prediction technique for a particular context of prediction. In this paper, we present a recommendation system that facilitates the selection of appropriate technique(s) to build fault prediction model. First, we have reviewed the literature to elicit the various characteristics of the fault dataset and the appropriateness of the machine learning and statistical techniques for the identified characteristics. Subsequently, we have formalized our findings and built a recommendation system that helps in the selection of fault prediction techniques. We performed an initial appraisal of our presented system and found that proposed recommendation system provides useful hints in the selection of the fault prediction techniques.