This paper proposes a distributed adaptive robust voltage/var control (DAR-VVC) method in active distribution networks to minimize power loss while keeping operating constraints under uncertainties. The DAR-VVC aims to coordinate on-load tap changers, capacitor banks and PV inverters in multiple operation stages through a distributed algorithm. To improve efficiency of the distributed algorithm, an affinity propagation clustering algorithm is employed to divide the distribution network by aggregating “the close nodes” together and setting “the far nodes” apart, leading to the network partition where the information exchange between adjacent sub-networks is reduced. Moreover, the virtual load which describes load characteristics of the sub-networks is applied to enhance the boundary conditions. To fully deal with the uncertainties, the proposed DAR-VVC is formulated in a robust optimization model which considers the worst case to guarantee solution robustness against uncertainty realization. Besides, this paper develops an alternating optimization procedure integrating a column-and-constraint generation algorithm and an alternating direction method of multipliers to solve the DAR-VVC problem. The proposed approach is tested on IEEE 33 and IEEE 123 bus distribution test system and numerical simulations verify high efficiency and full solution robustness of the DAR-VVC.

Distributed Adaptive Robust Voltage/VAR Control With Network Partition in Active Distribution Networks

Combined decentralized and local voltage control strategy of soft open points in active distribution networks

This paper proposes a novel model-free/data-driven centralized training and decentralized execution multi-agent deep reinforcement learning (MADRL) framework for distribution system voltage control with high penetration of PVs. The proposed MADRL can coordinate both the real and reactive power control of PVs with existing static var compensators and battery storage systems. Unlike the existing DRL-based voltage control methods, our proposed method does not rely on a system model during both the training and execution stages. This is achieved by developing a new interaction scheme between the surrogate modeling of the original system and the multi-agent soft actor critic (MASAC) MADRL algorithm. In particular, the sparse pseudo-Gaussian process with a few-shots of measurements is utilized to construct the surrogate model of the original environment, i.e., power flow model. This is a data-driven process and no model parameters are needed. Furthermore, the MASAC enabled MADRL allows to achieve better scalability by dividing the original system into different voltage control regions with the aid of real and reactive power sensitivities to voltage, where each region is treated as an agent. This also serves as the foundation for the centralized training and decentralized execution, thus significantly reducing the communication requirements as only local measurements are required for control. Comparative results with other alternatives on the IEEE 123-nodes and 342-nodes systems demonstrate the superiority of the proposed method.

Data-Driven Multi-Agent Deep Reinforcement Learning for Distribution System Decentralized Voltage Control With High Penetration of PVs

With the increasing integration of distribution generations (DGs), active distribution networks (ADNs) need to address the new challenges in voltage control. Meanwhile, the traditional centralized method in reactive power optimization usually leads to a heavy computational burden for large-scale distribution networks. A distributed voltage control model with a novel network partitioning method is proposed in this study, where the capacity curve of DGs and the regulation of electrical storage systems (ESS) are specially addressed. According to the number of decision variables and the number of constraints in the optimization model, the computational complexity can be utilized in the network partitioning method with the power balance and the electrical distance to automatically partition a distribution network in an optimal manner. Taking the coordination between two boundary sub-areas into consideration, a distributed voltage control strategy is developed and solved by a distributed solution method based on Lagrangian dual relaxation. Finally, numerical studies based on two IEEE test systems and a practical 152-bus system are performed to verify the effectiveness of the proposed method.

Distributed Voltage Control in Active Distribution Network Considering Renewable Energy: A Novel Network Partitioning Method

This paper proposes attention enabled multi-agent deep reinforcement learning (MADRL) framework for active distribution network decentralized Volt-VAR control. Using the unsupervised clustering, the whole distribution system can be decomposed into several sub-networks according to the voltage and reactive power sensitivity relationships. Then, the distributed control problem of each sub-network is modeled as Markov games and solved by the improved MADRL algorithm, where each sub-network is modeled as an adaptive agent. An attention mechanism is developed to help each agent focus on specific information that is mostly related to the reward. All agents are centrally trained offline to learn the optimal coordinated Volt-VAR control strategy and executed in a decentralized manner to make online decisions with only local information. Compared with other distributed control approaches, the proposed method can effectively deal with uncertainties, achieve fast decision makings, and significantly reduce the communication requirements. Comparison results with model-based and other data-driven methods on IEEE 33-bus and 123-bus systems demonstrate the benefits of the proposed approach.

