OPTIMIZATION OF LCL-T RESONANT CONVERTER FOR CC-CV OPERATION

S.Praveena

PG student- Department of EEE – Mepco Schlenk Engineering College, Sivakasi Tamilnadu, India.

            The LCL-Type series resonant converter has nearly load independent output voltage under some operating conditions. It behaves as a constant current source (CC) when operated under resonant frequency. The output voltage of a CC power supply increases linearly with load resistance. Therefore a constant voltage limit must be incorporated for its use for applications like arc-welding power supplies etc .This limit can be done by incorporating clamp diodes. Since no feedback control is used, the proposed circuit is more reliable and rugged.

View full abstract...

MODIFIED HYBRID CASCADED H- BRIDGE MULTILEVEL INVERTER

                                                                A.Rajkumar

PG student- Department of EEE - SBM college of Engineering and technology, Dindigul, Tamilnadu, India.

            This paper presents a cascaded H-bridge multilevel inverter that can be implemented using only a single dc power source and capacitors. Standard cascaded multilevel inverters require N dc sources for 2N+1 level. Without requiring transformers, the scheme proposed here allows the use of a single dc power source (example: a battery or a fuel cell stack) with the remaining dc sources being capacitors, which is referred to as Hybrid Cascaded H-bridge Multilevel Inverter (HCMLI) in this paper. It is shown that the inverter can simultaneously maintain the dc voltage level of the capacitors and choose a fundamental frequency switching pattern to produce a nearly sinusoidal output. HCMLI using only a single dc source for each phase is promising for high power motor drive applications as it significantly decreases the number of required dc power supplies, provides high quality output power due to its high number of output levels, and results in high conversion efficiency and low thermal stress as it uses a fundamental frequency switching scheme.

View full abstract...

PERSONALIZED SERVICE RECOMMENDATION SYSTEM USING RANK BOOSTING ALGORITHM ON HADOOP

Abinaya R¹, Selva Lakshmi C B ²  

¹PG student- Department of CSE - Vellammal college of Engineering and technology, Madurai, Tamilnadu, India.

²Assistant Professor - Department of CSE- Vellammal college of Engineering and technology, Madurai, Tamilnadu, India.

  

                A recommendation system helps us to generate the appropriate service to users. In recent years, the amount of data in our world has been raising quickly, to yield the big data analysis problem. As a result, service recommender system rapidly suffers from inefficiency and scalability problems when processing large scale data. In addition, most of the existing approach provides service recommendations based on the user preference without considering users combined preference. For this purpose, rank boosting algorithm is proposed for combined preferences to generate appropriate recommendations. It aims at getting the input as combined preferences, based on the preferences it process the similarities with the reviews of the existing users then it provides the ranking to the services. Based on the ranking it generates the recommendation list as result to the end users for their combined preference. To improve its scalability and efficiency in big data environment, user preferences based Service recommendation system is implemented on a distributed computing platform, hadoop which uses Map Reduce as computing framework.

View full abstract...

A SAFE EVALUATION OF HEALTH CARE RESULT VALUES USING CLASSIFICATION

        B.Suriya¹, J.Santhalakshmi Revathy²

¹PG student- Department of CSE - Vellammal college of Engineering and technology, Madurai, Tamilnadu, India.

²Assistant Professor - Department of CSE- Vellammal college of Engineering and technology, Madurai, Tamilnadu, India.

                The clinical assessment arrangement provides fitness information about the patients and this will provide health care assessment values to the clinicians. It is built by using medical datasets. It contains large amount of data. From the large set of database, mining the details of diabetic patient and classify the symptoms and forecast whether the patient have diabetes at wicked level or at initial level. These healthcare clarifications should help the clinicians for upgraded healthcare assessments. This information is located in remote servers. In our day to day life directing data over large networks more securely is being a great challenge. So we’re guarding our information using some of the techniques like cryptography. So propose a privacy preserving clinical decision support system which conserves the privacy of the patient data. By providing privacy, unauthenticated persons cannot able to mine information about the patients. For investigational results this paper uses therapeutic datasets from UCI-database.

View full abstract...

POWER QUALITY ENHANCEMENT USING FUZZY LOGIC CONTROLLER BASED SHUNT HYBRID POWER FILTER WITH THYRISTOR CONTROLLED REACTOR

M.Vijayalakshmi ¹, R.Ganesan  ², S.Srinivasan  ³

¹PG student- Department of EEE – Sethu Institute of Technology, Kariapatti, Tamilnadu, India.

²Professor - Department of EEE – Sethu Institute of Technology, Kariapatti, Tamilnadu, India.

³ Associate Professor - Department of EEE – Sethu Institute of Technology, Kariapatti, Tamilnadu, India.

            This paper focuses on the combined system of a Thyristor Controlled Reactor (TCR) and a Shunt Hybrid Power Filter (SHPF) for the harmonic and reactive power compensation. The SHPF is the combination of a small-rating Active Power Filter (APF) and a fifth-harmonic-tuned LC passive filter. The tuned passive filter and the TCR form a Shunt Passive Filter (SPF) to compensate reactive power. The small-rating APF is used to improve the filtering characteristics of SPF and to suppress the possibility of resonance between the SPF and line inductances. A Fuzzy logic controller was used to control the TCR. A nonlinear control of APF was developed for current tracking and voltage regulation. The latter is based on a decoupled control strategy, which considers that the controlled system may be divided into an inner fast loop and an outer slow one. Thus, an exact linearization control was applied to the inner loop, and a nonlinear feedback control law was used for the outer voltage loop. Integral compensators were added in both current and voltage loops in order to eliminate the steady-state errors due to system parameter uncertainty. The simulation results are found to be quite satisfactory to mitigate harmonic distortions and reactive power compensation.

View full abstract...