Filters








645,997 Hits in 4.7 sec

Energy-Based Models for Continual Learning [article]

Shuang Li, Yilun Du, Gido M. van de Ven, Igor Mordatch
2021 arXiv   pre-print
We motivate Energy-Based Models (EBMs) as a promising model class for continual learning problems.  ...  Our proposed version of EBMs for continual learning is simple, efficient and outperforms baseline methods by a large margin on several benchmarks.  ...  First, we introduce energy-based models for classification continual learning problems.  ... 
arXiv:2011.12216v2 fatcat:vnn4ahwa6nbd3l2cu24vlo43me

Base Station Power Optimization for Green Networks Using Reinforcement Learning

Semih AKTAŞ, Hande ALEMDAR
2021 Sakarya University Journal of Computer and Information Sciences  
We develop a reinforcement-based learning model by using deep deterministic policy gradient algorithm.  ...  Reducing the energy consumption of base stations is essential for going green and also it helps service providers to reduce operational expenses.  ...  They transform continuous decision variables into discrete ones to reduce complexity and to fit their models which are based on regret learning based RL and fictitious play based RL.  ... 
doi:10.35377/saucis.04.02.932709 fatcat:oa2pvu6wc5egvat36tarn7euaa

Regularizing Model-Based Planning with Energy-Based Models [article]

Rinu Boney, Juho Kannala, Alexander Ilin
2019 arXiv   pre-print
Learned dynamics models can be directly used for planning actions but this has been challenging because of inaccuracies in the learned models.  ...  Model-based reinforcement learning could enable sample-efficient learning by quickly acquiring rich knowledge about the world and using it to improve behaviour without additional data.  ...  Acknowledgments We would like to thank Saeed Saremi for valuable discussions about his work on deep energy estimator networks and neural empirical Bayes.  ... 
arXiv:1910.05527v1 fatcat:jdxulfgvlndvbp5isenacq2pe4

Simulating user learning in authoritative technology adoption: An agent based model for council-led smart meter deployment planning in the UK [article]

Tao Zhang, Peer-Olaf Siebers, Uwe Aickelin
2016 arXiv   pre-print
With the agent-based model we carry out experiments to validate the model and test different energy interventions that local authorities can use to facilitate energy consumers' learning and maintain their  ...  Energy consumers gain experience of using smart meters based on the learning curve in behavioural learning.  ...  model using Agent-Based Simulation (ABS).  ... 
arXiv:1607.05912v1 fatcat:rcewenvkvfhmfh6zr7vkr3y7vy

Deep Reinforcement Learning for Optimal Control of Space Heating [article]

Adam Nagy, Hussain Kazmi, Farah Cheaib, Johan Driesen
2018 arXiv   pre-print
The proposed algorithm outperforms rule based control by between 5-10% in a simulation environment for a number of price signals.  ...  Classical methods to control heating systems are often marred by suboptimal performance, inability to adapt to dynamic conditions and unreasonable assumptions e.g. existence of building models.  ...  Quality of model learnt in model-based learning For model-based RL, the quality of the learnt model is of paramount importance.  ... 
arXiv:1805.03777v1 fatcat:gavxolu2qbaezdpvatrx6735ja

Reinforcement Learning and Its Applications in Modern Power and Energy Systems: A Review

Di Cao, Weihao Hu, Junbo Zhao, Guozhou Zhang, Bin Zhang, Zhou Liu, Zhe Chen, Frede Blaabjerg
2020 Journal of Modern Power Systems and Clean Energy  
With the growing integration of distributed energy resources (DERs), flexible loads, and other emerging technologies, there are increasing complexities and uncertainties for modern power and energy systems  ...  Among them, reinforcement learning (RL) is one of the most widely promoted methods for control and optimization problems.  ...  advantage actor-critic (A3C) Value-based (model-free) Q-learning Model-based regulator (ILQR) learn control (PILCO) Environment Agent State Action Reward Fig. 6.  ... 
doi:10.35833/mpce.2020.000552 fatcat:42bllvvymfhfxbh42t6a46q4tq

Energy Management Model for HVAC Control Supported by Reinforcement Learning

Pedro Macieira, Luis Gomes, Zita Vale
2021 Energies  
learning and predictive model able to predict if users will be working in a given location, and (iii) the proposed decision model to manage the HVAC units according to the prediction of users, current  ...  The method applied in this paper divides the addressed problem into three steps: (i) the continuous acquisition of data provided by an open-source building energy management systems, (ii) the proposed  ...  Energy Management Model for HVAC Control Supported by Reinforcement Learning. Energies 2021, 14, 8210. https://doi.org/ 1.  ... 
doi:10.3390/en14248210 fatcat:a4tbt4l4wbaljknoc3xlml3qhy

Energy-Based Continuous Inverse Optimal Control [article]

Yifei Xu, Jianwen Xie, Tianyang Zhao, Chris Baker, Yibiao Zhao, Ying Nian Wu
2022 arXiv   pre-print
generator is used to fast initialize the synthesis step of the energy-based model.  ...  Moreover, to make the sampling or optimization more efficient, we propose to train the energy-based model simultaneously with a top-down trajectory generator via cooperative learning, where the trajectory  ...  We also train the energy-based model with a trajectory generator as a fast initializer for the Langevin sampling of the energy-based model in a cooperative learning scheme.  ... 
arXiv:1904.05453v6 fatcat:ai62xvl7fng6xip2ptjj63cruu

