---
title: Station vs. Route Charging Options
url: https://www.emergentmind.com/topics/station-based-or-route-based-charging-options
type: topic
---

# Station vs. Route Charging Options

Station-based or route-based charging options are two broad ways of organizing electric-vehicle energy replenishment. In station-based charging, energy is delivered at fixed physical facilities such as public charging stations, depots, hubs, EVSE-equipped sites, or base-station-centric hubs. In route-based charging, the charging decision is embedded in route execution, either because vehicles are assigned to stations on their path, because charging is planned at intermediate points during service, or because energy is transferred while the vehicle is moving through dynamic wireless charging or Mobile Energy Disseminators. A substantial part of the recent literature is hybrid: it retains fixed stations while jointly optimizing route choice, station choice, charging quantity, timing, congestion management, and grid interaction [2508.15833], [1903.06388], [2509.07402], [1710.03726].

## 1. Taxonomy and definitional boundaries

The clearest station-based formulations model charging as a local service process at a fixed site. A single charging station under peak and non-peak traffic is modeled as an \(M/M/1/k\) queue in "IDEAS: Information-Driven EV Admission in Charging Station Considering User Impatience to Improve QoS and Station Utilization" [2403.06223]. A high-demand fast-charging location with reservation, parking, and charging bundled as one product is studied in "Online Dynamic Pricing for Electric Vehicle Charging Stations with Reservations" [2410.05538]. A base-station-centric Energy-Communication-Transportation Hub reconstructs 5G base stations into fixed charging hubs by adding EVSE, batteries, and renewable generation in "Towards Integrated Energy-Communication-Transportation Hub: A Base-Station-Centric Design in 5G and Beyond" [2508.15833]. Fixed-hub charging also appears in depot-centered and terminal-centered transit settings, including depot charging and intermediate fast-charging stations for electric modular autonomous units [2504.04408].

Route-based charging is not a single modeling convention. One meaning is literal in-motion charging: "Electric Vehicle Routing Problem with Time Windows and Station-based or Route-based Charging Options" models route-based charging as dynamic wireless charging on arcs with partial coverage \(W_{ij}\in[0,1]\), while "Route Optimization of Electric Vehicles based on Dynamic Wireless Charging" models charging from Mobile Energy Disseminators that operate as mobile charging stations on predefined routes [2509.07402], [1710.03726]. Another meaning is operator-routed or route-embedded access to fixed stations: in "Pricing and Routing Mechanisms for Differentiated Services in an Electric Vehicle Public Charging Station Network," users do not directly choose a station; the Charging Network Operator assigns them to stations on their path through routing probabilities \(\mathbf r_{i,j,\ell}\) [1903.06388]. A further variant arises in service scheduling, where charging occurs only at predefined depots or intermediate stations, but those charging decisions are embedded in the route or duty structure [2504.04408], [2209.07305].

This suggests that the literature uses "route-based" in at least two senses: charging while moving and charging planned as part of route or schedule optimization. Conversely, station-based charging is not merely a location decision; it often includes queueing, admission control, pricing, reservation, charging-speed selection, storage dispatch, and V2G contract design [2403.06223], [1612.00106].

## 2. Station-based charging as local infrastructure and service process

Station-based models treat the charging site as the primary optimization locus. In the queueing formulation of [2403.06223], arrivals are Poisson with interarrival distribution
\[
P(\Delta t)=\lambda e^{-\lambda \Delta t},
\]
traffic intensity is \(\rho=\lambda/\mu\), and the model distinguishes forced balking, voluntary balking, and reneging. The paper’s central mechanism is real-time sharing of estimated wait time with arriving EVs, so that users decide whether to join the queue on the basis of station information rather than guesses. It also models the slowdown of fast charging beyond \(80\%\) SoC and proposes a two-mode, two-port charger: fast charge to \(80\%\), then automatic slow charging to \(100\%\) while releasing fast capacity to another EV [2403.06223].

