Samenvatting
Electric vehicles are gradually gaining attention due to our unsustainable transport system, but are still coping with adoption barriers such as range anxiety. Placing public charging stations is one possible solution for mitigating the anxiety. In this paper, the availability of charging infrastructure is modeled
by regression techniques that aim at forecasting the activity of charging infrastructure. Two regression models are assumed, namely, a global linear regression and a local geographically weighted regression. The data originates from the Flemish Living Labs and is used for fitting the models. The results show that the models can properly fit the data, but that the data is not spatial dependent unlike expected. Therefore, geographically weighted regression does not always provide the expected improvements.
by regression techniques that aim at forecasting the activity of charging infrastructure. Two regression models are assumed, namely, a global linear regression and a local geographically weighted regression. The data originates from the Flemish Living Labs and is used for fitting the models. The results show that the models can properly fit the data, but that the data is not spatial dependent unlike expected. Therefore, geographically weighted regression does not always provide the expected improvements.
Originele taal-2 | English |
---|---|
Status | Unpublished - 1 dec. 2015 |
Evenement | European Electric Vehicle Congress 2015 - Diamant Centre, Brussels, Belgium Duur: 1 dec. 2015 → 4 dec. 2015 |
Conference
Conference | European Electric Vehicle Congress 2015 |
---|---|
Land/Regio | Belgium |
Stad | Brussels |
Periode | 1/12/15 → 4/12/15 |