Search

This paper proposes a novel data driven framework for detecting topology transitions in a distribution network. The framework analyzes data from phasor measurement units (PMUs) and relies on the fact that changes in network topology results in changes in the structure and admittance of the network. Using voltage and current phasors recorded by PMUs, the proposed method approximates network parameters using an ensemble-based deep learning model and thus, it does not require any knowledge of network parameters and load models. Using the prediction error of the proposed model, a connectivity matrix which shows the status of switches is constructed. In contrast to other methods, this proposed framework does not require a library of voltage and current transients associated with possible network transitions. It can also detect simultaneous switching actions and is robust to noise and load variations. The model yields a lower error detection rate, and its performance is validated using a modified version of the IEEE 33 bus network and a real feeder located in Queensland, Australia, under full and partial observability conditions. The proposed model has also been compared with another data driven method in terms of inference time and error detection rates. [ABSTRACT FROM AUTHOR]

Volt/VAr Optimization (VVO) is becoming increasingly crucial in distribution management systems with renewable sources that require setting their reactive power in coordination with legacy voltage and reactive power control devices. This paper presents a VVO method that operates in complex variables based on the Wirtinger calculus; it employs a penalty method that keeps voltage magnitudes and controller parameters within limits together with a probabilistic rounding technique for handling switched device variables. The complex variable implementation permits using the compensation technique when solving the Karush-Kuhn Tucker (KKT) system, thus achieving a significant speedup due to limiting the number of linear system factorizations. The proposed method is contrasted with a classical VVO implementation employing discrete coordinate search from the current operating point and an enhanced version that uses sensitivity information. Numerical results on distribution networks having up to 3147 nodes show that the proposed method is significantly faster than classical methods and gives operating points free of voltage magnitude violations and lower power loss. [ABSTRACT FROM AUTHOR]

This paper proposes a distributed coordinated voltage control scheme for distribution networks with distributed generation (DG) and on-load tap changer (OLTC). In this scheme, static synchronous compensators (STATCOMs), DG units and OLTC are coordinated to regulate voltages of all buses to be close to the nominal value in the distribution network, mitigate voltage fluctuations, and minimize the number of operations of OLTC while considering different temporal characteristics of voltage regulation devices. The optimization problem of coordinating DG units and STATCOMs is decomposed by the gradient projection (GP) method. The local controller optimizes the reactive power outputs of DGs and STATCOMs according to local voltage and reactive power measurements, and still achieves the optimal coordination of DG units and STATCOMS in a decentralized manner without a central controller or communication between local controllers. The OLTC control scheme is designed to correct the long-term voltage deviations based on model predictive control (MPC) while minimizing the number of operations. The local controllers send the calculated reactive power references of DG and STATCOMs to the OLTC controller, which achieves distributed coordinated voltage control and mitigates the computation burden. A distribution network with two 20 kV feeders and 8 DG units was used to validate the control performance of the proposed coordinated voltage control scheme. [ABSTRACT FROM AUTHOR]

Soft normally open points (SNOPs) connected to distribution networks are instrumental in maintaining uninterrupted power supply and improving power quality during the fault period. In order to achieve the optimal service restoration scheme, a novel bi-level service restoration model based on the robust optimization method is proposed in this paper. The upper level of the proposed model aims at minimizing the risk of load loss. The switch statuses and range of power transmitted by SNOP terminals can be obtained. In the lower level of the model, the interval robust optimization method is adopted to cope with the uncertainties of DGs and loads. If there are no optimal solutions, the upper level model needs to be solved once more to adjust the network topology. Finally, a test system is established to verify the effectiveness of the proposed model and solution method. The effects of forecasting errors, the DG penetration, and the ratio of single terminal capacity of SNOP to line capacity on the optimization results are analyzed. [ABSTRACT FROM AUTHOR]

The uptake of variable megawatts from photovoltaics (PV) challenges distribution system operation. The primary problem is significant voltage rise in the feeder that forces existing voltage control devices such as on-load tap-changers and line voltage regulators to operate continuously. The consequence is the deterioration of the operating life of the voltage control mechanism. Also, conventional non-coordinated reactive power control can result in the operation of the line regulator at its control limit (runaway condition). This paper proposes an optimal reactive power coordination strategy based on the load and irradiance forecast. The objective is to minimize the number of tap operations so as not to reduce the operating life of the tap control mechanism and avoid runaway. The proposed objective is achieved by coordinating various reactive power control options in the distribution network while satisfying constraints such as maximum power point tracking of PV and voltage limits of the feeder. The option of voltage support from PV plant is also considered. The problem is formulated as constrained optimization and solved through the interior point technique. The effectiveness of the approach is demonstrated in a realistic distribution network model. [ABSTRACT FROM PUBLISHER]