College of Engineering, Pune

About: College of Engineering, Pune is a based out in . It is known for research contribution in the topics: Sliding mode control & Control theory. The organization has 4264 authors who have published 3492 publications receiving 19371 citations. The organization is also known as: COEP.

Modelling and simulation of DSTATCOM for power quality improvement in distribution system using MATLAB SIMULINK tool

Abstract: This paper presents Distribution Static Compensator (DSTATCOM) modeled in the MATLAB SIMULINK toolbox for the mitigation of the power quality issues in the distribution system. DSTATCOM is one of the custom power device used in distribution system for power conditioning. DSTATCOM is developed for the compensating reactive power demanded by non-linear and unbalanced load. Also power factor of the source is improved and the Total Harmonic Distortion in the source currents is reduced. DSTATCOM can correct voltage sag, swell, unbalance by injecting the reactive current into the system. Instantaneous reactive power theory is used for obtaining reference source current for controlling DSTATCOM. The performance of the DSTATCOM by using the IRP theory for unbalanced and non linear load is demonstrated with the MATLAB simulation results.

Visualizing semantic web

Abstract: Extensive volume of information spread across the web poses a major challenge for identification of relevant information. Attaching concise meaning to this information allows for effective retrieval and usage. In this paper, we propose an interactive component that extracts pertinent information from this structured data and presents it to the user through an innovative graph mechanism, namely the spring graph. Adobe Flash builder forms the base for spring graph generation, whereas deduction of desired result from semantic data is handled by Jena framework and SPARQL query language. RDF technology is used to structure information aggregated from various web sources. The current implementation reads RDF file input and visualizes it in the form of nodes of the spring graph. The solution is validated through relevant experiments that test the efficiency and proper functionality of the system. The proposed mechanism renders better quality graphs, visual output and concrete results instead of long list of documents.

Impacts of increasing photovoltaic penetration on distribution grid — Voltage rise case study

Abstract: Increasing electricity prices and unreliable grid supply, has led to increasing PV installation in a distribution network. While the PV installations on a distribution grid reduces the power demand, and eases grid congestion, the increasing PV penetration has several other impacts on the grid, both positive and negative. The paper makes a comprehensive study of the various impacts. Simulations analysis of the most prominent impact on the grid — voltage rise, in a distribution line is performed to understand the severity of the problem for different PV penetration levels. The paper also performs simulation analysis to estimate the impact of the voltage rise on transformer tap changes to regulate the voltage within permissible limits as per standards. It was seen that as the PV penetration increases, the number of tap changes required increase resulting in the reduced life time of the transformer.

A Review of Machine Learning Methodologies for Network Intrusion Detection

Abstract: The Internet continues to spread itself over the globe, providing a great opportunity for various threats which are growing on a daily basis. Current static detection techniques only detect known malicious attacks and they also require frequent updates to signature-based databases. To reduce this work, systems are proposed for network intrusion detection systems capable of analyzing contents of the network by means of machine learning techniques to analyze and classify the malicious contents. Various machine learning algorithms are used for developing a Network Intrusion Detection System. The review intends to provide an exhaustive survey of the currently proposed machine learning based intrusion detection systems in order to assist Network Intrusion Detection System developers to gain a better intuition.