Zbigniew Waclawek

Computation of voltage sag initiation with Fourier based algorithm, Kalman filter and Wavelets

Dynamic Voltage Restorers (DVR) have been successfully applied for voltage dip mitigation in the last years. Especially in systems with nonlinear loads and wind turbine generation DVR units support the Power Quality enhancement. The reliability and quality of DVR operation depends mostly on fast and accurate voltage dip detection. Detection methodologies must be able to detect a voltage dip as fast as possible and be immune to other types of perturbations. In this paper we address the problem of voltage dip estimation using carefully selected advanced signal processing methods such as Fourier based algorithm, Kalman filtering and Wavelets. Additionally, the traditional and common technique of RMS value tracking has been mentioned. The algorithms have been tested under different conditions: voltage dip with phase jump, noise, frequency variations.

Application of advanced signal processing methods for accurate detection of voltage dips

Custom Power Devices like the dynamic voltage restorer have been applied for voltage dip mitigation in the last years. These electronic equipments need fast and reliable voltage dip detection algorithms. Such detection methodologies must be able to detect a voltage dip as fast as possible and be immune to other types of perturbations. In this paper we address the problem of voltage dip estimation by using different advanced signal processing methods such as Kalman filtering, Fourier based algorithms and Wavelet processing. The three algorithms have been tested under different conditions: voltage dip with phase jump, noise, frequency variations. The final implementation on a Texas Instrument DSP TMS320F2812DSP has been done.

Signal Analysis in Converter-Fed Drives Using Adaptive Neural Networks

Frequency converter-fed induction motor drives were simulated using the EMTP-ATP (Electromagnetic Transients Program-Alternative Transients Program). The aim of the simulation investigations is to develop a feasible method for drive visualization and for diagnosis of different unsymmetric operations. Visualization of a converter-supplied drive by means of modal space phasors is a very useful and compact observation and diagnosis method. Fast and exact estimation of the frequency of distorted voltages and currents at the motor input enables us to obtain stationary space-phasors and to determine an appropriate sampling window for a DFT analysis. For this purpose, an adaptive neural network is proposed. Computer simulation results confirm the validity and performance of the proposed method.Frequency converter-fed induction motor drives were simulated using the EMTP-ATP (Electromagnetic Transients Program-Alternative Transients Program). The aim of the simulation investigations is to develop a feasible method for drive visualization and for diagnosis of different unsymmetric operations. Visualization of a converter-supplied drive by means of modal space phasors is a very useful and compact observation and diagnosis method. Fast and exact estimation of the frequency of distorted voltages and currents at the motor input enables us to obtain stationary space-phasors and to determine an appropriate sampling window for a DFT analysis. For this purpose, an adaptive neural network is proposed. Computer simulation results confirm the validity and performance of the proposed method.

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...

Adaptation of SVD and Prony method for precise computation of current components in networks with wind generation

precise computation of current components is a key prerequisite for reliable assessment of power quality. Especially in networks with wind generation we may observe increased number of possible disturbing phenomena. This paper presents an approach to accurate computation of currents components with two similar parametric methods based on singular value decomposition (SVD) and Prony model. Those methods seem to be applicable for the detection of non integer multiples of the main frequency in decaying signals. Results of both methods have been compared and evaluated. with respect to traditional Fourier method.

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.

Sizing of photovoltaic power and storage system for optimized hosting capacity

This paper presents an approach to optimized sizing of a combined photovoltaic and storage system in order to reduce fluctuations in local supply network and increase the hosting capacity of the distribution network, on the example of remote rural area. All the computations were based on real production and consumption data acquired over long period of time in two different test installations, located in Poland and Germany. Genetic algorithms were applied optimize corresponding sizes of PV and storage guarantying the research objective.

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.

Application of SVD method for signal parameters estimation in systems with DFIG

Signal parameters estimation is an important prerequisite for assessment of power quality (PQ) indices. Nowadays, large amounts of measured data need to be automatically processed for appropriate and useful data mining in PQ. Especially, modern wind generators are often seen as sources of PQ disturbances, which should be constantly supervised. The authors propose an application of modified singular value decomposition (SVD) method for signal parameters estimation. Results of the proposed method are compared with broadly used Fourier Transform. A mechanical model of doubly fed induction generator (DFIG) operating in various conditions was chosen as a source of disturbed signals. Research results verify the usefulness of SVD based method. Additionally, the dependence between operating mode of a DFIG and generated disturbance parameters was observed.

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.