Robust Optimal Allocation of Decentralized Reactive Power Compensation in Three-Phase Four-Wire Low-Voltage Distribution Networks Considering the Uncertainty of Photovoltaic Generation

• Published in: Energies (Basel) Vol. 12; no. 13; p. 2479
• Main Authors: Lin, Shunjiang, He, Sen, Zhang, Haipeng, Liu, Mingbo, Tang, Zhiqiang, Jiang, Hao, Song, Yunong
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 2019
• Subjects: Arrays; Compensation; distributed photovoltaics; Distribution management; Electric potential; Energy management; Inverters; Light intensity; Load; low-voltage distribution network; Lower bounds; Luminous intensity; Maximum power; Optimization; Photovoltaics; Power factor; Power loss; Process controls; Reactive power; reactive power optimization; robust optimization; Switching theory; Temperature; three-phase four-wire; uncertainty; Voltage; China
• ISSN: 1996-1073; 1996-1073
• DOI: 10.3390/en12132479

Due to the unbalanced three-phase loads, the single-phase distributed photovoltaic (PV) integration, the long feeders, and the heavy loads in a three-phase four-wire low voltage distribution network (LVDN), the voltage unbalance factor (VUF), the network loss and the voltage deviation are relatively high. Considering the uncertain fluctuation of the PV output and the load power, a robust optimal allocation of decentralized reactive power compensation (RPC) devices model for a three-phase four-wire LVDN is proposed. In this model, the uncertain variables are described as box uncertain sets, the three-phase simultaneous switching capacity and single-phase independent switching capacity of the RPC devices are taken as decision variables, and the objective is to minimize the total power loss of the LVDN under the extreme scenarios of uncertain variables. The bi-level optimization method is used to transform the robust optimization model with uncertain variables into bi-level deterministic optimization models, which could be solved alternately. The nonlinear programming solver IPOPT in the mature commercial software GAMS is adopted to solve the upper and lower deterministic optimization models to obtain a robust optimal allocation scheme of decentralized RPC devices. Finally, the simulation results for an actual LVDN show that the obtained decentralized RPC scheme can ensure that the voltage deviation and the VUF of each bus satisfied the secure operation requirement no matter how the PV output and load power changed within their own uncertain sets, and the network loss could be effectively reduced.