Compressive Sensing-Based Harmonic Sources Identification in Smart Grids

Abstract

Identifying the prevailing polluting sources would help the distribution system operators in acting directly on the cause of the problem, thus reducing the corresponding negative effects. Due to the limited availability of specific measurement devices, ad-hoc methodologies must be considered. In this regard, Compressive Sensing-based solutions are perfect candidates. This mathematical technique allows recovering sparse signals when a limited number of measurements are available, and thus overcoming the lack of Power Quality meters.In this paper, a new formulation of the 1-minimization algorithm for Compressive Sensing problems, with quadratic constraint, has been designed and investigated in the framework of the identification of the main polluting sources in Smart Grids. A novel whitening transformation is proposed for this context. This specific transformation allows the energy of the measurement errors to be appropriately estimated and thus better identification results are obtained. The validity of the proposal is proved by means of several simulations and tests performed on two distribution networks for which suitable measurement systems are considered along with a realistic quantification of the uncertainty sources.

Enhanced PMU-Based Line Parameters Estimation and Compensation of Systematic Measurement Errors in Power Grids Considering Multiple Operating Conditions

“…where ϵ w = W Ch ϵ Γ is the whitened measurement error vector, having the identity matrix as covariance matrix. In order to obtain a suitable estimation of the measurement error norm, the method illustrated in [27] has been used and adapted to the specific problem. To solve (24), the method described in [28] has been followed (see Appendix B).…”

A Comprehensive Review of Harmonic Issues and Estimation Techniques in Power System Networks Based on Traditional and Artificial Intelligence/Machine Learning

“…It is possible to identify the harmonic sources in a power system by calculating the harmonic components injected into the network by each source, load, or generator for all the harmonic orders of interest. The research in [43] has the main objective of the identification of the harmonic sources since it will enable system operators to take immediate action against the origin of the issue. This strategy is known as harmonic source estimation (HSoE).…”

A Tuned Whitening-Based Taylor-Kalman Filter for P Class Phasor Measurement Units

“…For the same reason, even if ' ε σ can be chosen arbitrarily small, in practice it is pointless to choose a value lower than the largest uncertainty affecting the elements of ' 0 Λ . In any case, if 2 2 ' ε ε σ σ < , then the signal noise floor after disturbance whitening is certainly lower than before applying (10).…”

“…Even though the limits reported in the IEEE/IEC Standard were established mainly for transmission systems monitoring, at the moment they are usually a reference also for distribution-level PMUs. However, the inherent characteristics of distribution systems (e.g., the use of shorter lines with a lower X/R ratio than in transmission systems [9]) as well the presence of stronger harmonic and inter-harmonic interferers caused by a variety of sources pose two further crucial and contrasting problems [10]. On one hand, the phase angle differences between bus voltage phasors are very small (in the order of a few mrad) [11], which requires higher accuracy than traditional PMUs.…”