Attention Enabled Multi-Agent DRL for Decentralized Volt-VAR Control of Active Distribution System Using PV Inverters and SVCs

In order to solve the voltage violation problem caused by high penetration of photovoltaic (PV) units in distribution power networks, this paper proposes a double-layer voltage control strategy based on the distribution network partition. By optimizing the active and reactive power outputs of PV units, the minimization objective of PV active power curtailment and network active power loss can be achieved. In the basis of community detection algorithm, a novel cluster performance index based on the electrical distance and regional voltage regulation capability is presented to partition a distribution network into several clusters. The proposed double-layer voltage control strategy combines the cluster autonomous optimization and distributed intercluster coordination optimization in different time scales. The cluster autonomous optimization can eliminate intracluster voltage violation rapidly by alternately updating the intracluster optimal solution and the virtual slack bus voltage. The distributed intercluster coordination optimization with a long-time scale is based on the alternating direction method of multipliers and realizes global optimal control through finite boundary data exchange among adjacent clusters. The effectiveness and feasibility of the proposed method have been demonstrated via simulation tests on a real 10.5 kV feeder in China and the IEEE 123-bus system.

Network Partition and Voltage Coordination Control for Distribution Networks With High Penetration of Distributed PV Units

With the increasing penetration of distributed photovoltaics (PVs), the operation and control of distribution networks (DNs), especially voltage control, have become more complicated. To deal with the voltage violation problem caused by large-scale PV access, this paper presents a fully distributed optimization method that combines the alternative direction multiplier method (ADMM) with the branch and bound method (BBM) for regional DNs. The total cost of active power losses and PV generation losses is minimized by making full use of the voltage regulation resources, e.g., reactive power compensators, step voltage regulators (SVR), and PV inverters, and the ADMM is employed to realize the intra-regional optimization and inter-regional coordination. To overcome the non-convex problem that is introduced by the SVR, the constraints of real-value tap positions are reformulated as linear inequality constraints of boundary voltages and added to the original problem, then the integer optimal solutions of SVR tap positions are obtained by BBM. The effectiveness of the proposed method is verified via numerical simulations on a practical 32-bus DN in China and a modified IEEE123-bus system.

/pdf/a-fully-distributed-voltage-optimization-method-for-1j6wle01fb.pdf

A Fully Distributed Voltage Optimization Method for Distribution Networks Considering Integer Constraints of Step Voltage Regulators

The increasing penetration of distributed photovoltaics (PVs) has brought great challenges to the security and stability of distribution networks (DNs), especially voltage violation. Current voltage optimization methods rarely consider the mutual voltage support between high voltage (HV) and medium voltage (MV) DNs, which may cause PV generation losses in extreme scenarios. Limited by the convergence conditions of common distributed optimization algorithms, mixed integer optimization problem can hardly be calculated in a distributed manner. This paper tries to exert the mutual voltage support capability of HV and MV DNs and realize optimal operation of continuous and discrete voltage regulation devices. Thus, a hierarchical distributed voltage optimization method for HV and MV DNs is proposed in this paper. It is assumed that HV and MV DNs are operated by different control systems and real-time communication between them is available. First, the voltage optimization models of HV and MV DNs are developed to minimize the total cost of voltage adjustment and network losses, which optimize the dispatch of on-load tap changers, reactive power compensators, feeder switches, and distributed PVs. Then the generalized Benders decomposition is applied to solve the mixed integer optimization model in a master-slave distributed manner and the accuracy of LinDistFlow equation is improved through the compensation of peak voltage and boundary power flow in MV DNs. Finally, the effectiveness of proposed method is demonstrated via simulation tests in Jinzhai DNs of China. The results of proposed optimization method are very close to those of global centralized optimization method.

Hierarchical Distributed Voltage Optimization Method for HV and MV Distribution Networks

Multi-stage resilience scheduling of electricity-gas integrated energy system with multi-level decentralized reserve

High penetration of distributed photovoltaics (PVs) has caused severe voltage violation problem in distribution networks. In view of the actual situations, this paper proposes a coordinated optimization method for distributed PVs clusters considering inadequate measurement data. Based on pivotal measurement data, each cluster firstly uses network simplification method to obtain a simplified network with complete measurement information. And then each cluster implements coordinated optimization with adjacent clusters to minimize its network losses, PVs generation losses and voltage violation amount of downstream cluster. The coordinated optimization includes two stages, i.e. evaluating voltage regulation capability of intra-cluster reactive power resources and implementing cluster autonomous optimization control. These two stages can use the reactive power resources of upstream clusters to solve the voltage violation of downstream clusters, which avoids unnecessary generation losses of distributed PVs. The effectiveness and accuracy of the proposed method are demonstrated via simulation tests on a real 10kV feeder in China and modified IEEE 123-bus system. Compared with distributed optimization method, the proposed coordinated optimization requires much less calculation time and communication data between clusters.

Yuanyuan Chai

Papers

Network Partition and Voltage Coordination Control for Distribution Networks With High Penetration of Distributed PV Units

A Fully Distributed Voltage Optimization Method for Distribution Networks Considering Integer Constraints of Step Voltage Regulators

Hierarchical Distributed Voltage Optimization Method for HV and MV Distribution Networks

Multi-stage resilience scheduling of electricity-gas integrated energy system with multi-level decentralized reserve

Network Simplification-Based Cluster Coordinated Optimization Method for Distributed PVs With Inadequate Measurement