Predicting solar generation from weather forecasts using machine learning

Sharma, Naveen; Sharma, Pranshu; Irwin, David; Shenoy, Prashant

doi:10.1109/smartgridcomm.2011.6102379

Cited by 368 publications

(207 citation statements)

References 3 publications

Supporting

Mentioning

202

Contrasting

Unclassified

Order By: Relevance

“…Such methods may consist of using techniques such as artificial neural networks (ANN-Mellit 2008;Wang et al 2012), autoregressive models (Hassanzadeh et al 2010;Yang et al 2012), Markov process models (Morf 2014) or support vector machines (Sharma et al 2011;Bouzerdoum et al 2013) to recognize patterns in the changes of wind speed or solar radiation. Blended techniques, such as that of Pedro and Coimbra (2012), that use a genetic algorithm to optimize an ANN have also been effective.…”

Section: Nowcastingmentioning

confidence: 99%

Short-Range Forecasting for Energy

Haupt

2018

Weather &Amp; Climate Services for the Energy Industry

View full text Add to dashboard Cite

Short-range forecasts for periods on the order of hours to days and up to two weeks ahead are necessary to smoothly run transmission and distribution systems, plan maintenance, protect infrastructure and allocate units. In particular, forecasting the renewable energy resources on a day-to-day basis enables integration of increasing capacities of these variable resources. This chapter describes the basics of this short-range forecasting, beginning with the observation-based "nowcasting" of the first 15 minutes and ranging up to two weeks using numerical weather prediction. We discuss how blending multiple forecasts can increase accuracy and how probabilistic forecasts are being used to quantify the forecast uncertainty.

show abstract

Section: Nowcastingmentioning

confidence: 99%

Short-Range Forecasting for Energy

Haupt

2018

Weather &Amp; Climate Services for the Energy Industry

View full text Add to dashboard Cite

show abstract

“…The model is able to predict next day energy harvesting based on weather forecasts. They improved accuracy by 27% with machine learning techniques (Sharma et al, 2011).…”

Section: Related Workmentioning

confidence: 99%

A CO2 Emissions Accounting Framework with Market-based Incentives for Cloud Infrastructures

Margery

Guyon²,

Orgerie

et al. 2017

Proceedings of the 6th International Conference on Smart Cities and Green ICT Systems

View full text Add to dashboard Cite

Abstract:CO2 emissions related to Cloud computing reach nowadays worrying levels, without any reduction in sight. Often, Cloud users, asking for virtual machines, are not aware of such emissions which concern the entire Cloud infrastructures and are thus difficult to split into the actual resources utilization, such as virtual machines. We propose a CO2 emissions accounting framework giving flexibility to the Cloud providers, predictability to the users and allocating all the carbon costs to the users. This paper shows the architecture of our accounting framework and ideas on how to practically implement it.

show abstract

“…The commonly used statistical forecast methods are time series method [7], artificial neural network (ANN) method [8] and support vector machine (SVM) [9]. In [11], prediction models for solar power generation are built. Comparing with multiple regression techniques for generating prediction models, including linear least squares and support vector machines, simulation shows that SVM-based prediction models are more accurate.…”

Section: Introductionmentioning

confidence: 99%

Photovoltaic power forecasting with a rough set combination method

Yang

Ren

Yue

2016

2016 UKACC 11th International Conference on Control (CONTROL)

View full text Add to dashboard Cite

Abstract-One major challenge with integrating photovoltaic (PV) systems into the grid is that its power generation is intermittent and uncontrollable due to the variation in solar radiation. An accurate PV power forecasting is crucial to the safe operation of the grid connected PV power station. In this work, a combined model with three different PV forecasting models is proposed based on a rough set method. The combination weights for each individual model are determined by rough set method according to its significance degree of condition attribute. The three different forecasting models include a past-power persistence model, a support vector machine (SVM) model and a similar data prediction model. The case study results show that, in comparison with each single forecasting model, the proposed combined model can identify the amount of useful information in a more effective manner.

show abstract

Predicting solar generation from weather forecasts using machine learning

Cited by 368 publications

References 3 publications

Short-Range Forecasting for Energy

Short-Range Forecasting for Energy

A CO2 Emissions Accounting Framework with Market-based Incentives for Cloud Infrastructures

Photovoltaic power forecasting with a rough set combination method

Contact Info

Product

Resources

About