Other station-based papers move the focus from queueing to resource pricing and local energy management. In [2410.05538], the charging station is a set of time-slot resources with finite capacity; a reservation request is a product vector \(p\in\mathbb N_0^n\), feasibility is componentwise \(c\ge p\), and the operator chooses prices online through a finite-horizon MDP. In [1609.09037], the station offers a menu of contracts \(p_{k,l,t}\) indexed by energy level \(l\) and deadline \(t\); the user chooses the contract maximizing surplus \(u_{k,l,t}-p_{k,l,t}\), while the station computes the incremental fulfillment cost \(v_{l,t}-v_{-k}\). In [1612.00106], the contract is expanded to \((l,t,BU)\), where \(BU\) is maximum additional battery utilization, so that V2G discharging can be priced jointly with energy and deadline. In [2604.15227], the discrete access variable \(x_{ij}\in\{0,1\}\) explicitly represents whether EV \(i\) is assigned to EVSE \(j\), and station occupancy is constrained by
\[
\sum_{i\in\mathcal I} s_i(t)x_{ij}\le 1,\quad \forall j,t.
\]

Station-based charging also extends beyond conventional charging plazas. The ECT-Hub architecture in [2508.15833] uses base-station backup batteries as a BESS, augments some sites with PV or wind turbines, and couples EV charging revenue to battery scheduling and incentive pricing. Its grid power requirement is
\[
P_{grid}(t)=\max\{0,P_{BS}(t)+P_{CS}(t)+P_{BP}(t)-P_{WT}(t)-P_{PV}(t)\},
\]
and its overall objective maximizes
\[
\Psi = CR - OC.
\]
Similarly, in the BEB charging scheduler of [2405.11365], station charging is a Position Allocation Problem: each charging visit occupies a time interval and a queue position, with charger assignment variables \(w_{iq}\), linearized charging-duration variables \(g_{iq}\), SOC propagation, and an objective that minimizes charger usage while prioritizing slow charging for battery health.

The common significance of these station-based models is that they internalize local scarcity. The scarce resource may be queue positions, charging posts, EVSE occupancy, time-slot capacity, charging power, battery energy, or station-space usage. The mathematical form changes, but the operational unit remains the station.

## 3. Route-based charging and in-motion replenishment

The most explicit route-based charging models place the energy gain directly on the route. In [2509.07402], a traversed arc \((i,j)\) updates battery charge according to
\[
y_j \le y_i + (w d_{ij} W_{ij} - r d_{ij})x_{ij} + Q(1-x_{ij}),
\]
so route-based wireless charging provides continuous energy transfer while moving, depends on distance and coverage \(W_{ij}\), and can be partial rather than binary. Station-based charging remains available in the same model: a charging-station visit restores the battery to full \(Q\) and incurs fixed charging time \(gQ\). This makes the paper a direct comparison between stopped full recharge and arc-based dynamic replenishment [2509.07402].

In [1710.03726], route-based charging is implemented by Mobile Energy Disseminators, usually buses in cities or trucks on highways, that act as mobile charging stations on predefined routes. An EV contacts a MED, makes an appointment \((time, location)\), forms a platoon, and receives inductive power transfer while both vehicles are moving. The routing objective minimizes total travel time, charging time, and waiting time:
\[
\min \sum_{k \in K} \left( \sum_{(i,j)\in A,\, i\neq j} dt_{ij}x_{ij}^k + \sum_{i \in S \cup S'} (ct_i + wt_i) z_i^k + \sum_{i \in M \cup M'} wt_i q_i^k \right).
\]
The model remains graph-based, but its charging points include both static stations and mobile route resources [1710.03726].

A distinct route-based interpretation appears in operator-controlled networks. In [1903.06388], each service option \((i,j,\ell)\) is associated with routing probabilities over feasible stations on the user’s path, and expected delay is
\[
W_{i,j,\ell} = \sum_{q=1}^{Q} (d_q + \varrho_q) r^q_{i,j,\ell}.
\]
Users do not directly select stations; they choose differentiated service options, and the CNO manages route-compatible assignment to shape wait times, grid costs, and station utilization. In this sense, routing is not in-motion charging, but access to charging is still route-based because station assignment is path-constrained and centrally controlled [1903.06388].

A common misconception is that route-based charging must mean charging while moving. The literature also uses the term for charging decisions embedded in route or schedule feasibility. For electric modular autonomous units, charging occurs at depots or selected intermediate fast-charging stations, yet the charging location and charging duration are encoded in a time-space-SoC path \(p=\{n_p^1,\dots,n_p^m\}\) [2504.04408]. For metropolitan taxi fleets, charging detours and charging times are decision variables between fixed customer trips, so infrastructure siting is station-based while operational feasibility is intraroute [2209.07305].

