Weather data modelling for energy system models using machine learning

2020 ◽  
Author(s):  
Alexander Kies ◽  
Nishtha Srivastava ◽  
Kai Zhou ◽  
Jan Steinheimer ◽  
Horst Stoecker
Keyword(s):  
Machine Learning ◽  
Large Scale ◽  
Numerical Models ◽  
Energy Systems ◽  
Energy System ◽  
Weather Data ◽  
Generative Adversarial Networks ◽  
System Models ◽  
Wind Speeds ◽  
Adversarial Networks

<p>Weather data is essential to model and optimise energy systems, which are based on high shares of renewable generation sources. However, differences between data sources can be significant and often little emphasis is put on energy-related variables such as hub-height wind speeds.</p><p>In this work, we use generative adversarial networks (GAN), a class of machine learning systems, to model weather data for large-scale energy system models and optimise energy systems of different scales and sizes.</p><p>We show that generating weather data using GAN saves effort as required for processing large amounts of weather data and that it can reliably reproduce results from using weather data produced by numerical models.</p>

ORGANIC (1).pdf

2017 ◽  
Author(s):  
Benjamin Sanchez-Lengeling ◽  
Carlos Outeiral ◽  
Gabriel L. Guimaraes ◽  
Alan Aspuru-Guzik
Keyword(s):  
Machine Learning ◽  
Learning Community ◽  
Chemical Species ◽  
Material Design ◽  
Organic Photovoltaic ◽  
Generative Adversarial Networks ◽  
Generative Adversarial Network ◽  
Adversarial Network ◽  
Adversarial Networks ◽  
Photovoltaic Material

Molecular discovery seeks to generate chemical species tailored to very specific needs. In this paper, we present ORGANIC, a framework based on Objective-Reinforced Generative Adversarial Networks (ORGAN), capable of producing a distribution over molecular space that matches with a certain set of desirable metrics. This methodology combines two successful techniques from the machine learning community: a Generative Adversarial Network (GAN), to create non-repetitive sensible molecular species, and Reinforcement Learning (RL), to bias this generative distribution towards certain attributes. We explore several applications, from optimization of random physicochemical properties to candidates for drug discovery and organic photovoltaic material design.

scholarly journals Dynamic Stability Enhancement of a Hybrid Renewable Energy System in Stand-Alone Applications

Computation ◽  
2021 ◽  
Vol 9 (2) ◽  
pp. 14
Author(s):  
Ezzeddine Touti ◽  
Hossem Zayed ◽  
Remus Pusca ◽  
Raphael Romary
Keyword(s):  
Renewable Energy ◽  
Energy Systems ◽  
Experimental Tests ◽  
Energy System ◽  
Induction Generator ◽  
Technical Solution ◽  
Hybrid Energy ◽  
Wind Speeds ◽  
Hybrid Renewable Energy System ◽  
Stability Enhancement

Renewable energy systems have been extensively developed and they are attractive to become widespread in the future because they can deliver energy at a competitive price and generally do not cause environmental pollution. However, stand-alone energy systems may not be practical for satisfying the electric load demands, especially in places having unsteady wind speeds with high unpredictability. Hybrid energy systems seem to be a more economically feasible alternative to satisfy the energy demands of several isolated clients worldwide. The combination of these systems makes it possible to guarantee the power stability, efficiency, and reliability. The aim of this paper is to present a comprehensive analysis and to propose a technical solution to integrate a self-excited induction generator in a low power multisource system. Therefore, to avoid the voltage collapsing and the machine demagnetization, the various parameters have to be identified. This procedure allows for the limitation of a safe operating area where the best stability of the machine can be obtained. Hence, the load variation interval is determined. An improvement of the induction generator stability will be analyzed. Simulation results will be validated through experimental tests.