Simulating user learning in authoritative technology adoption: An agent based model for council-led smart meter deployment planning in the UK

Tao Zhang, Peer-Olaf Siebers, Uwe Aickelin
2016 Technological forecasting & social change  
Zhang, Tao and Siebers, Peer-Olaf and Aickelin, Uwe (2016) Simulating user learning in authoritative technology adoption: an agent based model for councilled smart meter deployment planning in the UK.  ...  With the agent-based model we carry out experiments to validate the model and test different energy interventions that local authorities can use to facilitate energy consumers' learning and maintain their  ...  model using Agent-Based Simulation (ABS).  ... 
doi:10.1016/j.techfore.2016.02.009 fatcat:iejzmav2endungjy2xarhwzcbi

Deep Reinforcement Learning for Backscatter-Aided Data Offloading in Mobile Edge Computing [article]

Shimin Gong, Yutong Xie, Jing Xu, Dusit Niyato, Ying-Chang Liang
2020 arXiv   pre-print
By continuously interacting with the environment, deep reinforcement learning (DRL) provides a mechanism for different network entities to build knowledge and make autonomous decisions to improve network  ...  We then discuss the applications of DRL for mobile edge computing (MEC), which can be used for the low-power IoT devices, e.g., wireless sensors in healthcare monitoring, to offload computation workload  ...  There are mainly value-and policy-based approaches for solving reinforcement learning problems [4] .  ... 
arXiv:2001.10183v1 fatcat:uu77msyizbfyjezlo3t2sot74a

Real Time Demand Response Modelling for Residential Consumers in Smart Grid Considering Renewable Energy with Deep Learning Approach

S. Sofana Reka, V. Prakash, Hassan Haes Alhelou, Pierluigi Siano, M.E.H. Golshan
2021 IEEE Access  
from the utility and this helps in maximizing the reward obtain and the model continues for every strategy created by an agent developed in this learning model.  ...  This learning model inhibits for the consumers to learn the agents' policy created generally from best two strategies created from the consumers in a continuous period.  ...  modeling.  ... 
doi:10.1109/access.2021.3071993 fatcat:5opfbqejqzdqneiffklcetwsam

A Relearning Approach to Reinforcement Learning for Control of Smart Buildings [article]

Avisek Naug and Marcos Quiñones-Grueiro and Gautam Biswas
2020 arXiv   pre-print
This paper demonstrates that continual relearning of control policies using incremental deep reinforcement learning (RL) can improve policy learning for non-stationary processes.  ...  We demonstrate this approach for a data-driven 'smart building environment' that we use as a test-bed for developing HVAC controllers for reducing energy consumption of large buildings on our university  ...  In the next section, we present a continuous learning approach for optimal control of non stationary processes based on this idea.  ... 
arXiv:2008.01879v1 fatcat:epddy7mzqvgohjxspykwp5px7q

Modeling Instruction to Promote Student's Understanding of System and Model of System of Mechanical Energy

Zainul Mustofa, Anik Asmichatin
2019 Abjadia  
<p>Understanding of system and model of the system are necessary for conducting an analysis of scientific phenomena. One concept in science that requires system is energy.  ...  Moreover, any misconception about energy was able to prevent and overcome by this learning. By learning, student's skill to solve new problem had changed approaching expert problem solver.  ...  This activity was called model development. Picture 1. Sequence Activity of Learning of Work-Energy Theorem Then students continue to model deployment.  ... 
doi:10.18860/abj.v3i1.5939 fatcat:6poqivd5enaclpfbc2rvosntu4

Deep reinforcement learning-based vehicle energy efficiency autonomous learning system

Xuewei Qi, Yadan Luo, Guoyuan Wu, Kanok Boriboonsomsin, Matthew J. Barth
2017 2017 IEEE Intelligent Vehicles Symposium (IV)  
In this study, a deep reinforcement learning based PHEV energy management system is designed to autonomously learn the optimal fuel use from its own historical driving record.  ...  It is a fully datadriven and learning-enabled model that does not rely on any prediction or predefined rules.  ...  The DQN based model consumes more battery energy at these time steps with significant SOC drops to avoid unnecessary engine energy consumption.  ... 
doi:10.1109/ivs.2017.7995880 dblp:conf/ivs/QiLWBB17 fatcat:7xgjoltwgrd3nii6qrqg6bfrva

A Relearning Approach to Reinforcement Learning for control of Smart Buildings

Avisek Naug, Marcos Q'uiñones -Grueiro, Gautam Biswas
2020 Proceedings of the Annual Conference of the Prognostics and Health Management Society, PHM  
This paper demonstrates that continual relearning of control policies using incremental deep reinforcement learning (RL) can improve policy learning for non-stationary processes.  ...  This approach has been demonstrated in a data-driven "smart building environment" that we use as a test-bed for developing HVAC controllers for reducing energy consumption of large buildings on our university  ...  In the next section, we present a continuous learning approach for optimal control of non stationary processes based on this idea.  ... 
doi:10.36001/phmconf.2020.v12i1.1296 fatcat:jownxtnslra2bjmv3k6gj7kb4q
« Previous Showing results 1 — 15 out of 645,997 results