Multi-Objective Adaptive Robust Voltage/VAR Control for High-PV Penetrated Distribution Networks

Voltage/VAR control (VVC) methods implemented by on-load tap-changers, capacitor banks and photovoltaic (PV) associated inverters can efficiently maintain power quality (re-duce voltage deviation) and reduce energy loss for active distribution networks (ADNs). However, minimization of voltage deviation is in competition with minimization of energy loss. Besides, uncertainties such as PV power generation and load demand impair control results. To address these issues, this paper proposes a multi-objective VVC model based on a multi-stage coordination framework for ADNs, aiming to minimize voltage deviation and energy loss simultaneously. More importantly, this paper proposes a multi-objective robust optimization approach to robustly optimize the multi-objective VVC model against uncertainties. Correspondingly, a solution algorithm based on adaptive weighted-sum and column-and-constraint generation algorithms is developed to solve the multi-objective robust optimization problem. Simulation results show well distributed solutions and high solution robustness of the proposed multi-objective robust VVC against uncertainties.

This paper presents a three-stage inverter-based peak shaving and Volt-VAR control (VVC) framework in active distribution systems using the online safe deep reinforcement learning (DRL) method. The proposed framework aims to reduce the peak load, voltage violations, and real power loss by coordinating three stages with different control timescales. In the first stage, a day-ahead charging/discharging scheduling of energy storage systems (ESSs) with a 30 min resolution is performed via their inverters for peak shaving. In the second stage, the discharging power of ESSs is adjusted through measurements with a 1 min resolution to completely shave peak loads. A model-free DRL algorithm integrated with a safety module is also implemented in the second stage. Using this algorithm, the reactive powers of photovoltaic (PV) systems and ESSs are controlled by the DRL agent to reduce the voltage violation and real power loss, whereas no voltage violation occurs during the online training process. In the third stage, a proportional-integral controller with real-power compensation is integrated into inverters of PV systems and ESSs to rapidly mitigate local voltage violations with a 0.1 s resolution. The high efficiency and safety of the proposed method were validated on the IEEE 33-bus and IEEE 123-bus systems.

Modern power systems are heading towards a sustainable future with rapidly increasing penetration of distributed renewable energy sources such as solar photovoltaics (PV), especially in active distribution networks (ADNs). However, temporal and spatial variations and intermittency of renewable power generation bring challenges on voltage control. As controllable VAR sources, PV associated inverters can supply flexible and fast-responsive reactive power to achieve volt-age/V AR control (VVC) in the ADNs. Conventional VVC mainly relies on rule-based, mathematical or heuristic methods, which can be inefficient or even infeasible as the system becomes large and when complex uncertainty conditions are considered. To this end, this paper proposes a deep reinforcement learning (DRL) based optimization model for the inverter-based VVC under uncertainties. The DRL agent learns to optimally determine inverter reactive power output setpoints through exploration in the virtual environment and neural network training, with multiple objectives of minimizing bus voltage deviations, PV active power curtailments and network power losses. Temporally and spatially uncertain PV power generation and loads are fully considered in the proposed DRL model. A numerical case study on the IEEE 123-bus test system has indicated the VVC decisions are efficiently obtained and are robust against uncertainty realization.

Joint optimization of Volt/VAR control and mobile energy storage system scheduling in active power distribution networks under PV prediction uncertainty

Multi-objective voltage/VAR control for integrated port energy system considering multi-network integration

