Reinforcement learning for demand response: A review of algorithms and modeling techniques

Vázquez-Canteli, José R.; Nagy, Zoltán

doi:10.1016/j.apenergy.2018.11.002

Cited by 520 publications

(227 citation statements)

References 99 publications

Supporting

Mentioning

184

Contrasting

Order By: Relevance

“…Several studies in the literature have also considered DSM schemes that involve heating, ventilation, and air conditioning (HVAC) systems, as it is one of the dominating electricity consuming systems in buildings [33][34][35]. Although HVAC systems are associated with occupants' comfort [36], different DSM strategies are proposed which do not breach thermal comfort of the occupants. In addition, impacts on comfort also depend on thermal characteristics of the buildings [37].…”

Section: Dsm In Developed and Developing Countriesmentioning

confidence: 99%

“…Numerous studies have explained 'non-stationary' or 'dynamic' demand response using different algorithms [12,36,[77][78][79][80]. Most of these studies are simulation based.…”

Section: Types Of Demand Response and Its Potentialitymentioning

confidence: 99%

“…Most of these studies are simulation based. To date, only a few studies have tested the proposed algorithm in physical systems [36]. One of the most promising algorithms is reinforcement learning (RL), which is an agent-based artificial intelligent algorithm [81].…”

Section: Types Of Demand Response and Its Potentialitymentioning

confidence: 99%

“…One of the most promising algorithms is reinforcement learning (RL), which is an agent-based artificial intelligent algorithm [81]. Study suggests that RL is not only able to design effective dynamic demand response schemes but also considers occupants' comfort at houses [12,36]. In addition, Lu et al (2018) found that RL-based DSM scheme is able to maintain balance between supply and demand in the electricity market [12].…”

Section: Types Of Demand Response and Its Potentialitymentioning

confidence: 99%

See 3 more Smart Citations

Energy-saving behaviour as a demand-side management strategy in the developing world: the case of Bangladesh

Khan

2019

Int J Energy Environ Eng

View full text Add to dashboard Cite

Although demand-side management (DSM) needs to be more customer centred, either with or without smart technologies (e.g. smart grid), less attention has been paid to the developing world in relation to DSM strategy development. The main reasons have been lack of appropriate technology and capital costs. Importantly, there are alternative DSM strategies that require minimum or no cost to implement and provide immediate results, of which energy-saving behaviour of the occupants at residences is one. This study explores the potentiality of this energy-saving behaviour as a DSM strategy for the least developed economies, focusing particularly on Bangladesh. The literature suggests that energy-saving behaviour could reduce energy demand by a maximum of 21.9%. However, this potential DSM scheme seems underestimated in the national DSM programme of Bangladesh. The Energy Efficiency and Conservation Master Plan (EECMP) of Bangladesh (a DSM program) shows that efficiency improvement in the use of home appliances could reduce electricity demand in the residential sector by about 28.8%, but this does require a long time to be implemented, whereas the inclusion of energy-saving behaviour as a demand response strategy in residences along with the EECMP might achieve demand reduction of up to 50.7%. Although the findings from this study are specific to Bangladesh, these could be useful guidelines for the policymakers of other developing nations where national DSM strategy development is underway.

show abstract

Section: Dsm In Developed and Developing Countriesmentioning

confidence: 99%

“…Numerous studies have explained 'non-stationary' or 'dynamic' demand response using different algorithms [12,36,[77][78][79][80]. Most of these studies are simulation based.…”

Section: Types Of Demand Response and Its Potentialitymentioning

confidence: 99%

Section: Types Of Demand Response and Its Potentialitymentioning

confidence: 99%

Section: Types Of Demand Response and Its Potentialitymentioning

confidence: 99%

See 2 more Smart Citations

Energy-saving behaviour as a demand-side management strategy in the developing world: the case of Bangladesh

Khan

2019

Int J Energy Environ Eng

View full text Add to dashboard Cite

show abstract

“…RL methods can solve sequential decisionmaking problems in real time [10]. The last two decades have seen increasing efforts to apply conventional RL methods, such as Q-learning and fitted Q-iteration [10], in various decisionmaking and control problems in power systems; these range from demand response [11], energy management, and automatic generation control to transient stability and emergency control [9], [12], [13]. Due to scalability issues, applications of conventional RL methods are mainly focusing on problems with low-dimensional state and action spaces.…”

Section: Introductionmentioning

confidence: 99%

Adaptive Power System Emergency Control Using Deep Reinforcement Learning

Huang

Hao

et al. 2020

IEEE Trans. Smart Grid

274

157

View full text Add to dashboard Cite

Power system emergency control is generally regarded as the last safety net for grid security and resiliency. Existing emergency control schemes are usually designed off-line based on either the conceived "worst" case scenario or a few typical operation scenarios. These schemes are facing significant adaptiveness and robustness issues as increasing uncertainties and variations occur in modern electrical grids. To address these challenges, this paper developed novel adaptive emergency control schemes using deep reinforcement learning (DRL) by leveraging the high-dimensional feature extraction and non-linear generalization capabilities of DRL for complex power systems.Furthermore, an open-source platform named Reinforcement Learning for Grid Control (RLGC) has been designed for the first time to assist the development and benchmarking of DRL algorithms for power system control. Details of the platform and DRL-based emergency control schemes for generator dynamic braking and under-voltage load shedding are presented. Robustness of the developed DRL method to different simulation scenarios, model parameter uncertainty and noise in the observations is investigated. Extensive case studies performed in both the two-area, four-machine system and the IEEE 39-bus system have demonstrated excellent performance and robustness of the proposed schemes.

show abstract

Deep learning and reinforcement learning approach on microgrid

Kumar¹,

Kandasamy

Ramanathan

2020

Int Trans Electr Energ Syst

View full text Add to dashboard Cite

Microgrid is a new era in the power system and it has more scope of investigation on research. Due to an increase in demand and future expansion of the power system, analyzing the complexities of the network becomes a challenging task. Artificial intelligence plays a vital role in resolving such issues in a microgrid in various aspects. Owing to the rapid growth of periodical update in computational cost reduction, enhanced data analysis-based algorithm artificial intelligence enters into new epoch Artificial Intelligence AI 2.0. Based on such approach, machine learning has been evolved as AI 2.0 initially. Now, it develops branches like deep learning, reinforcement learning, and a combination of both deep reinforcement learning algorithms. These algorithms are precise to attain higher priority in decision-making under a complex network. This paper deals with numerous challenges of the above algorithm to state the significance of AI 2.0 and summarization of their application toward microgrid is useful to analyze the power system. K E Y W O R D S deep learning, deep reinforcement learning, microgrid, reinforcement learning 1 | INTRODUCTION Microgrid is a cluster of distributed generators (DG), stored energy system, local loads along with protecting and monitoring devices. Microgrid has its own specialty in operating under two modes of operation: autonomous and grid-connected mode, mode selection attained by the use of static transfer switch that bridges operation selection in microgrid. The presence of multiple DGs creates a distribution access problem in multipoint is transformed into a concentrated

show abstract

Reinforcement learning for demand response: A review of algorithms and modeling techniques

Cited by 520 publications

References 99 publications

Energy-saving behaviour as a demand-side management strategy in the developing world: the case of Bangladesh

Energy-saving behaviour as a demand-side management strategy in the developing world: the case of Bangladesh

Adaptive Power System Emergency Control Using Deep Reinforcement Learning

Deep learning and reinforcement learning approach on microgrid

Contact Info

Product

Resources

About