Nodes in electric distribution networks are greatly differentiated and are very often nonlinear and/or unbalanced. They can create significant harmonic pollution, with harmonics that inevitably spread along the grid. Monitoring harmonic propagation and correlated power quality phenomena requires specific measurement devices and methodologies. Nevertheless, because of the unavailability of a rapid diffusion of synchronized and dedicated devices (due to technical and economic reasons) on every node and branch of the network, estimating the harmonic status of the entire grid by means of a complete or even redundant monitoring system can be practically unfeasible. A more feasible, though always meaningful, goal can thus be pursued, that is estimating the main harmonic sources in the network, rather than its complete harmonic status. This approach, of course, can be based on a simpler and cheaper upgrade of the distributed monitoring system. Even more, by considering the common scenario where the number of significant harmonic sources is lower than the number of loads connected to the grid, specific estimation procedures can be defined to further reduce the complexity of the monitoring system. In this scenario, this paper presents an efficient compressive sensing harmonics detector (CSHD) for the identification and the estimation of the principal pollution sources. The proposed CSHD method is validated by means of appropriate tests performed on an example of distribution grid.

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 ℓ1-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.

In this paper, an approach to perform model reduction of distribution grids for phasor measurement unit (PMU)-based state estimation (SE) purposes is presented. The method is based on ac load flow equations. An arbitrary number of lines and nodes can be selected and reduced to an equivalent one. The performance of the method is tested using the IEEE34 and IEEE123 test cases. An evaluation of the computational performance of the SE process is also reported. To conclude, the SE process based on the reduced model is tested against uncertainty in the parameters of the system and in the measurements so to evaluate how the performance of the SE is deteriorated by these phenomena.

This paper proposes a new planning framework to determine the optimal location, capacity, and types of EV charging stations (EVCSs) in multi-energy systems (MESs). We propose a two-stage stochastic programming approach -with scenario-based algorithms-that explicitly considers vehicle-to-grid (V2G) peculiarities (four-quadrant operation and stochastic human factors influence: V2G willingness, walking distance, and charging patterns). Considering those factors together with MES uncertainties -RES generation, load demands, and electricity price-enables a comprehensive study of V2G and MES impact on EVCS planning. The proposed approach is applied to both a purely electric distribution network (EDN) and an MES to analyze the interplay of EVCSs in different energy domains, in consideration of different V2G contracts. The obtained results underline that the sole consideration of the EDN can lead to non-optimal results, while the more comprehensive analysis leads to optimal planning of all energy resources and cost savings. Finally, we analyse how each considered factor individually impacts EVCSs planning.

Non-intrusive load monitoring (NILM) is a process of determining the operating states and the energy consumption of single electric devices using a single energy meter providing aggregate load measurements. Due to the large spread of power electronic-based and nonlinear devices connected to the network, the time signals of both voltage and current are typically non-sinusoidal. The effectiveness of a NILM algorithm strongly depends on determining a set of discriminative features. In this paper, voltage and current signals were combined to define, according to the definitions provided in Standard IEEE 1459, different power quantities, that can be used to distinguish different types of appliance. Multi-layer perceptron (MLP) classifiers were trained to solve the appliance detection problem as a multi-class event classification problem, varying the electric features in input. This allowed to select an optimal set of features guarantying good classification performance in identifying typical electric loads.

Grid management undoubtedly benefits from an accurate monitoring of actual network operating conditions. Such monitoring can be obtained starting from a widespread presence of measurement points and different types of estimation techniques. To exploit all the possible measurements present on the network, data coming from the protection systems can be considered in addition to those from the measurement systems. In this scenario, however, accuracy should not be underestimated, because in the presence of dynamic signals it can be dramatically reduced. In this context, this paper presents an improved fault detection and location method based on synchronized measurements, estimation techniques and an appropriate accuracy modeling aimed at reducing the uncertainty in the fault analysis. The proposed approach is validated by simulations carried out by means of a Digital Real Time Simulator (DRTS) on the three-phase CIGRE European Medium Voltage distribution network.