A Clustering Approach for the Optimal Siting of Recharging Stations in the Electric Vehicle Routing Problem with Time Windows

Transportation has been incorporating electric vehicles (EVs) progressively. EVs do not produce air or noise pollution, and they have high energy efficiency low maintenance costs. In this context, the development of efficient techniques to overcome vehicle routing problem becomes crucial with proliferation EVs. The concerns freight capacity battery autonomy limitations in different delivery-service scenarios, challenge best locating recharging stations. This work proposes a mixed-integer linear programming model solve location time windows (E-LRPTW) considering state charge, capacities, customer decision model. A clustering strategy based on k-means algorithm is proposed divide set vertices into small areas define potential sites for stations, while reducing number binary variables. E-LRPTW was implemented Python solved using mathematical modeling language AMPL together CPLEX. Performed tests instances 5 10 clients showed large reduction required find solution (by about 60 times one instance). It concluded that dividing customers by sectors be applied generate solutions larger geographical numbers determine station locations as part planning decisions more realistic scenarios.