Energy variables for benchmarking in technical schools in São Paulo, Brazil

Published: 27 March 2019| Version 1 | DOI: 10.17632/zwv5d46pp5.1
Haroldo Luiz Nogueira da Silva


Variables: DMU: Decision Making Unit; PDC: Peek demand contracted (peek demand value contracted with the energy company); PDS: Peek demand suggested (peek demand value suggested as ideal); PDD: Peek demand deviation (difference between the contracted and ideal peak demand); AAE_obs: Anual active energy observed (annual energy consumption real); AAE_pred: Anual active energy predicted (annual energy consumption predicted by predictive model - e.g. machine learning methods, linear regression); RAE: Reactive anual energy (annual reactive energy consumption); AEN_nsd: Active energy in non-school months (energy consumption in non-school months); TNS: Total number of students (total number of students in the school).


Steps to reproduce

The experiment in its original formulation consists of using Data Envelopment Analysis (DEA) to perform an energy benchmarking model to categorize schools through their efficiencies. Input variables: PDD, AAE_obs, RAE, AEN_nsd. Output variables: AEE_pred, TNS.


Universidade Federal do ABC