This paper addresses the voltage control problem in medium-voltage distribution networks. The objective is to cost-efficiently maintain the voltage profile within a safe range, in presence of uncertainties in both the future working conditions, as well as the physical parameters of the system. Indeed, the voltage profile depends not only on the fluctuating renewable-based power generation and load demand, but also on the physical parameters of the system components. In reality, the characteristics of loads, lines and transformers are subject to complex and dynamic dependencies, which are difficult to model. In such a context, the quality of the control strategy depends on the accuracy of the power flow representation, which requires to capture the non-linear behavior of the power network. Relying on the detailed analytical models (which are still subject to uncertainties) introduces a high computational power that does not comply with the real-time constraint of the voltage control task. To address this issue, while avoiding arbitrary modeling approximations, we leverage a deep reinforcement learning model to ensure an autonomous grid operational control. Outcomes show that the proposed model-free approach offers a promising alternative to find a compromise between calculation time, conservativeness and economic performance.

This paper presents a new technique for voltage regulation of a radial medium voltage (MV) distribution grid in presence of distributed generation (DG) units. The proposed technique consists of the coordinated actions of on-load tap changer (OLTC) of transformer and reactive power compensation by distribution static synchronous compensator (D-STATCOM). Managing the system voltage using the action of OLTC is one of the most common ways for voltage regulation of MV systems. However, OLTC cannot be used for voltage regulation of long radial distribution feeders as it changes the sending point voltage of the feeder. In this study, the problem of using OLTC for the voltage regulation of a radial distribution feeder will be solved by using reactive power compensation at the DG connected bus. Simulation results reveal that the proposed control method is capable of maintaining the system voltages within the permitted range. Thanks to the coordinated control method, the drawbacks of each voltage control method are covered by the other one that results in minimization of the system losses as well.

—In this paper, a novel voltage sensitivity analysis method is proposed. It presents a complementary formulation of the direct sensitivity analysis approach which has been previously developed based on the topological structure of the network. The proposed method named improved direct sensitivity analysis (IDSA) incorporates variations of power losses in the system branches due to the nodal power changes and their eventual impacts on the node voltages. Effectiveness of the IDSA in voltage estimation is investigated and compared with the voltage results obtained by the direct, Jacobian-based, as well as the perturb-and-observe sensitivity analysis methods. To this end, firstly, the introduced sensitivity analysis methods are tested when active or reactive power is changed at the selected nodes of the studied test systems. Accuracy of voltage responses obtained by each of the considered sensitivity analysis methods is evaluated with respect to the exact voltage value obtained from the load flow study. Moreover, performance of the introduced sensitivity analysis methods is examined when they are separately embedded in a multi-step voltage control algorithm which manages active and reactive powers of distributed generation units in order to keep the system voltages within the permitted voltage limits. Simulation results confirm that when the power losses impact is considerable, the IDSA outperforms the direct, perturb-and-observe, and Jacobian-based sensitivity analysis methods in terms of accuracy of the voltage estimation.

Low voltage distribution networks have not been traditionally designed to accommodate the large-scale integration of decentralized photovoltaic (PV) generations. The bidirectional power flows in existing networks resulting from the load demand and PV generation changes as well as the influence of ambient temperature led to voltage variations and increased the leakage current through the cable insulation. In this paper, a machine learning-based framework is implemented for the identification of cable degradation by using data from deployed smart meter (SM) measurements. Nodal voltage variations are supposed to be related to cable conditions (reduction of cable insulation thickness due to insulation wear) and to client net demand changes. Various machine learning techniques are applied for classification of nodal voltages according to the cable insulation conditions. Once trained according to the comprehensive generated datasets, the implemented techniques can classify new network operating points into a healthy or degraded cable condition with high accuracy in their predictions. The simulation results reveal that logistic regression and decision tree algorithms lead to a better prediction (with a 97.9% and 99.9% accuracy, respectively) result than the k-nearest neighbors (which reach only 76.7%). The proposed framework offers promising perspectives for the early identification of LV cable conditions by using SM measurements.

In this paper, a novel framework is proposed in order to evaluate impacts of the uncertain models of the system components on the voltage regulation problem of the medium‐voltage distribution systems. The investigation focuses on the model uncertainty associated with voltage dependency of loads, power factor of loads, thermal dependency of lines, shunt admittances of lines and internal resistance of substation transformer. To this end, firstly, voltage constraints are managed using a centralised voltage control algorithm (VCA) by relying on the simplified models of the system components. The system loads and lines as well as the substation transformer are then modelled with the uncertain variables which are bounded in the predefined ranges. Monte Carlo (MC) simulations are used to create wide series of scenarios that cover the possible values that the parameters of the system components can take due to their uncertain nature. The model uncertainty impacts on the voltage regulation problem are finally evaluated by the load flow calculations considering the scenarios created by the MC simulations and the set‐point obtained by the VCA. The proposed investigation brings useful information regarding the possible deviations that the node voltages can have due to the uncertain models of the studied components.