9,129 Hits in 6.8 sec

Deep Reinforcement Learning for Real-Time Optimization of Pumps in Water Distribution Systems

Gergely Hajgató, György Paál, Bálint Gyires-Tóth
2020 Journal of water resources planning and management  
Real-time control of pumps can be an infeasible task in water distribution systems (WDSs) because the calculation to find the optimal pump speeds is resource-intensive.  ...  Deep reinforcement learning (DRL) is presented here as a controller of pumps in two WDSs.  ...  The research presented in this paper has been supported by the BME-Artificial Intelligence  ... 
doi:10.1061/(asce)wr.1943-5452.0001287 fatcat:vy6wgrwhiragdgtshjragc3iii

Gigawatt-hour scale savings on a budget of zero: Deep reinforcement learning based optimal control of hot water systems

Hussain Kazmi, Fahad Mehmood, Stefan Lodeweyckx, Johan Driesen
2018 Energy  
In this paper, we present a novel reinforcement learning based methodology which optimizes hot water production.  ...  Furthermore, optimization usually presupposes existence of a detailed dynamics model for the hot water system. These assumptions lead to suboptimal energy efficiency in the real world.  ...  The authors also thank the referees for their constructive input.  ... 
doi:10.1016/ fatcat:kvua5ub5gvhata7oulhksjiea4

Deep Reinforcement Learning-Based Smart Joint Control Scheme for On/Off Pumping Systems in Wastewater Treatment Plants

Giup Seo, Seungwook Yoon, Myungsun Kim, Changho Moon, Euiseok Hwang
2021 IEEE Access  
for operation in a real-world system [17] .  ...  Shiue et al. proposed a Q-learning based real-time scheduling approach for a smart factory [16] .  ... 
doi:10.1109/access.2021.3094466 fatcat:i53gkyesofcrhfyfvhnhbd3ks4

Machine Learning and Urban Drainage Systems: State-of-the-Art Review

Soon-Ho Kwon, Joong-Hoon Kim
2021 Water  
This review paper provides a state-of-the-art review of ML-based UDS modeling/application based on three categories: (1) operation (real-time operation control), (2) management (flood-inundation prediction  ...  distributions.  ...  Mullapudi et al. (2020) [32] formulated and analyzed a real-time operation control model (i.e., pumps) using reinforcement learning.  ... 
doi:10.3390/w13243545 fatcat:l3vq6vefkfaybh4nl22zzk5kte

Flow Rate Control in Smart District Heating Systems Using Deep Reinforcement Learning [article]

Tinghao Zhang, Jing Luo, Ping Chen, Jie Liu
2019 arXiv   pre-print
A real-world case study shows the effectiveness and practicability of the proposed system controlled by humans, and the simulated experiments for deep reinforcement learning show about 1985.01 gigajoules  ...  In this paper, we describe the theoretical methods to solve this problem by deep reinforcement learning and propose a cloud-based heating control system for implementation.  ...  In this study, we resorted to deep reinforcement learning (DRL) for the DHS's flow rates control. Reinforcement learning (RL) has been proven to be efficient for optimal control problems [29] .  ... 
arXiv:1912.05313v1 fatcat:hbshsz6uzrffpdj3varr5vxhua

Optimal demand response strategy of commercial building‐based virtual power plant using reinforcement learning

Tao Chen, Qiushi Cui, Ciwei Gao, Qinran Hu, Kexing Lai, Jianlin Yang, Ran Lyu, Hao Zhang, Jinyuan Zhang
2021 IET Generation, Transmission & Distribution  
.: Optimal demand response strategy of commercial building-based virtual power plant using reinforcement learning. IET  ...  In this paper, the optimal demand response strategy of a commercial building-based virtual power plant with real-world implementation in heavily urbanised area is studied.  ...  The authors also thank Economic and Technical Research Institute (ETRI) in State Grid Corporation of China-Shanghai Company and Shanghai Huangpu District Development and Reform Commission's support to  ... 
doi:10.1049/gtd2.12179 fatcat:dc4jiwcy3nakxjkofb2g7vvp4a

Comparative Evaluation of Different Multi-Agent Reinforcement Learning Mechanisms in Condenser Water System Control

Shunian Qiu, Zhenhai Li, Zhengwei Li, Qian Wu
2022 Buildings  
For comparison, quantitative simulations are conducted based on a virtual environment established using measured data of a real condenser water system.  ...  Model-free reinforcement learning (RL) techniques are currently drawing attention in the control of heating, ventilation, and air-conditioning (HVAC) systems due to their minor pre-conditions and fast  ...  Variable Description Unit 𝑃 Real-time overall system electrical power kW 𝐶𝐿 System cooling load kW 𝑇 Ambient wet-bulb temperature °C 𝑓 Common frequency of running condenser water pump(s) Hz 𝑛 Current  ... 
doi:10.3390/buildings12081092 fatcat:tvigll4isrdclllaejjtz5u6ce

Real-Time Demand Response Management for Controlling Load Using Deep Reinforcement Learning

Yongjiang Zhao, Jae Hung Yoo, Chang Gyoon Lim
2022 Computers Materials & Continua  
The results show that through the coordination of the SAC to control load in CityLearn, realizes the goal of reducing both the peak load demand and the operation costs on the premise of regulating voltage  ...  Enhancing and improving the power distribution infrastructure requires high maintenance costs. 2) The user's electricity schedule is unreasonable due to personal behavior, which will cause a waste of electricity  ...  Unlike traditional model-based methods that require an explicit physical or mathematical model of the system, deep reinforcement learning (DRL), a combination of reinforcement learning and deep learning  ... 
doi:10.32604/cmc.2022.027443 fatcat:5lj43vabwjbefbd6sxbjdihxo4

