Branch-and-cut-and-price for the Electric Vehicle Routing Problem with Time Windows, Piecewise-Linear Recharging and Capacitated Recharging Stations

https://doi.org/10.1016/j.cor.2022.105870Get rights and content

Highlights

  • This paper considers the routing of electric vehicles and their recharging en route.

  • Recharging the batteries of the vehicles is governed by a piecewise-linear function.

  • Vehicles must recharge at stations with a limited number of chargers.

  • Hence, the vehicles must be scheduled around the availability of a charger.

  • The problem is tackled using a branch-and-cut-and-price algorithm.

  • The algorithm decomposes the routing to integer programming and the scheduling to constraint programming.

  • Results show that this hybrid algorithm solves instances with up to 100 customers.

Abstract

The Electric Vehicle Routing Problem with Time Windows, Piecewise-Linear Recharging and Capacitated Recharging Stations aims to design minimum-cost routes for a fleet of electric vehicles subject to intra-route and inter-route constraints. Every vehicle is equipped with a rechargeable battery that depletes while it transports goods along its route. A vehicle must detour to a recharging station to recharge before draining its battery. To approximate a real recharging process, the amount of energy restored is modeled as a piecewise-linear function of the time spent recharging. Furthermore, each station has a small number of chargers, and hence, when and where a vehicle can recharge must be scheduled around the availability of a charger. This interaction between vehicles does not appear in classical vehicle routing problems and motivates the development of new methods that can exploit the joint routing and scheduling structure. This paper proposes a branch-and-cut-and-price algorithm that designates the routing to integer programming using Dantzig–Wolfe decomposition and the scheduling to constraint programming using logic-based Benders decomposition. Experimental results indicate that this hybrid method solves 34% of the instances with 100 customers.

Introduction

This paper proposes the Electric Vehicle Routing Problem with Time Windows, Piecewise-Linear Recharging and Capacitated Recharging Stations (EVRPTW-PLR-CRS). The EVRPTW-PLR-CRS adds two complications to the Electric Vehicle Routing Problem with Time Windows (EVRPTW) which is itself based on the classical Vehicle Routing Problem with Time Windows (VRPTW).

The VRPTW appoints a fleet of identical vehicles to deliver packages from a central depot to customers before returning to the depot at the end of their routes (see Vigo and Toth, 2014, Costa et al., 2019). Every package is associated with a quantity of load. Each vehicle can carry packages up to a maximum total load called the vehicle capacity. The customer of every package is given a time frame, called the time window, within which the package will be delivered. Vehicles can arrive at a customer prior to the time window but must wait until the window opens before commencing delivery. Performing a delivery requires an amount of time called its service duration. The goal of the VRPTW is to compute routes that deliver every package while minimizing the total travel distance.

The EVRPTW extends the VRPTW to electric vehicles, each equipped with a rechargeable battery (Desaulniers et al., 2016). The battery of a vehicle is initially fully charged at the depot and depletes while the vehicle travels along its route. The battery must maintain a minimum amount of energy, denoted as 0, at all times. Vehicles must detour to a recharging station to recharge before draining their battery. The recharge rate is assumed to be constant. Therefore, the energy replenished is a linear function of the time spent recharging.

In practice, the recharge rate progressively slows to avoid damaging the battery. It can be accurately modeled as a piecewise-constant function (Montoya et al., 2017), which leads to a piecewise-linear function for representing the energy restored. The EVRPTW-PLR-CRS includes this piecewise-linear recharging function absent in the EVRPTW.

Much of the literature on routing electric vehicles assumes that the recharging stations can simultaneously recharge an unlimited number of vehicles. In practice, the stations are often equipped with only a few chargers. Hence, a vehicle may arrive at a recharging station to find that all chargers are already in use by other vehicles and, consequently, has to wait until a charger becomes available. The issue of waiting is exacerbated by the slow recharging of batteries today. The EVRPTW-PLR-CRS supplements the routing structure of the EVRPTW with a novel scheduling structure at each recharging station that both determines when a vehicle should recharge and bounds the total number of vehicles simultaneously recharging. This type of scheduling is seen at privately-owned recharging facilities where the operator has full control of the chargers. Appendix A presents an example that illustrates the combined routing and scheduling.

