Use of Available Daylight to Improve Short-Term Load Forecasting Accuracy

Miguel López,Carolina Senabre,Sergio Valero,Carlos Sans

doi:10.3390/en14010095

Miguel López, Carolina Senabre + Show 2 more

Open Access

https://doi.org/10.3390/en14010095

Copy DOI

Journal: Energies	Publication Date: Dec 26, 2020
Citations: 6	License type: CC BY 4.0

Affiliation: Miguel Hernandez University

Abstract

This paper introduces a new methodology to include daylight information in short-term load forecasting (STLF) models. The relation between daylight and power consumption is obvious due to the use of electricity in lighting in general. Nevertheless, very few STLF systems include this variable as an input. In addition, an analysis of one of the current STLF models at the Spanish Transmission System Operator (TSO), shows two humps in its error profile, occurring at sunrise and sunset times. The new methodology includes properly treated daylight information in STLF models in order to reduce the forecasting error during sunrise and sunset, especially when daylight savings time (DST) one-hour time shifts occur. This paper describes the raw information and the linearization method needed. The forecasting model used as the benchmark is currently used at the TSO’s headquarters and it uses both autoregressive (AR) and neural network (NN) components. The method has been designed with data from the Spanish electric system from 2011 to 2017 and tested over 2018 data. The results include a justification to use the proposed linearization over other techniques as well as a thorough analysis of the forecast results yielding an error reduction in sunset hours from 1.56% to 1.38% for the AR model and from 1.37% to 1.30% for the combined forecast. In addition, during the weeks in which DST shifts are implemented, sunset error drops from 2.53% to 2.09%.

Highlights

Short-term load forecasting (STLF) is one of the key functions of a Transmission System Operator (TSO) in maintaining technical viability of the system with low operational costs
These deep neural networks (DNNs) have been applied to load forecasting [18,19,20] and they have allowed researchers to reduce the workload of hand-designed feature inputs, which makes them good candidates for smart grid applications [21]
This particular week is especially sensitive to the modeled phenomenon and, it is a good reference value: The results show that the inclusion of available daylight information causes a reduction of the forecasting error in sunrise and sunset times of 0.12 and 0.18 percentage points respectively

Summary

Introduction

Short-term load forecasting (STLF) is one of the key functions of a Transmission System Operator (TSO) in maintaining technical viability of the system with low operational costs. STLF includes forecasts made from one hour to several days ahead It provides useful information for system operators to guarantee the reliability of the system, for generators to optimize schedules [1] and for market participants to generate market biddings on both sides of the market. Recent advances in technology have freed the use of neural networks with more than the typical three layers [17]. These deep neural networks (DNNs) have been applied to load forecasting [18,19,20] and they have allowed researchers to reduce the workload of hand-designed feature inputs, which makes them good candidates for smart grid applications [21]. For a 40 GW system like Spain’s inland network, it is still worth designing these stages for the system

Objectives

Methods

Results

Conclusion