Transfer learning applied to DRL-Based heat pump control to leverage microgrid energy efficiency

Paulo Lissa, Michael Schukat, Marcus Keane, Enda Barrett
2021 Smart Energy  
This paper investigates the application of transfer learning applied to a deep reinforcement learning-based heat pump control to leverage energy efficiency in a microgrid.  ...  For a particular state, the agent receives a J o u r n a l P r e -p r o o f Highlights  Transfer Learning can reduce time to train control policies more than a factor of 5.  Deep Reinforcement Learning-based  ...  learning to a deep reinforcement learning (DRL)based heat pump control to leverage energy efficiency in a microgrid, focusing on DHW.  ... 
doi:10.1016/j.segy.2021.100044 fatcat:uhkcf4bstbhztcd4v54jwjvb2a

Semi-analytical Industrial Cooling System Model for Reinforcement Learning [article]

Yuri Chervonyi, Praneet Dutta, Piotr Trochim, Octavian Voicu, Cosmin Paduraru, Crystal Qian, Emre Karagozler, Jared Quincy Davis, Richard Chippendale, Gautam Bajaj, Sims Witherspoon, Jerry Luo
2022 arXiv   pre-print
This model is designed for reinforcement learning (RL) applications and balances simplicity with simulation fidelity and interpretability.  ...  The model's fidelity is evaluated against real world data from a large scale cooling system. This is followed by a case study illustrating how the model can be used for RL research.  ...  Examples of applying RL in real cooling/heating systems include reducing energy consumption for cooling Google data centers by 40% [2] and controlling the HVAC systems, hot water tank and heat pumps in  ... 
arXiv:2207.13131v1 fatcat:kesfk5km7vbztgkajkrwsn2hfi

Using Reinforcement Learning for Demand Response of Domestic Hot Water Buffers: a Real-Life Demonstration [article]

Oscar De Somer, Ana Soares, Tristan Kuijpers, Koen Vossen, Koen Vanthournout, Fred Spiessens
2017 arXiv   pre-print
This paper demonstrates a data-driven control approach for demand response in real-life residential buildings.  ...  The objective is to optimally schedule the heating cycles of the Domestic Hot Water (DHW) buffer to maximize the self-consumption of the local photovoltaic (PV) production.  ...  ACKNOWLEDGMENT The research leading to these results has received funding from the European Commission in the H2020 Programme -EU. under grant agreement nr.680603 -REnnovates.  ... 
arXiv:1703.05486v1 fatcat:hqayahaum5gifpqj2vokg7iphi

An Advanced Learning-Based Multiple Model Control Supervisor for Pumping Stations in a Smart Water Distribution System

Alexandru Predescu, Ciprian-Octavian Truică, Elena-Simona Apostol, Mariana Mocanu, Ciprian Lupu
2020 Mathematics  
to changing operating conditions.The high-level processing and components for smart water distribution systems are supportedby the smart meters, providing real-time data, push-based and decoupled software  ...  Water distribution is fundamental to modern society, and there are many associatedchallenges in the context of large metropolitan areas.  ...  modeling in water distribution systems [22] .  ... 
doi:10.3390/math8060887 fatcat:5xfrwifmr5hx5mnj432cohp6ma

Deep Reinforcement Learning with Uncertain Data for Real-Time Stormwater System Control and Flood Mitigation

Sami M. Saliba, Benjamin D. Bowes, Stephen Adams, Peter A. Beling, Jonathan L. Goodall
2020 Water  
One method of automating RTC is reinforcement learning (RL), a general technique for sequential optimization and control in uncertain environments.  ...  Retrofitting standard passive systems with controllable valves/pumps is promising, but requires real-time control (RTC).  ...  Reinforcement learning (RL) [16] , a type of machine learning, is an emerging approach to stormwater system RTC that allows the creation of policies for flow control valves, pumps, and ponds within a  ... 
doi:10.3390/w12113222 fatcat:qoilawmkzbaevm7mep7ijo6iwe

Direct load control of thermostatically controlled loads based on sparse observations using deep reinforcement learning

2019 CSEE Journal of Power and Energy Systems  
CONCLUSION In this paper, we demonstrated the effectiveness of combining different deep learning techniques with reinforcement learning for two demand response applications that are hindered by sparse  ...  LITERATURE REVIEW This section provides a short literature overview of Reinforcement Learning (RL) related to demand response and discusses some relevant applications of deep learning in RL. A.  ... 
doi:10.17775/cseejpes.2019.00590 fatcat:xr5afuayjbav5o2uq36vxaplle

Preheating Quantification for Smart Hybrid Heat Pumps Considering Uncertainty

Mingyang Sun, Goran Strbac, Predrag Djapic, Danny Pudjianto
2019 IEEE Transactions on Industrial Informatics  
In particular, the fully-optimized control technology can provide flexible heat that redistributes the heat demand across time for improving the utilization of low-carbon generation and enhancing the overall  ...  Varieties of fine-grained data from a real-world trial are exploited to estimate the baseline heat demand using Bayesian deep learning while jointly considering epistemic and aleatoric uncertainties.  ...  In other words, "thinking" is one of the core, fundamental abilities for deep learning to build a real artificial intelligence system.  ... 
doi:10.1109/tii.2019.2891089 fatcat:b7khhegvyzgwvikgw24ko7icpy
« Previous Showing results 1 — 15 out of 9,129 results