The scheduling structure in the EVRPTW-PLR-CRS substantially complicates the VRPTW. Every vehicle in the VRPTW can be considered independent: as long as the packages delivered along a route fit within the vehicle capacity and can be delivered on-time, delaying the route does not impact the routes of the other vehicles. In the EVRPTW, adding the energy constraints retains the independence of the vehicles: as long as a vehicle does not completely drain its battery, the routes of the other vehicles remain unaffected. However, the recharging station scheduling of the EVRPTW-PLR-CRS makes the problem much more difficult. The vehicles are no longer independent as they interact at the charging stations. To deliver the on-board packages on time, the vehicles are required to self-organize at the stations to determine the order in which the vehicles recharge.

The main contributions of this paper are an exponential-size formulation of the EVRPTW-PLR-CRS and an accompanying exact optimization algorithm that elegantly decomposes the joint routing and scheduling structure to the best technologies available for exploiting these structures. It is well-known that integer programming and particularly branch-and-cut-and-price (BCP) power the state-of-the-art exact methods for many vehicle routing problems including the VRPTW (see Vigo and Toth, 2014, Costa et al., 2019) and the EVRPTW (Desaulniers et al., 2020). It is also well-known that constraint programming generally outperforms integer programming at scheduling problems (Schutt et al., 2010, Heinz, 2018). This paper introduces a hybrid BCP algorithm that uses integer programming for routing via a Dantzig–Wolfe decomposition (see Lübbecke and Desrosiers, 2005, Desaulniers et al., 2005) and defers the scheduling to a constraint programming subproblem using the generic form (Lam and Van Hentenryck, 2017, Davies et al., 2017, Lam et al., 2020) of logic-based Benders decomposition (Hooker and Ottosson, 2003). The paper also presents an ad-hoc technique to strengthen the Benders cuts and a polyhedral analysis to further lift the Benders cuts. Experimental results indicate that the BCP algorithm solves 34% of the instances with 100 customers.

The remainder of the paper is organized as follows. Section 2 reviews related problems and solution methods. Section 3 presents a mathematical formulation of the EVRPTW-PLR-CRS and describes the hybrid BCP algorithm. Section 4 compares the empirical performance of the algorithm on different variants of the problem. Section 5 concludes this paper.

Section snippets

Background and related work

This section reviews background material and surveys relevant studies.

The branch-and-cut-and-price algorithm

This section presents an exponential-size model of the EVRPTW-PLR-CRS and an accompanying BCP algorithm for solving it.

Experimental results

This section reports the performance of the BCP algorithm and analyzes the impacts of piecewise-linear recharging and the station capacity constraints.

Conclusion and future work

This paper proposes an exact method for solving the EVRPTW-PLR-CRS. The EVRPTW-PLR-CRS includes a more realistic non-linear recharging procedure and recharging stations with a limited number of chargers which must be shared among the vehicles. To ensure that the vehicles achieve their collective goal of delivering the packages to all customers on time, the vehicles are required to self-organize at the recharging stations. This interaction leads to a novel scheduling structure rarely seen in

CRediT authorship contribution statement

Edward Lam: Conceptualization, Methodology, Software, Investigation, Writing – original draft, Writing – review & editing. Guy Desaulniers: Investigation, Writing – review & editing, Supervision. Peter J. Stuckey: Supervision, Funding acquisition.

Funding sources

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

References (43)

  • CostaL. et al.

    Exact Branch-Price-and-Cut algorithms for vehicle routing

    Transp. Sci.

    (2019)
  • Davies, T.O., Gange, G., Stuckey, P.J., 2017. Automatic Logic-Based Benders Decomposition with MiniZinc. In:...
  • DesaulniersG. et al.

    Cutting planes for branch-and-price algorithms

    Networks

    (2011)
  • DesaulniersG. et al.

    Exact algorithms for electric vehicle-routing problems with time windows

    Oper. Res.

    (2016)
  • DesaulniersG. et al.

    Variable fixing for two-arc sequences in Branch-Price-and-Cut algorithms on path-based models

    Transp. Sci.

    (2020)
  • DrexlM.

    Synchronization in vehicle routing–a survey of VRPs with multiple synchronization constraints

    Transp. Sci.

    (2012)
  • El HachemiN. et al.

    A hybrid constraint programming approach to the log-truck scheduling problem

    Ann. Oper. Res.

    (2011)
  • FeydyT. et al.

    Lazy clause generation reengineered

  • FrogerA. et al.

    The electric vehicle routing problem with capacitated charging stations

    Transp. Sci.

    (2021)
  • GamrathG. et al.

    The SCIP Optimization Suite 7.0Technical Report

    (2020)
  • Cited by (34)

    • A ridesharing routing problem for airport riders with electric vehicles

      2024, Transportation Research Part E: Logistics and Transportation Review
    View all citing articles on Scopus
    View full text