Pawel Kostyla

Adaptive neural networks for robust estimation of signal parameters

In many applications, very fast methods are required for estimating of parameters of harmonic signals distorted by noise. Most of the known digital algorithms are not fully parallel, so that the speed of processing is quite limited. In this paper new parallel algorithms are proposed, which can be implemented by analogue adaptive circuits employing some neural networks principles. Algorithms based on the least‐squares (LS) and the total least‐squares (TLS) criteria are developed and compared. The problems are formulated as optimization problems and solved by using the steepest descent continuous‐time optimization algorithm. The corresponding architectures of analogue neuron‐like adaptive processors are also shown. The developed networks are more robust against noise in the measured signal than other known neural network algorithms. The network based on the TLS criterion optimizes the estimation under the assumption that the signal model can also be perturbated (frequency or sampling interval fluctuation an...

Wide-area system of registration and processing of power quality data in power grid with distributed generation: Part I. System description, functional tests and synchronous recordings

This paper presents the developed and built comprehensive system of registration, archiving and data processing for the wide-area monitoring of power quality in a separated part of real power grid with distributed renewable generation. Real case studies related to localization of sources of voltage disturbances are presented.

Adaptive On-line Learing Algorithm for Robust Estimation of Parameters of Noisy Sinusoidal Signals

In many applications, very fast methods are required for estimating of parameters of harmonic signals distorted by noise. Most of the known digital algorithms are not fully parallel, so that the speed of processing is quite limited. In this paper new parallel algorithms are proposed, which can be implemented by analogue adaptive circuits employing some neural networks principles. Algorithms based on the least-squares (LS) and the total least-squares (TLS) criteria are developed and compared. Extensive computer simulations confirm the validity and performance of the proposed algorithms.

Potentials of microgrids - RES infeed, stationary storage and controllable loads modelling

The general possibility of combining electric vehicles, PV systems and storage devices is useful for the reliability and quality of electricity distribution. It contributes to optimal operation of distribution networks and minimization of transmission losses. The potential of a microgrid concept can be evaluated using a reliable and reality oriented simulation allowing various case studies and resizing of microgrid elements. This paper presents an attempt to develop and simulate and robust system based on long time collection of real data.

Artificial neural network for detection of asynchronous state

An asynchronous state of a synchronic machine may be identified through determining the amplitudes of particular components of stators current provided that a constant slip value is assumed. Following a synchronism loss, this adopted value is assumed to be achieved and, for sure, exceeded. New parallel algorithms for detection of asynchronous state of synchronic machines, are proposed. The algorithms can be implemented by analogue adaptive circuits employing some neural networks principles. This chapter provides a description of artificial neural networks realising this task, whose operation algorithm is based on minimum square error criteria and maximum loss method.

Adaptive Neural Networks for Robust Estimation of Signal Parameters

Abstract In many applications, very fast methods are required for estimating of parameters of harmonic signals distorted by noise. Most of the known digital algorithms are not fully parallel, so that the speed of processing is quite limited. In this paper new parallel algorithms are proposed, which can be implemented by analogue adaptive circuits employing some neural networks principles. Algorithms based on the least-squares (LS) and the total least-squares (TLS) criteria are developed and compared. Extensive computer simulations confirm the validity and performance of the proposed algorithms.

Synchronized profiles of power quality parameters in assessment of disturbances in power systems with distributed generation

This paper presents implementation of instantaneous voltage-current trajectory (I, U) and its linearization. The aim of the work is to investigate possibility of using the slope of the linearized trajectories obtained synchronously in few measurement points. Field measurement results are presented using wide-area power quality monitoring system installed in distribution network with small hydro power plants.

Detection of power system transients disturbances in distributed generation systems using Hubert transform and signal decomposition

An idea of empirical mode decomposition (EMD) with Hilbert Transform (HT) as well as cross-section of smoothed pseudo Wigner-Ville distribution (SPWVD) and instantaneous distortion index (IDIN) are investigated in point of detection of high frequency inrush power quality disturbances. Application and assessment of mentioned techniques are performed using disturbances recorded by wide area power quality monitoring system installed in the three substations connected to the same medium voltage line i.e. two substations of small hydro power plants (SHP) and one urban substation. Dynamic aspect of SHP integration with power systems are tracked using mentioned techniques in point of local and area mutual influence.

Adaptive artificial neural network for robust estimation of parameters of asynchronous state

In many applications, very fast methods are required for estimating of amplitudes of current signals distorted by noise. In this paper new parallel algorithms are proposed, which can be implemented by analogue adaptive circuits employing selected neural networks principles. Algorithms based on the total least-squares (TLS) and the robust total least-squares (RTLS) criteria are developed and used to detection of asynchronous state of synchronous machines. The problems are formulated as optimization tasks and solved using the steepest descent continuous-time optimization algorithm. The corresponding architectures of analogue neuron-like adaptive processors are also shown. The developed networks are more robust against noise in the measured signal than other known neural network algorithms. The network based on the TLS criterion realizes the optimization process under the assumption that the signal model can also be deteriorated (frequency or sampling interval fluctuation and so forth). The TLS estimation effect is better and more reliable than the corresponding LS structure, when higher sampling frequency and a wider sampling window is applied. An asynchronous state of a synchronous machine may be identified through determining the amplitudes of particular components of stators current provided that a constant slip value is assumed. Following a synchronism loss, this adopted value is assumed to be achieved and, for sure, exceeded. This work provides a description of artificial neural networks realising mentioned task of asynchronous state detection. Extensive computer simulations confirm the validity and performance of the proposed algorithms.

Detection of arc in the transmission power lines

Automatic one-and three-phase re-closing power lines, after short circuit shutdown, is a very effective way to improve the reliability of power delivery. Re-closing while the fault is not cleared can be dangerous for some electrical appliances. To prevent re-closing of a short circuit a method allowing stable arc detection has developed. The method is based on the estimation of real-time parameters in voltage signals and an analysis of error estimation. A neural network has been proposed to solve the problem in real time.