In recent years, the violation and fluctuation of system voltage has occurred with greater frequency with the integration of high-penetration distributed photovoltaic generation. In this paper, the voltage violation and fluctuation in a high-penetration distributed photovoltaic integrated system is analyzed, and then a corresponding suppression strategy is proposed. Firstly, based on solar cell and photovoltaic control system models, the influence factors of photovoltaic output are analyzed. Secondly, the voltage violation and fluctuation caused by photovoltaic integration is analyzed, and the quadratic parabola relationship between bus voltage fluctuation and photovoltaic power variation is constructed. Next, according to the virtual synchronous generator characteristic of distributed photovoltaics, a double-hierarchical suppression strategy is proposed to make full use of reactive power regulation capability, which can maintain the symmetry of power supply while meeting standard requirements. The proposed strategy can conveniently realize quick response and support the photovoltaic extensive access. Moreover, with the employment of the proposal, the system voltage violation and fluctuation can be suppressed effectively. Finally, considering the photovoltaic access location, capacity, and partial shading, the effectiveness of the proposed strategy is verified in IEEE 33-bus distribution system with field measured data. After distributed photovoltaic accesses the system, more than 60% of buses appear to have undergone bus voltage violation. With the proposed method, more than 20% of the voltage deviation and more than 6% of the voltage fluctuation are effectively suppressed so that the system voltage can be kept below 1.07 p.u. and the voltage fluctuation can be kept within 4%, meeting the requirements of power quality standards.

Kriging surrogate model enabled heuristic algorithm for coordinated Volt/Var management in active distribution networks

Advancements in data-driven voltage control in active distribution networks: A Comprehensive review

The high penetration of intermittent renewable energy resources in active distribution networks (ADN) results in a great challenge for the conventional Volt-Var control (VVC). This article proposes a two-stage deep reinforcement learning (DRL)-based real-time VVC method to mitigate fast voltage violation while minimizing the network power loss. In the first stage, on-load tap changer (OLTC) and capacitor banks (CBs) are dispatched hourly based on the optimal power flow method. The optimization problem is formulated as a mixed-integer second-order cone programming (MISOCP) which can be effectively solved. In the second stage, the reactive power of photovoltaics (PVs) is regulated dynamically to mitigate fast voltage fluctuation based on the well-learned control strategy and local measurements. The real-time VVC problem is formulated and solved using a multi-agent deep deterministic policy gradient (MADDPG) method, which features offline centralized training and online decentralized application. Rather than using the critic network to evaluate the output of the actor-network, the gradient of the action-value function to action is derived analytically based on the voltage sensitivity method. The proposed approach is tested on the IEEE 33-bus distribution system and comparative simulation results show the enhanced control effect in mitigating voltage violations.

Source-grid-load-storage has represented an interactive characteristic in the active distribution network (ADN). Moreover, power electronic devices have been widely used for source-grid-load-storage with the rapid development of power electronics technology. In this condition, the large-scale distributed source may cause voltage quality degradation, while the application of large-scale power electronics equipment may also lead to serious harmonic distortion. Power quality has become one of the most important issues in the development of ADN. In this paper, the source-grid-load-storage interactive power quality characteristic of the ADN is analyzed. Firstly, considering the source-grid-load-storage interaction in ADN, the voltage deviation and fluctuation are analyzed and the degrees are further quantified. Then, the source-load-storage harmonic models for the power electronic components are built, which is the basis for harmonic analysis. Moreover, the decoupled harmonic power flow algorithm for ADN is proposed to analyze the system harmonic distributions. Finally, considering the location and capacity of photovoltaic and energy storage, the interaction and power quality are analyzed in the IEEE 33-bus distribution system. With the access of energy storage, more than 20% of the voltage deviation and more than 6% of the voltage fluctuation caused by photovoltaics are effectively suppressed, while the harmonic distortion may be further increased.

This paper proposes a two-stage combined central and distributed Volt/Var control (VVC) strategy to coordinate both utility-scale and customer-owned photovoltaics (PVs) for voltage regulation in active distribution networks (ADN). In the first stage, with day-ahead PV and load predictions, the central controller (CC) performs optimal power flow (OPF) to determine the hourly dispatch of on-load tap changers (OLTC) and capacitor banks (CBs). In the second stage, the ADN is partitioned into different zones only based on PV and load types. The zone controllers (ZC), which is formulated by deep neural networks (DNN), then communicate with their neighbors and track the optimal total reactive power support of PVs in each zone. In the offline training process, an analytical derivation to minimize power losses is also provided that can help the input and output selection of DNN. To online dispatch reactive power for individual utility-scale and customer-owned PVs in corresponding zones, a multi-mode reactive power allocation strategy is proposed. Three different modes namely loss reduction mode (LRM), fair power sharing mode (FPM) and voltage security mode (VSM) are designed to meet varying operation requirements, which can mitigate voltage violations with fair participation for customers and reduced power losses for system operators. Numerical simulations are performed on the IEEE 33-bus distribution system to demonstrate the effectiveness of the proposed method in finding a trade-off between voltage security and loss minimization with fairness.

