Vladimir Celebic

Neural network ensemble for power transformers fault detection

Electrical transformers are the most important elements in the process of transmission and distribution of electricity. Depending on the size and position of the transformer, the sudden device failure can cause tremendous damage. Neural networks are widespread technique for transformer health monitoring. Neural Network Ensembles are an advanced neural technique that improves the accuracy and reliability in the transformers health diagnosis and failure prognosis. This paper describes a technique how to identify causal relation of dissolved gases in transformers oil and the current state of the transformers health. The described algorithm improves the interpretation of results obtained by dissolved gas analysis (DGA) technique. The most important result of this algorithm is a timely and reliable prediction of transformers failure based on incipient faults detection.

DSP-Based Approach to Generating Computer Model of Impulse Noise Using Kernel Data Sequence

In this paper, we present a new computer model to simulate the impulse-type noise that appears in bursts on a high- voltage (HV) power line during transient and emergency situations. The computer model is designed for investigations of the effects of impulse noise on communication channels implemented over an HV power line. The simulation process relies on an authentic record of measured noise whose modified replicas we use to compose a train of short-time impulses evenly distributed in time. The recorded noise sequence is referred to as a kernel data sequence over which we apply digital-signal-processing (DSP) techniques to randomize its length and its amplitudes. Noise impulses produced by the simulator achieve and retain the sharp front ends and spectral properties of the original noise. A variable digital filter is implemented to examine the influences of impulse noise on communication channels. Performances of simulated noise have been compared with those of measured noise published in CIGRÉ, IEEE, and IEC documents. Comparisons demonstrate a good agreement between the characteristics of measured and simulated noise.

Design and realization of FPGA platform for real time acoustic signal acquisition and data processing

This paper describes the realization of an FPGA platform designed for real time acoustic signal acquisition and data processing. The paper presents complete development process, starting from high level system design including signal integrity simulations to real system measurement results based on described platform. A real system for localization of the dominant noise source based on the platform is presented.

Software realization for the communication according to the IEC61850 standard on the nanoRISC MSC hardware platform

This paper describes software realization and implementation for the communication according to the IEC61850 Standard, between the module for supervising teleprotection equipments and the power substation server. Teleprotection equipment have important role in transmission of messages for power line section tripping. Software is implemented on the hardware platform type: MSC “nanoRISC-S3C2416 MB2” which belongs to the “COM - computer on module” systems.

Concept design of a system for partial discharges diagnostics using single board computer

This paper presents architecture of a system for signal acquisition, conditioning and processing, used in partial discharges diagnostics. System uses acoustic method for partial discharges detection, as noninvasive and interference resistant method, which can prevent transformer breakdown due to insulation degradation. The main processing platform is industrial embedded computer based on the newest OMAP5 processor from Texas Instruments. Besides digital signal processing, computer covers wide range of interfaces which could be used in mobile instrumentation.

Realization of the teleprotection equipment interface for working with PLC equipment using FPGA logic

This paper describes the realization of the interface in the teleprotection equipment for work with PLC equipment, and so to transmit teleprotection commands over power lines. The properties of the teleprotection equipment are shortly described, and after the realization of the line unit using FPGA technology (XILINX XC3S700AN) is described in details. Fourth part describes the simulation model realized in MATLAB-Simulink program.

Laboratory model for partial discharge diagnostics using acoustic emission measurement

One of the major causes of transformer failures is dielectric breakdown. Partial discharges cause gradual insulation degradation thus partial discharge activity monitoring provides transformer state insight. This paper gives an overview of common methods for partial discharges detection and source location in transformers, with special reference to the acoustic method as non invasive and interference resistant method suitable for application. For laboratory testing a model of a measuring system for acoustic emission measurement was developed.