Big Data Analytics for Spatio-Temporal Service Orders Demand Forecasting in Electric Distribution Utilities

Vitor Hugo Ferreira,Natanael Rodrigues Pereira,Angelo Cesar Colombini,Márcio Zamboti Fortes,Flávio Luis De Mello,Rubens Lucian Da Silva Correa,Fernando Carvalho Cid De Araujo

doi:10.3390/en14237991

Vitor Hugo Ferreira, Natanael Rodrigues Pereira + Show 5 more

Open Access

https://doi.org/10.3390/en14237991

Copy DOI

Journal: Energies	Publication Date: Nov 30, 2021
Citations: 3	License type: CC BY 4.0

Affiliation: Universidade Federal Fluminense

Abstract

This paper presents a big data analytics-based model developed for electric distribution utilities aiming to forecast the demand of service orders (SOs) on a spatio-temporal basis. Being fed by robust history and location data from a database provided by an energy utility that is using this innovative system, the algorithm automatically forecasts the number of SOs that will need to be executed in each location in several time steps (hourly, monthly and yearly basis). The forecasted emergency SOs demand, which is related to energy outages, are stochastically distributed, projecting the impacted consumers and its individual interruption indexes. This spatio-temporal forecasting is the main input for a web-based platform for optimal bases allocation, field team sizing and scheduling implemented in the eleven distribution utilities of Energisa group in Brazil.

Highlights

The electric distribution utilities are responsible for supplying energy to an extensive number of consumers spanned across vast geographic regions
This paper presents a big data analytics-based algorithm for spatio-temporal service orders demand forecasting in electric distribution utilities
Combining georeferenced algorithms for historical data processing and statistical methods for projection and analysis, the methodology is implemented as a module in an intelligent workforce planning system called AWDEC (Brazilian acronym for web-based workforce planning system—Aplicação Web para Dimensionamento de Equipes de Campo)

Summary

Introduction

The electric distribution utilities are responsible for supplying energy to an extensive number of consumers spanned across vast geographic regions. Workforce planning aims to optimally allocate and size work teams to perform a given task on a specific time horizon This problem includes the proposition (strategic planning) of operational facilities (depots) where the teams start their daily work routine, sizing (tactical planning) the number of workers in each depot and distributing (operational planning) each staff member in a work schedule. These studies are necessarily based on some spatio-temporal service demand predictions, which must be conducted for different forecasting horizons (long, medium and short term) and for distinct geographical scales (city, district, zone)

Methods

Results

Conclusion