Search

Implementing accurate Distribution System State Estimation (DSSE) faces several challenges, among which the lack of observability and the high density of the distribution system. While data-driven alternatives based on Machine Learning models could be a choice, they suffer in DSSE because of the lack of labeled data. In fact, measurements in the distribution system are often noisy, corrupted, and unavailable. To address these issues, we propose the Deep Statistical Solver for Distribution System State Estimation (DSS^2), a deep learning model based on graph neural networks (GNNs) that accounts for the network structure of the distribution system and the governing power flow equations of the problem. DSS^2 is based on GNN and leverages hypergraphs to model the network as a graph into the deep-learning algorithm and to represent the heterogeneous components of the distribution systems. A weakly supervised learning approach is put forth to train the DSS^2: by enforcing the GNN output into the power flow equations, we force the DSS^2 to respect the physics of the distribution system. This strategy enables learning from noisy measurements and alleviates the need for ideal labeled data. Extensive experiments with case studies on the IEEE 14-bus, 70-bus, and 179-bus networks showed the DSS^2 outperforms the conventional Weighted Least Squares algorithm in accuracy, convergence, and computational time while being more robust to noisy, erroneous, and missing measurements. The DSS^2 achieves a competing, yet lower, performance compared with the supervised models that rely on the unrealistic assumption of having all the true labels., Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public., Multimedia Computing, Intelligent Electrical Power Grids

In this paper, a real-time state estimation platform for distribution grids monitored by Phasor Measurement Units (PMUs) is developed, tested, and validated using Real Time Digital Simulator (RTDS). The developed platform serves as a proof-of-concept for potential implementation in an existing 50 kV ring network of the Dutch distribution utility Stedin medium voltage distribution grid located in the southwest (Zeeland area) of the Netherlands. To catch up with the fast sampling rates of PMUs, the platform incorporates computationally efficient techniques for state estimation and detection, discrimination and identification of anomalies like bad data and sudden load changes. Forecasting Aided State Estimation has been utilized to enable measurement innovations needed for fast anomaly detection, discrimination, and identification, whilst the Extended Kalman Filter (EKF) algorithm is selected to provide fast state forecasting and filtering. The platform has been tested under various normal and abnormal operating conditions considering different statistical properties of measurement noise as well as different bad data and sudden load change scenarios. To demonstrate advantages and disadvantages for embedding EKF into the platform, EKF is compared with Unscented Kalman Filter (UKF) in terms of estimation accuracy, computational efficiency, and compatibility with the module for anomaly detection, discrimination, and identification. The results of extensive simulations provide good hints about the feasibility of PMU-based real-time state estimation for the Stedin distribution grid., Intelligent Electrical Power Grids

This study proposes a method for the determination of flicker contribution based on synchronised measurements of root mean square (RMS) voltages and currents. The advantage of this approach is the avoidance of time aggregation which is present in the analysis based on the short- or long-term flicker coefficient. The applicability of the method is demonstrated using phasor measurement unit measurements from a 50 kV network and compared with the results of an analysis based on instantaneous flicker coefficients.

The installation of phasor measurement units (PMUs) by Enduris, one of the distribution network operators in the Netherlands to monitor their 50 kV ring distribution grid has resulted in increased observability of the power flows and voltages. However, with the increased observability, there is a need to analyse the distribution grid based on the data generated by the PMU monitoring system. This study provides steady-state and dynamic analysis of the 50 kV Enduris ring distribution grid in terms of voltage and power flow behaviour based on the real-time data acquired from the PMUs. First, the Enduris 50 kV distribution grid under study is modelled in PowerFactory software. This model is subsequently used for analysis of the distribution grid behaviour for two contingencies. The contingencies are actual grid events observed by the PMUs, and using this data, the contingencies are simulated and analysed. The analysis results of each contingency are compared with the data from the PMU monitoring system., Session 4: Distributed energy resources and active demand integration, Intelligent Electrical Power Grids

Part of the energy transition will take place in residential areas. In order to investigate the decisions consumers have to make in the coming decade and their consequences, a forecasting model is designed. By using characteristics of the environment, a mix of energy systems for a complete neighbourhood is calculated. The corresponding low-voltage (LV) network is analysed by performing capacity calculations with the forecasted mix as input. From the case study on a typical neighbourhood, it becomes clear that bottlenecks can be expected in typical LV networks in the coming decades. By using scenarios, the outcomes are given for multiple projected futures.