Time filter

Source Type

Thangaprakash S.,Sri Shakthi Institute of Engineering and Technology
Journal of Engineering Research | Year: 2010

This paper presents a modified control algorithm for Space Vector Modulated (SVM) Z-Source inverters. In traditional control strategies, the Z-Source capacitor voltage is controlled by the shoot through duty ratio and the output voltage is controlled by the modulation index respectively. Proposed algorithm provides a modified voltage vector with single stage controller having one degree of freedom wherein traditional controllers have two degrees of freedom. Through this method of control, the full utilization of the dc link input voltage and keeping the lowest voltage stress across the switches with variable input voltage could be achieved. Further it offers ability of buck-boost operation, low distorted output waveforms, sustainability during voltage sags and reduced line harmonics. The SVM control algorithm presented in this paper is implemented through Matlab/Simulink tool and experimentally verified with Z-source inverter prototype in the laboratory.


Thangaprakash S.,Sri Shakthi Institute of Engineering and Technology
Journal of Power Electronics | Year: 2010

This paper presents a current mode integrated control technique (CM-ICT) using a modified voltage space vector modulation (MSVM) for Z-source inverter (ZSI) fed induction motor drives. MSVM provides a better DC voltage boost in the dc-link, a wide range of AC output voltage controllability and a better line harmonic profile. In a voltage mode ICT (VM-ICT), the outer voltage feedback loop alone is designed and it enforces the desired line voltage to the motor drive. An integrated control technique (ICT), with an inner current feedback loop is proposed in this paper for the purpose of line current limiting and soft operation of the drive. The current command generated by the PI controller and limiter in the outer voltage feedback loop, is compared with the actual line current, and the error is processed through the PI controller and a limiter. This limiter ensures that, the voltage control signal to the Z-source inverter is constrained to a safe level. The rise and fall of the control signal voltage are made to be gradual, so as to protect the induction motor drive and the Z-source inverter from transients. The single stage controller arrangement of the proposed CM-ICT offers easier compensation. Analysis, Matlab/Simulink simulations, and experimental results have been presented to validate the proposed technique.


Charles S.,Sri Shakthi Institute of Engineering and Technology | Venkateshkumar P.,Karunaya University
Australian Journal of Electrical and Electronics Engineering | Year: 2012

The use of active power filters is widely accepted and implemented as a solution to the power quality problems in utility, industry and commercial applications. This paper presents an optimisation based control algorithm for determining the reference compensation currents of shunt active power filter (SAPF) to maintain sinusoidal source currents supplied to non-linear loads and to maintain the total harmonic distortion (THD) less than 5% according to IEEE-519 standard and improve power factor. Optimisation is due to the combined effect of perfect harmonic cancellation (PHC) and unity power factor (UPF) strategies. Reference compensation current is calculated through MATLAB using discrete Fourier transform. The optimisation-based control algorithm gives the best power factor while satisfying the constraints such as THD limits and average power balance of source currents. It is also flexible to adopt different compensation cluiracteristics based on the supply voltage conditions. The proposed technique is implemented using MATLAB/Simulink and optimisation tool boxes for both balanced and unbalanced load conditions. The algorithm is validated by using a prototype of digital signal processor (TMS320F2812) based SAPF. Detailed experimental results are presented. © Institution of Engineers Australia, 2012.


Kannammal K.E.,Sri Shakthi Institute of Engineering and Technology | Purusothaman T.,Government of Tamilnadu | Manjusha M.S.,Sri Shakthi Institute of Engineering and Technology
Journal of Theoretical and Applied Information Technology | Year: 2014

Wireless Sensor Network (WSN) consists of hundreds or thousands of sensor nodes which have limited energy, computation and memory resources. These sensors are randomly deployed in a specific area to collect useful information periodically for few months or even few years. The applications of WSN in some extreme environment make sensor nodes difficult to replace once the battery lifetime expires. Since the wireless transmission is the most energy consuming operation, designing an energy efficient routing protocol becomes the main goal for the wireless sensor network. LEACH is considered as the most popular routing protocol which has better performance in saving the energy consumption. However, the cluster head choosing formula neglects the change of node's energy will make the nodes acting as cluster heads too many times leading the cluster head die early by consuming too much energy. This paper presents K means clustering approach for clustering and a two level fuzzy logic approach to Cluster Head (CH) election based on four parameters namely - number of neighbor nodes, remaining energy, energy dispersion and distance from the base station. The proposed method has proved that it prolong the network lifetime than the LEACH protocol by comparison simulations using Matlab. © 2005 - 2014 JATIT & LLS. All rights reserved.


