Hierarchical learning optimisation method for the coordination dispatch of the inter‐regional power grid considering the quality of service index

Lv, Kai; Tang, Hao; Bak‐Jensen, Birgitte; Pillai, Jayakrishnan Radhakrishna; Tan, Qingyuan; Zhang, Qianli

doi:10.1049/iet-gtd.2019.1869

Cited by 9 publications

(4 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2) The adopted data-driven AI technique, which refers to the DRL method with the advantage actor-critic (A2C) algorithm [34], effectively avoids the system modeling errors and the parameter estimation deviations. In contrast to the conventional complex modeling process with respect to RESs and loads, the proposed method can skip the dynamical modeling process and therefore is regarded as an advantage over the existing works [15], [16], [21], [25], [34]. When solving the considered highdimensional stochastic energy management problem, we demonstrate that the adopted DRL algorithm is more advantageous than the conventional approaches such as proportional integral (PI) based and OPF based control schemes.…”

Section: Introductionmentioning

confidence: 93%

“…Via the coordinative operation of intelligent agents at the MG level and ER level, energy management solutions regarding the future power and energy system generally come along with a multi-layer (or, hierarchical) architecture; see, e.g., [15], [16], [20]- [25]. In [15] a distributed control scheme for multiple MGs is designed to manage the balance between control performance and computation feasibility.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Data-Driven Dynamical Control for Bottom-up Energy Internet System

Qin

Dong

et al. 2022

IEEE Trans. Sustain. Energy

103

View full text Add to dashboard Cite

With the increasing concern on climate change and global warming, the reduction of carbon emission becomes an important topic in many aspects of human society. The development of energy Internet (EI) makes it possible to achieve better utilization of distributed renewable energy sources with the power sharing functionality introduced by energy routers (ERs). In this paper, a bottom-up EI architecture is designed, and a novel data-driven dynamical control strategy is proposed. Intelligent controllers augmented by deep reinforcement learning (DRL) techniques are adopted for the operation of each microgrid independently in the bottom layer. Moreover, the concept of curriculum learning (CL) is integrated into DRL to improve the sample efficiency and accelerate the training process. Based on the power exchange plan determined in the bottom layer, considering the stochastic nature of electricity price in the future power market, the optimal power dispatching scheme in the upper layer is decided via model predictive control. The simulation has shown that, under the bottom-up architecture, compared with the conventional methods such as proportional integral and optimal power flow, the proposed method reduces overall generation cost by 7.1% and 37%, respectively. Meanwhile, the introduced CL-based training strategy can significantly speed up the convergence during the training of DRL. Last but not least, our method increases the profit of energy trading between ERs and the main grid.

show abstract

Section: Introductionmentioning

confidence: 93%

Section: Introductionmentioning

confidence: 99%

Data-Driven Dynamical Control for Bottom-up Energy Internet System

Qin

Dong

et al. 2022

IEEE Trans. Sustain. Energy

103

View full text Add to dashboard Cite

show abstract

“…It adopts a quality-of-service approach to analyze the service attributes in power dispatch. In addition, the study also constructed a model free hierarchical optimization method based on learning technology, and used reinforcement learning algorithms to effectively solve the economic dispatch problem of cross regional power grids [10]. Scholars such as Xu D have designed a maximum minimum two-layer optimization model and a two-stage robust optimization model to solve the problem of unpredictability during power grid backup, and used column constraint generation algorithm to solve it.…”

Section: Related Workmentioning

confidence: 99%

Economic Dispatch of Multi Regional Power Systems Based on CMOPSO Algorithm

Guo

2024

SCPE

View full text Add to dashboard Cite

With the progress of society and the aggravation of environmental pollution, the economic dispatch of the power system is developing towards multiple environmental and economic goals. To improve energy utilization efficiency, this study innovatively proposes a multi-objective particle swarm optimization algorithm based on competitive learning, and uses this algorithm to solve multi regional environmental and economic scheduling problems. In addition, the study solves static and dynamic economic (S-DE) scheduling problems in multiple regions through improved competitive group optimization algorithms. The research results show that under different testing systems, the average distribution uniformity indicators of the research algorithm built on competitive learning are 0.8058 and 0.8457, and the average anti generation distance is 67.6316 and 1664.0978. The improved competitive group optimization algorithm solves the maximum, minimum, and average fuel costs for static economic scheduling in multiple regions, which are 656.2243 $/h, 655.8592 $/h, and 655.9866 $/h, respectively. Thus, the designed algorithm can effectively solve economic scheduling problems, which is of great significance for resource integration, saving power generation costs, and reducing pollution emissions.

show abstract

“…Chen et al developed a DDPG algorithm based on hybrid energy scheduling, which can learn the optimal policies from historical experiences, avoid inadequate exploration by introducing decaying noise (Chen et al, 2022). Reference (GUAN et al, 2020;Lv et al, 2020) has undertaken initial explorations into the utilization of deep reinforcement learning for real-time grid scheduling optimization. While these preliminary forays have delved into the optimization of grid scheduling, they have not yet been extended to address intricacies such as intra-day rolling scheduling, multi-objective grid scheduling, and the dynamic considerations arising from maintenance or minor faults in the system's topology.…”

Section: Introductionmentioning

confidence: 99%

Risk-averse stochastic dynamic power dispatch based on deep reinforcement learning with risk-oriented Graph-Gan sampling

Huang,

Dai,

Hou

et al. 2023

Front. Energy Res.

View full text Add to dashboard Cite

The increasing penetration of renewable energy sources (RES) brings volatile stochasticity, which significantly challenge the optimal dispatch of power systems. This paper aims at developing a cost-effective and robust policy for stochastic dynamic optimization of power systems, which improves the economy as well as avoiding the risk of high costs in some critical scenarios with small probability. However, it is hard for existing risk-neutral methods to incorporate risk measure since most samples are normal. For this regard, a novel risk-averse policy learning approach based on deep reinforcement learning with risk-oriented sampling is proposed. Firstly, a generative adversarial network (GAN) with graph convolutional neural network (GCN) is proposed to learn from historical data and achieve risk-oriented sampling. Specifically, system state is modelled as graph data and GCN is employed to capture the underlying correlation of the uncertainty corresponding to the system topology. Risk knowledge is the embedded to encourage more critical scenarios are sampled while aligning with historical data distributions. Secondly, a modified deep reinforcement learning (DRL) with risk-measure under soft actor critic framework is proposed to learn the optimal dispatch policy from sampling data. Compared with the traditional deep reinforcement learning which is risk-neutral, the proposed method is more robust and adaptable to uncertainties. Comparative simulations verify the effectiveness of the proposed method.

show abstract

Hierarchical learning optimisation method for the coordination dispatch of the inter‐regional power grid considering the quality of service index

Cited by 9 publications

References 26 publications

Data-Driven Dynamical Control for Bottom-up Energy Internet System

Data-Driven Dynamical Control for Bottom-up Energy Internet System

Economic Dispatch of Multi Regional Power Systems Based on CMOPSO Algorithm

Risk-averse stochastic dynamic power dispatch based on deep reinforcement learning with risk-oriented Graph-Gan sampling

Contact Info

Product

Resources

About