Machine Learning Data Center Workloads Using Generative Adversarial Networks

2020 ◽  
Vol 48 (2) ◽  
pp. 21-23
Author(s):  
Boudewijn R. Haverkort ◽  
Felix Finkbeiner ◽  
Pieter-Tjerk de Boer
Keyword(s):  
Machine Learning ◽  
Data Center ◽  
Generative Adversarial Networks ◽  
Adversarial Networks ◽  
Learning Data

scholarly journals Day-ahead Market Modelling of Large-scale Highly-renewable Multi-energy Systems: Analysis of the North Sea Region Towards 2050

2020 ◽  
Author(s):  
Juan Gea Bermúdez ◽  
Kaushik Das ◽  
Hardi Koduvere ◽  
Matti Juhani Koivisto
Keyword(s):  
North Sea ◽  
Large Scale ◽  
Unit Commitment ◽  
District Heating ◽  
Energy Systems ◽  
Energy System ◽  
Computational Time ◽  
Integer Variables ◽  
The North ◽  
The North Sea

This paper proposes a mathematical model to simulate Day-ahead markets of large-scale multi-energy systems with high share of renewable energy. Furthermore, it analyses the importance of including unit commitment when performing such analysis. The results of the case study, which is performed for the North Sea region, show the influence of massive renewable penetration in the energy sector and increasing electrification of the district heating sector towards 2050, and how this impacts the role of other energy sources such as thermal and hydro. The penetration of wind and solar is likely to challenge the need for balancing in the system as well as the profitability of thermal units. The degree of influence of the unit commitment approach is found to be dependent on the configuration of the energy system. Overall, including unit commitment constraints with integer variables leads to more realistic behaviour of the units, at the cost of increasing considerably the computational time. Relaxing integer variables reduces significantly the computational time, without highly compromising the accuracy of the results. The proposed model, together with the insights from the study case, can be specially useful for system operators for optimal operational planning.

scholarly journals Day-ahead renewable scenario forecasts based on generative adversarial networks

2020 ◽  
Author(s):  
Congmei Jiang ◽  
Yongfang Mao ◽  
Yi Chai ◽  
Mingbiao Yu
Keyword(s):  
Power Systems ◽  
Renewable Resources ◽  
Large Scale ◽  
Generative Adversarial Networks ◽  
Power Sources ◽  
Renewable Power ◽  
Adversarial Networks ◽  
Future Behavior ◽  
Large Scale Integration ◽  
Scale Integration

<p>With the increasing penetration of renewable resources such as wind and solar, the operation and planning of power systems, especially in terms of large-scale integration, are faced with great risks due to the inherent stochasticity of natural resources. Although this uncertainty can be anticipated, the timing, magnitude, and duration of fluctuations cannot be predicted accurately. In addition, the outputs of renewable power sources are correlated in space and time, and this brings further challenges for predicting the characteristics of their future behavior. To address these issues, this paper describes an unsupervised method for renewable scenario forecasts that considers spatiotemporal correlations based on generative adversarial networks (GANs), which have been shown to generate high-quality samples. We first utilized an improved GAN to learn unknown data distributions and model the dynamic processes of renewable resources. We then generated a large number of forecasted scenarios using stochastic constrained optimization. For validation, we used power-generation data from the National Renewable Energy Laboratory wind and solar integration datasets. The experimental results validated the effectiveness of our proposed method and indicated that it has significant potential in renewable scenario analysis.</p>

scholarly journals RoCGAN: Robust Conditional GAN

2020 ◽  
Vol 128 (10-11) ◽  
pp. 2665-2683 ◽  
Author(s):  
Grigorios G. Chrysos ◽  
Jean Kossaifi ◽  
Stefanos Zafeiriou
Keyword(s):  
Large Scale ◽  
Real Data ◽  
Superior Performance ◽  
Target Space ◽  
Generative Adversarial Networks ◽  
Natural Scenes ◽  
Adversarial Networks ◽  
Target Manifold ◽  
The Face ◽  
Intense Noise

