The impact of gradually replacing fossil fuel-powered vehicles with electric ones: A bi-objective optimisation approach

In the last two decades, many cities imposed environmental regulations that mandate companies to stop using fuel-powered vehicles once or twice weekly depending on the air quality and their identification numbers. The regulation limits the companies ability to fulfil client demands. Many companies m...

Full description

Saved in:
Bibliographic Details
Published in:Expert systems with applications 2022-05, Vol.194, p.116546, Article 116546
Main Authors: Romero-Ocaño, A. Danisa, Cosío-León, M.A., Valenzuela-Alcaraz, Víctor M., Brizuela, Carlos A.
Format: Article
Language:English
Subjects:
Air quality
Conventional vehicles
Electric vehicles
Evolutionary algorithms
Fossil fuels
Impact analysis
Inspection
Mix fleet
Multiple objective analysis
Optimization
Vehicle restriction programs
Vehicle routing problem with time windows
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In the last two decades, many cities imposed environmental regulations that mandate companies to stop using fuel-powered vehicles once or twice weekly depending on the air quality and their identification numbers. The regulation limits the companies ability to fulfil client demands. Many companies may adopt alternative strategies to overcome this constraint by replacing the stopped vehicles with electric units. However, this replacement impacts the fleet’s performance. In the present study, we analyse the impact on fleet’s performances by modelling the problem as a Pareto front degradation, when replacing conventional vehicles (CVs) with electric vehicles (EVs). To this end, we base our analysis on a visual inspection of the non-dominated fronts and a coverage measure. To obtain good quality non-dominated fronts, we improve a multi-objective evolutionary algorithm (MOEA/D) by introducing a novel, simple and effective post-processing stage applied to each non-dominated solution obtained from the MOEA/D. The analysis and computational experiments offer three main results: i) The post-processing algorithm improves almost every single non-dominated front generated by the MOEA/D. ii) Contrary to the expected results, the number of CVs to be replaced without affecting the fleet performance is large, between half and two-thirds of the total number of vehicles, iii) The proposed model here will help companies to find the appropriate number of CVs to be replaced without affecting their service quality.
ISSN:0957-4174
1873-6793
DOI:10.1016/j.eswa.2022.116546