Title :
Analysis of forecasting algorithms for minimization of electric demand costs for electric vehicle charging in commercial and industrial environments
Author :
Jewell, Nicholas ; Turner, Matthew ; Naber, John ; McIntyre, Michael
Author_Institution :
Dept. of Electr. & Comput. Eng., Univ. of Louisville, Louisville, KY, USA
Abstract :
The large scale deployment of Electric Vehicle (EV) charging infrastructure can result in high added utility costs due to the peak demand cost structure in utility bills for commercial and industrial users. A method to minimize this disincentive to EV adoption is proposed that relies on forecasting demand so that EV charging activity can be intelligently controlled. This study examines multiple forecasting models and techniques to determine the optimal algorithm for use in the proposed control system. Simulation results are presented for each of the forecasting algorithms with the best mean absolute percent error of 1.26% using a neural network with averaging. This results in a reduction in the peak demand electric costs of approximately 95%.
Keywords :
electric vehicles; electrical engineering computing; neural nets; EV charging infrastructure; electric demand cost minimization; electric vehicle charging; forecasting algorithm analysis; high added utility costs; industrial environments; multiple forecasting models; neural network; peak demand cost structure; peak demand electric cost reduction; utility bills; Algorithm design and analysis; Electricity; Forecasting; Linear regression; Neural networks; Predictive models; Regression tree analysis; Charge Control; Electric Vehicle (EV); Electric Vehicle Service Equipment (EVSE); Electricity Demand; Forecast; PMCS;
Conference_Titel :
Transportation Electrification Conference and Expo (ITEC), 2012 IEEE
Conference_Location :
Dearborn, MI
Print_ISBN :
978-1-4673-1407-7
Electronic_ISBN :
978-1-4673-1406-0
DOI :
10.1109/ITEC.2012.6243461