Abstract Conditional image generation lies at the heart of computer vision and conditional generative adversarial networks (cGAN) have recently become the method of choice for this task, owing to their superior performance. The focus so far has largely been on performance improvement, with little effort in making cGANs more robust to noise. However, the regression (of the generator) might lead to arbitrarily large errors in the output, which makes cGANs unreliable for real-world applications. In this work, we introduce a novel conditional GAN model, called RoCGAN, which leverages structure in the target space of the model to address the issue. Specifically, we augment the generator with an unsupervised pathway, which promotes the outputs of the generator to span the target manifold, even in the presence of intense noise. We prove that RoCGAN share similar theoretical properties as GAN and establish with both synthetic and real data the merits of our model. We perform a thorough experimental validation on large scale datasets for natural scenes and faces and observe that our model outperforms existing cGAN architectures by a large margin. We also empirically demonstrate the performance of our approach in the face of two types of noise (adversarial and Bernoulli).

scholarly journals Effect of Heat Demand on Integration of Urban Large-Scale Renewable Schemes—Case of Helsinki City (60 °N)

Energies ◽  
2020 ◽  
Vol 13 (9) ◽  
pp. 2164
Author(s):  
Vahid Arabzadeh ◽  
Peter D. Lund
Keyword(s):  
Wind Power ◽  
Large Scale ◽  
Energy Use ◽  
Energy Systems ◽  
Energy System ◽  
Heat Pumps ◽  
Electricity Production ◽  
Population Increase ◽  
Building Energy Efficiency ◽  
Heat Demand

Heat demand dominates the final energy use in northern cities. This study examines how changes in heat demand may affect solutions for zero-emission energy systems, energy system flexibility with variable renewable electricity production, and the use of existing energy systems for deep decarbonization. Helsinki city (60 °N) in the year 2050 is used as a case for the analysis. The future district heating demand is estimated considering activity-driven factors such as population increase, raising the ambient temperature, and building energy efficiency improvements. The effect of the heat demand on energy system transition is investigated through two scenarios. The BIO-GAS scenario employs emission-free gas technologies, bio-boilers and heat pumps. The WIND scenario is based on large-scale wind power with power-to-heat conversion, heat pumps, and bio-boilers. The BIO-GAS scenario combined with a low heat demand profile (−12% from 2018 level) yields 16% lower yearly costs compared to a business-as-usual higher heat demand. In the WIND-scenario, improving the lower heat demand in 2050 could save the annual system 6–13% in terms of cost, depending on the scale of wind power.

scholarly journals Large-Scale Distributed Training Applied to Generative Adversarial Networks for Calorimeter Simulation

2019 ◽  
Vol 214 ◽  
pp. 06025
Author(s):  
Jean-Roch Vlimant ◽  
Felice Pantaleo ◽  
Maurizio Pierini ◽  
Vladimir Loncar ◽  
Sofia Vallecorsa ◽  
...  
Keyword(s):  
Large Scale ◽  
Message Passing Interface ◽  
Scale Up ◽  
High Energy ◽  
Neural Nets ◽  
Generative Adversarial Networks ◽  
Detector Response ◽  
Data Set ◽  
Distributed Training ◽  
Adversarial Networks

In recent years, several studies have demonstrated the benefit of using deep learning to solve typical tasks related to high energy physics data taking and analysis. In particular, generative adversarial networks are a good candidate to supplement the simulation of the detector response in a collider environment. Training of neural network models has been made tractable with the improvement of optimization methods and the advent of GP-GPU well adapted to tackle the highly-parallelizable task of training neural nets. Despite these advancements, training of large models over large data sets can take days to weeks. Even more so, finding the best model architecture and settings can take many expensive trials. To get the best out of this new technology, it is important to scale up the available network-training resources and, consequently, to provide tools for optimal large-scale distributed training. In this context, our development of a new training workflow, which scales on multi-node/multi-GPU architectures with an eye to deployment on high performance computing machines is described. We describe the integration of hyper parameter optimization with a distributed training framework using Message Passing Interface, for models defined in keras [12] or pytorch [13]. We present results on the speedup of training generative adversarial networks trained on a data set composed of the energy deposition from electron, photons, charged and neutral hadrons in a fine grained digital calorimeter.

Sign in / Sign up

Export Citation Format

Share Document