Volt/VAR control (VVC) techniques may be used for purposes such as technical loss reduction, voltage profile improvement, and conservation voltage reduction. The advent of distributed energy resources (DERs) at distribution and consumer levels imposes imperative VVC challenges for the distribution network operator. Innovative approaches have been proposed to use the inherent thermal inertia of buildings to provide ancillary services to the grid to tackle the problems posed by increasing trend of volatile DERs. Numerous state-of-the-art VVC strategies utilize traditional VVC devices and smart inverters to deal with voltage violations in active distribution networks (DNs) with increased DER penetration. However, they have not considered the potential service buildings can provide to mitigate this problem. The ability of smart buildings to provide reactive power support to the grid has not been exploited to date. Hitherto, the effect of the modulation of loads’ reactive power consumption on the grid's voltage profile has not been studied. In this article, a distributed model-predictive control strategy for VVC in the DN by utilizing smart buildings is presented. The robustness of this strategy is validated on a modified IEEE 13-bus system.

The penetration of photovoltaics (PVs) has been increasing in active distribution networks (ADN), which leads to severe voltage violation problems. PV inverters can provide fast and flexible reactive power support and are now allowed to participate in the voltage regulation process. This paper proposes a real-time combined central and local Volt/Var control (VVC) strategy to mitigate voltage violation problems while minimizing the network power loss. Based on the historical PV and load data, the load flow and optimal power flow are performed in the centralized controller (CC) to obtain multiple voltage and optimal power settings for each PV system. The local controller (LC) then generates voltage control curves with these optimal scatters. To improve the voltage control effect, a novel 3-Dimension voltage control curve is proposed considering both the measurements of node voltage and PV generation. Moreover, a data-driven deep convolution neural network is designed and trained to generate optimal local voltage control curves without prior knowledge of specific curve functions. The proposed approach is tested on the IEEE 33-bus distribution system and simulation results verify the effectiveness in voltage control compared with the optimal Q(V) and Q(P) method.

The increasing penetration of uncontrollable distributed generators (NDGs) exacerbates the risk of voltage violations in active distribution networks (ADNs). It is difficult for a centralized control strategy to meet the requirements of fast voltage and reactive power control because of the heavy computational and communication burdens. Local voltage control based on real-time measurements can respond quickly to the frequent fluctuations of distributed generators (DGs). In this paper, a local voltage control strategy of DGs with reactive power optimization based on a kriging metamodel is proposed. First, to build the metamodel for local voltage control, the steps for determining the input variables are presented in detail, and the effects of different variables on the accuracy of the metamodel are analyzed. Then, taking minimum network losses and voltage deviations as the objective function, we construct the metamodel for local voltage control based on kriging methods. Finally, operation strategies for DGs are developed by calculating the optimally weighted vector based on real-time measurements, and the operation strategies for DGs will be added into the original sample set to improve the accuracy of the metamodel. The proposed local voltage control strategy based on only the local measurements can quickly respond to the frequent DG fluctuations, reduce the communication burden for large networks and improve the adaptability of local voltage control in ADNs. Case studies under different scenarios on the IEEE 33-node system and the IEEE 123-node system are conducted to verify the effectiveness of the proposed method, and the results show that the proposed method can effectively solve the problems of voltage deviation and voltage fluctuation caused by the high penetration of DGs.

Generative adversarial network assisted stochastic photovoltaic system planning considering coordinated multi-timescale volt-var optimization in distribution grids