Summary We consider an application in electricity grid load prediction, where generalized additive models are appropriate, but the data set's size can make their use practically intractable with existing methods. therefore develop practical model fitting methods for large sets case which smooth terms represented by using penalized regression splines. The iterative update schemes to obtain factors of matrix while requiring only subblocks be computed at any one time. show that efficient...

Electricity load forecasting faces rising challenges due to the advent of innovating technologies such as smart grids, electric cars and renewable energy production. For distribution network managers, a good knowledge future electricity consumption stands central point for reliability investment strategies. In this paper, we suggest semi-parametric approach based on generalized additive models theory model electrical over more than 2200 substations French network, at both short middle term...

We propose a novel framework for fitting additive quantile regression models, which provides well calibrated inference about the conditional quantiles and fast automatic estimation of smoothing parameters, model structures as diverse those usable with distributional GAMs, while maintaining equivalent numerical efficiency stability. The proposed methods are at once statistically rigorous computationally efficient, because they based on general belief updating Bissiri et al. (2016) to loss...

In the last two decades, growth of computational resources has made it possible to handle generalized additive models (GAMs) that formerly were too costly for serious applications. However, in model complexity not been matched by improved visualizations development and results presentation. Motivated an industrial application electricity load forecasting, we identify areas where lack modern visualization tools GAMs is particularly severe, address shortcomings existing methods proposing a set...

The coronavirus disease 2019 (COVID-19) pandemic has urged many governments in the world to enforce a strict lockdown where all nonessential businesses are closed and citizens ordered stay at home. One of consequences this policy is significant change electricity consumption patterns. Since load forecasting models rely on calendar or meteorological information trained historical data, they fail capture break caused by have exhibited poor performances since beginning pandemic. In paper we...

Generalized additive models (GAMs) are flexible non-linear regression models, which can be fitted efficiently using the approximate Bayesian methods provided by mgcv R package. While GAM based on assumption that response distribution is modeled parametrically, here we discuss more do not entail any parametric assumption. In particular, this article introduces qgam package, an extension of providing fast calibrated for fitting quantile GAMs (QGAMs) in R. QGAMs a smooth version pinball loss...

We propose a hybrid approach for the modeling and short-term forecasting of electricity loads. Two building blocks our are (1) overall trend seasonality by fitting generalized additive model to weekly averages load (2) dependence structure across consecutive daily loads via curve linear regression. For latter, new methodology is proposed regression with both response regressors. The key idea behind dimension reduction based on singular value decomposition in Hilbert space, which reduces...

The development of smart grid and new advanced metering infrastructures induces opportunities challenges for utilities. Exploiting meters information forecasting stands as a key point energy providers who have to deal with time varying portfolio customers well managers needs improve accuracy local forecasts face distributed renewable generation development. We propose machine learning approach forecast the system load group exploiting individual measurements in real and/or exogenous like...

We present the winning strategy for IEEE DataPort Competition on Day-Ahead Electricity Load Forecasting: Post-Covid Paradigm. This competition was organized to design new forecasting methods unstable periods such as one starting in Spring 2020. First, we pre-process data with a statistical correction of meteorological variables. Second, apply standard and machine learning models. Third, rely state-space models adapt aforementioned forecasters. It achieves right compromise between two...

French electricity load forecasting has encountered major changes during the past decade. These are, among other things, due to opening of market and economic crisis, which require development new automatic time adaptive prediction methods. The advent innovating technologies also needs some methods because thousands or tens series have be studied. In this paper we adopt for a semi-parametric approach based on additive models. We present an procedure explanatory variable selection in model...

The transition to electric vehicles (EVs) presents challenges and opportunities for the management of electrical networks. This paper focuses on developing evaluating probabilistic forecasting algorithms understand predict EV charging behaviours, crucial optimising grid operations ensuring a balance between electricity demand generation. Several approaches tailored different time horizons are proposed across diverse model classes, including direct, bottom-up, adaptive approaches. In all...

To meet France’s CO2 emission reduction of 33 % by 2030 compared to 1990 and reach greenhouse gas neutrality in 2050, sustainable energy sources are key clean power production reduced emissions from the sector. However, non-dispatchable renewables such as wind solar photovoltaic (PV) require accurate forecasts improve their grid stability, reliability, penetration level not mention supply-demand matching. Indeed, those dependent on weather conditions radiation or speeds, making...

In the last two decades growth of computational resources has made it possible to handle Generalized Additive Models (GAMs) that formerly were too costly for serious applications. However, in model complexity not been matched by improved visualisations development and results presentation. Motivated an industrial application electricity load forecasting, we identify areas where lack modern visualisation tools GAMs is particularly severe, address shortcomings existing methods proposing a set...

Smart grids require flexible data driven forecasting methods. We propose clustering tools for bottom-up short-term load forecasting. focus on individual consumption analysis which plays a major role energy management and electricity The first section is dedicated to the industrial context review of electrical analysis. Then, we hierarchical time-series idea decompose global signal obtain disaggregated forecasts in such way that their sum enhances prediction. This done three steps: identify...