Search

Passive and active filters have traditionally been used to reduce harmonics in power systems. Only recently combined systems of shunt passive and active filters have been proposed in order to overcome the technical problems of passive filters and to reduce the costs of active ones. The paper tries to overcome some drawbacks of a combined filtering system (modified high-pass filter) which appeared in literature. In particular, improvements in the system structure and in the steady-state design methodology are proposed to obtain a filtering system with both satisfactory compensation characteristics and low costs. Finally, numerical applications are discussed considering a test case.

The paper deals with Smart Grids including renewable generation units, in particular concentrated photovoltaic systems and plug-in vehicles fleets. DG units and fleets are interfaced to the grid through static converters. Several services are performed thanks to the converters presence, with objectives aimed at meeting needs either internal or external to the Smart Grid. Optimization constrained problems are formulated and properly solved in order to perform the services in the best economical and technical way. Among the objectives to meet, the proposed optimal control strategy includes Power Quality issues such as the minimization of waveform distortions as well as the optimization of the voltage profile at fundamental frequency. A numerical application on a 30-bus low voltage Smart Grid test system gives evidence of the effectiveness of the proposed procedure.

Some evaluation methods to obtain the probability density functions of voltage and current harmonics in multiconvertor power systems are presented. The approaches are based on Monte Carlo simulation applied to the non linear and linear harmonic power flow equations, on convolution process and on Pearson distributions. These methods are compared in terms of time execution and accuracy in the evaluation of harmonic probability density functions, and in particular of their 95 percentiles and maximum values, being these quantities the statistical measures of greatest interest in many International harmonic Standards and Recommendations. Several numerical applications are presented and discussed with reference to a multiconvertor test system.

This is a companion paper to a paper of the same title, Part I, in which the theoretical aspects of some spectrum estimation methods for analysis of nonstationary voltages and currents in power systems with nonlinear loads are analysed. In particular, the Prony model and rootMusic method are taken into account. In this work, numerical applications of the methods presented in Part I are discussed, in order to illustrate their implementation problems and practical applications. To investigate the methods, several experiments were performed using nonstationary voltage and current waveforms in a supply system of a DC arc furnace. For comparison, similar experiments were repeated using the short-time Fourier transform. The comparison proved the effectiveness of the new methods; however, their computational efforts can be much more complex. Conclusions and recommendations are made concerning the framework of each method.

In this paper, a probabilistic method is proposed to analyze the very short-term steady-state performance of an unbalanced distribution electrical system characterized by the presence of wind farms. This method, which can take into account the uncertainties of loads and wind productions, is based on a Monte Carlo simulation procedure applied to the non-linear three-phase load flow equations, including wind farm models. Bayesian time series models are used to predict the next hour's wind speed probability density functions, making possible a predictive evaluation of the very short-term system steady-state behavior. Numerical applications are presented and discussed with reference to the three-phase unbalanced IEEE 34-bus test distribution system in the presence of wind farms connected at different busbars.

The problem of voltage disturbance compensation in an industrial power system is considered. Reference is made to a case-study where a transformer feeds an AC/DC boost power converter, a capacitor VAr compensator and a balanced AC load. The transformer supplying voltages are supposed to be affected by flicker, slow variations and unbalance. The proposed perturbation compensation consists in suitably controlling the boost power converter which so acts as both a rectifier and static VAr compensator. The power converter sizing is discussed and the controller design presented. The effectiveness of the proposed compensation is shown via simulations obtained by means of the SIMULINK implementation of the power system dynamic model.

The inherent structure theory of networks, already applied in the literature at the fundamental frequency to a variety of planning problems, is extended to the field of power system harmonics. In the context of this theory, the harmonic admittance matrices are reformulated in terms of their eigenvalues and eigenvectors, showing that these expansions can be useful to create a framework in which the power system response to harmonic disturbances can be better understood. After the presentation of the theoretical background, some power system harmonic problems are tackled to show possible applications of this theory. Numerical experiments on a 17-busbar distribution system give evidence of the significance of application to the field of power system harmonics.

Representation of signals in time and frequency domain has been of interest in power systems for many years, especially in the case of nonstationary signals. Standard tools of waveform distortion analysis are based on the Fourier transform while only recently the Prony and rootMusic methods were tested for this purpose. However, these methods show a strong sensitivity to the number of spectral components when applied to evaluate waveform distortion indices. This paper investigates this problem in order to evidence their optimal number. Numerical simulations on test waveforms and on DC arc furnace current waveforms allow to point out the most suitable choices.