Kannammal K.E.,Sri Shakthi Institute of Engineering and Technology | Purusothaman T.,Government College of Technology, Coimbatore
European Journal of Scientific Research | Year: 2011

Adhoc networks classify the sensor networks as one of the infrastructure less networks. Wireless Sensor Networks (WSN) are a self configurable, healing network which finds variety of emerging applications in today's world such as military, movement tracking, industries and medical fields. In Mobile Sensor Network (MSN) nodes move their position dynamically without any infrastructure and to operate under unattended mode. Therefore it is required an efficient protocol to establish a connection between the mobile nodes. Directed Diffusion and flooding are the data centric routing protocols in WSN. The diffusion routing protocols are application specific and their performance may vary across mobility model and node speed. So, it is significant in bringing out a comparison of the various routing protocols for better understanding of the performance. In this paper we have analyzed the significant comparison of these protocols with random way point mobility model using Ns2 simulator. The simulation results illustrate the performance of the routing protocol across different performance metrics like overhead, MAC load, power consumption and throughput. © EuroJournals Publishing, Inc. 2011.


Kannammal K.E.,Sri Shakthi Institute of Engineering and Technology | Purusothaman T.,Government College of Technology, Coimbatore
European Journal of Scientific Research | Year: 2012

Directed diffusion protocol in highly mobile environments will lead to much network overhead and poor performance. In Ns2, there are two kinds of directed diffusion protocols, rate based gradient mechanism (diffusion/rate) and probability based gradient mechanism (diffusion/prob).In this paper, we focus a simple solution to improve the performance of directed diffusion in mobile scenario. To achieve this, we have modified the interest propagation mechanism and control the interest propagation with respect to the mobility as well as the nearby density of the propagating node. As a proof of concept, we have added the mobility and density aware interest propagation mechanism with ns2 implementation of diffusion/prob. Comparison between proposed directed diffusion with existing diffusion protocols and flooding has been successfully observed. The proposed directed diffusion protocol implies the better performance than other protocols in terms of overhead, MAC load and dropped packets. © 2012 EuroJournals Publishing, Inc.


Dharshini R.,Sri Shakthi Institute of Engineering and Technology | Hemanandhini S.,Sri Shakthi Institute of Engineering and Technology
2016 International Conference on Computer Communication and Informatics, ICCCI 2016 | Year: 2016

One of the serious kind of disease in the medical field is considered to be the brain tumor. So it must to have the fast and accurate detection. Different algorithms are provided for the tumour detection and segmentation. The important approach in the brain tumor segmentation is to identify the various stages includes benign, malignant and the normal. For the process of classifying voxels, a classifier called Self Organising Map(SOM) is used. Self-Organizing Map (SOM) which includes the unsupervised learning algorithm and Learning Vector Quantization (LVQ) with high diversity data like tumour appearance and its contour deformation. On comparing with the other approaches, instead of using Multimodal MRI images for clustering of voxels SOM has been used. A feature vector has been constructed in which the features for the constructing vector have been gained from the Discrete Wavelet Transform (DWT) coefficients. This has been constructed mainly for identifying tissue types which includes White Matter (WM), Grey Matter (GM), Cerebrospinal Fluid (CSF) and sometimes pathological tissues. The accuracies and the performance are provided in terms of training performance and the classification accuracies. The results which has been been simulated, shows the better accuracies of classifier and segmentation on comparing with the previous method. © 2016 IEEE.


Selvam D.,Sri Shakthi Institute of Engineering and Technology | Jananee M.,Sri Shakthi Institute of Engineering and Technology
2016 International Conference on Computer Communication and Informatics, ICCCI 2016 | Year: 2016

Micro-reviews are a considered as a source of review content in the social media. In this work, we study the micro-reviews, and use them to solve the problem of review selection. We introduce a novel formulation of review selection, to ensure efficiency and increase the coverage, which leads to novel coverage problems. The mechanism for helping the users with comprehensive set of reviews which satisfies their need is critical. So, here we consider approximation and heuristic algorithms, and study them experimentally, in order to demonstrate quantitatively and qualitatively the benefits of our approach. We also propose an Integer Linear Programming (ILP) formulation and an optimal algorithm. This helps us to quantify the approximation quality of the greedy heuristics. Experimental results explain the performance of the system against the State of art approach in terms of precision and recall. © 2016 IEEE.


Shabariram C.P.,Sri Shakthi Institute of Engineering and Technology | Kannammal K.E.,Sri Shakthi Institute of Engineering and Technology | Manojpraphakar T.,Sri Shakthi Institute of Engineering and Technology
2016 International Conference on Computer Communication and Informatics, ICCCI 2016 | Year: 2016

Rainfall data is collected to predict the storm warnings from the hydrological data. This is considered as a research idea as it consumes huge number of records from the distributed system. This paper describes a novel solution to manage the data based on spatial temporal characteristics using a Map Reduce Framework. The workload is classified using Support Vector Machine (SVM). It uses feature selection and reduction algorithm associated with the dataset.Various rainstorm concept prediction is achieved using the big raw rainfall data. The dataset impact parameters are classified into local, hourly, and overall storms. The proposed system serves as a tool for predicting rainstorm from a large amount of rainfall data in a efficient manner. The result indicates the proposed system improves the performance in terms of accuracy and efficiency. © 2016 IEEE.

Loading Sri Shakthi Institute of Engineering and Technology collaborators
Loading Sri Shakthi Institute of Engineering and Technology collaborators