Dariusz Grabowski

Sizing of active power filters using some optimization strategies

Purpose – The change in the way of active power filters (APF) location can lead to overall cost reduction due to less number or less power of APFs required. The goal of this paper was to minimize the APF currents what is equivalent to solution with less apparent power of installed devices. The next step consists in development of new methods of APF optimal location.Design/methodology/approach – Some scripts integrating optimization and harmonic analysis methods in Matlab and PCFLO software environments have been developed in order to achieve the goal.Findings – Solution to the minimization problem determines the current spectrum of an APF connected to a selected system bus in accordance with some optimization strategies which among others enable minimization of THDV coefficients.Research limitations/implications – The APF control algorithm defined in the frequency domain and based on given current spectrum could lead to some problems with synchronization between APF instantaneous current and compensated c...

Strategies for optimal allocation and sizing of active power filters

Some new approaches to active compensator allocation and sizing in distribution networks have been proposed in the paper. A few objective functions as well as their advantages and disadvantages have been given. Moreover, software which enables practical verification of the strategies has been described.

Cost effective allocation and sizing of active power filters using genetic algorithms

The problem of cost effective active compensator allocation and sizing in distribution networks has been analyzed in the paper. A few objective functions have been compared from the economical point of view. Theoretical considerations have been illustrated by an example of active power filters (APF) allocation and sizing by means of the genetic algorithm.

Compensation based on active power filters - The cost minimization

The paper deals with the problem of power quality and its solution. Passive or active compensators are usually used to reduce distortions of voltage and current waveforms. Active power filters seem to be the best choice if the effect of compensation is the only criterion. However, the financial cost of such a solution put the potential investors off. The paper shows that the cost could be reduced considerably if the optimization of active power filter allocation, as well as their nominal currents, is carried out. Two cases have been considered - active power filters with a classical control algorithm, which operates locally and aims at obtaining sinusoidal shape of waveforms in the point of installation, and active power filters with a modified control algorithm, which aims at reduction of waveform distortions in all nodes of an analyzed power system. Goal functions and optimization problems for both cases have been defined.

Deterministic model of electric arc furnace – a closed form solution

Purpose – Electric arc furnaces are usually modelled using combined models which divide the phenomenon taking place in real objects into a deterministic and a stochastic or chaotic parts. The former is expressed by a nonlinear differential equation. The goal of this paper was to obtain a closed form of the solution to one of the most popular nonlinear differential equations used for the AC electric arc modelling.Design/methodology/approach – The solution has been obtained in the time domain by a sequence of transformations of the original nonlinear equation which lead to a linear equation, for which a closed form solution is known.Findings – The paper provides a set of parameters for which the solution to the nonlinear differential equation describing electric arc can be obtained in a closed form.Research limitations/implications – There are still some parameter values for which the solution can be found only numerically. Moreover, due to the nature of the phenomena occurring in electric arc furnaces, in ...

Analysis of deterministic model of electric arc furnace

A closed form of the solution to a differential equation used for the AC electric arc modeling has been derived in the paper. A special case of periodic excitations has been considered and the appropriate formulae have been given. Having a closed form of the solution rather than the numerical one enables better understanding of the phenomena described by the equation. Moreover, it makes easier the extension of the arc model in order to cover the time-varying character of the arc waveforms, e.g. using stochastic approach. Theoretical considerations have been illustrated by a simulation experiment.

Generalized Spectrum Analysis by Means of Neural Networks

The problem of determination of generalized signal spectrum using neural networks has been investigated. The paper contains the formalization of the problem and is focused on building of a neural network that may be used to calculate coefficients of the generalized Fourier series for various base functions. The theoretic considerations have been illustrated by an example of analysis of a signal belonging to the Hilbert space of signals with finite energy for Laguerre’s base and a wavelet base.

Genetic and Combinatorial Algorithms for Optimal Sizing and Placement of Active Power Filters

Abstract The paper deals with cost effective compensator placement and sizing. It becomes one of the most important problems in contemporary electrical networks, in which voltage and current waveform distortions increase year-by-year reaching or even exceeding limit values. The suppression of distortions could be carried out by means of three types of compensators, i.e., passive filters, active power filters and hybrid filters. So far, passive filters have been more popular mainly because of economic reasons, but active and hybrid filters have some advantages which should cause their wider application in the near future. Active power filter placement and sizing could be regarded as an optimization problem. A few objective functions have been proposed for this problem. In this paper we compare solutions obtained by means of combinatorial and genetic approaches. The theoretical discussion is followed by examples of active power filter placement and sizing

Neural Approach to Time-Frequency Signal Decomposition

The problem of time-frequency decomposition of signals by means of neural networks has been investigated. The paper contains formalization of the problem as an optimization task followed by a proposition of recurrent neural network that can be used to solve it. Depending on the applied base functions, the neural network can be used for calculation of several standard time-frequency signal representations including Gabor. However, it can be especially useful in research on new signal decompositions with non-orthogonal bases as well as a part of feature extraction blocks in neural classification systems. The theoretic considerations have been illustrated by an example of analysis of a signal with time-varying parameters.

Time-invariant and time-varying filters versus neural approach applied to DC component estimation in control algorithms of active power filters

Abstract This paper presents an application of digital filters and neural networks to the extraction of a DC signal component. This problem arises, among others, in control of active power filters (APF) used for power quality improvement. Solutions to the basic problem of DC component estimation are well-known and so the difficulty of the task comes rather from the required minimization of the calculation time. It should ensure fast reaction of the control system to load changes. As a result, lower value of the current total harmonic distortion coefficient (THD) and better efficiency of the APF can be obtained. The paper includes propositions of both time-varying and neural filters as well as comparison with the classical approach to the DC component estimation based on infinite impulse response (IIR) low-pass filters. The results obtained by simulations have been presented. The future work will be devoted to digital signal processor implementation of APF control algorithms based on the best solution.