## 4. Hybrid formulations: joint station choice, route choice, and system coupling

Several papers reject a clean separation between station-based and route-based charging because the two decisions are structurally coupled. In "Generalized Wardrop Equilibrium for Charging Station Selection and Route Choice of Electric Vehicles in Joint Power Distribution and Transportation Networks," each EV chooses both a road vector \(r_i=[r_i^e]\) and a charging-station vector \(t_i=[t_i^d]\), subject to local flow-conservation constraints that couple destination choice and path choice [2211.05635]. Road congestion enters through a BPR-type latency \(l_e(\sigma_e(r^e))\), station congestion enters through a utilization-based surcharge \(\psi_d(t^d)\), and charging prices are determined by DLMPs from the DSO’s OPF. The equilibrium concept is a variational generalized Wardrop equilibrium, so route congestion, station crowding, road tolls, station surcharges, and distribution-network prices are solved as one coupled game [2211.05635].

Adaptive trip planning models make the same coupling computational rather than equilibrium-theoretic. In [2102.06503], the total trip criterion is
\[
T_{\text{total}} = T_{\text{drive}} + T_{\text{wait}} + T_{\text{charge}},
\]
and station selection is integrated into route planning through a dynamic graph whose nodes are the origin, destination, and reachable charging stations. Edge weights are built from Pareto-optimal time-energy paths, expected waiting time at the next station, and charging time at the current and next station. The Central Charging Station Database stores current occupancy, planned charge stops, and historical utilization data, so route and station decisions are recalculated from the same information structure [2102.06503].

A related trilevel structure appears in commuting and hub charging. In [2012.06392], a commuter chooses a path to a Park-and-Ride hub and then either charges at the hub during working hours or charges later elsewhere. The lower-level congestion game determines route and hub choice, the Charging Service Operator sets charging prices at the hubs, and the Electrical Network Operator sets supply conditions. Charging is therefore station-based in execution but route-coupled in equilibrium [2012.06392].

These hybrid models are significant because they show why isolated station optimization can be misleading. If route detours, time windows, road congestion, or power-network constraints are ignored, the optimal station may cease to be optimal once the transportation and electrical layers are coupled.

## 5. Operational control, information systems, and pricing mechanisms

One major line of research improves charging performance by changing information and incentives rather than physical charger count alone. In [2403.06223], sharing estimated wait time at arrival reduces later reneging and stabilizes the queue; the paper reports reneging reduction up to \(94\%\) and throughput improvement up to \(5\%\) under the informed two-port design. In [1605.00434], Charging Stations publish location, instantaneous queuing time, supply price, and charging capability through a Publish/Subscribe framework with Road Side Units. Pull Mode caches the latest CS information at RSUs, and Advanced Pull Mode adds anonymous remote reservation information \((T^{arr}_{ev},T^{cha}_{ev})\), so EVs select the station with minimum expected waiting time rather than only current queue length [1605.00434].

Long-distance routing work reaches similar conclusions from a route-planning perspective. The CSDB framework in [2102.06503] reduces average waiting time at charging stations from \(6{:}03\) hours without CSDB to \(11\) minutes with \(100\%\) CSDB adoption, a reduction of about \(97\%\). Even partial adoption substantially lowers waiting time because demand is spread more evenly across stations [2102.06503].

Pricing mechanisms operate at several layers. Menu-based pricing in [1609.09037] posts contracts over energy and deadline, with a social-welfare pricing rule
\[
p_{k,l,t}=v_{l,t}-v_{-k},
\]
and a fixed-profit family
\[
p_{k,l,t}=v_{l,t}-v_{-k}+\beta.
\]
The paper reports reduced peak-demand and more efficient use of limited charging spots. The V2G extension in [1612.00106] adds battery utilization \(BU\), so the station can compensate users for additional cycling and exploit V2G when renewable energy is scarce. Online reservation pricing in [2410.05538] uses an MDP with state \(s=(c,t,p)\) and Bellman recursion \(V^*(s)=\max_{a\in A}\sum_{s'}T(s'|s,a)[R(s,a,s')+V^*(s')]\), while a Monte-Carlo tree search heuristic scales to larger station instances. Finally, [2604.15227] develops copositive marginal pricing for binary station access: payment decomposes into energy charge, capacity shadow-price charge, and station-access congestion charge, and the mechanism is proved revenue-adequate, with strong individual rationality under strong duality for the CPP.

A plausible implication is that station-based charging has become as much an information-design and mechanism-design problem as an electrical one. Queue observability, reservation granularity, contract menus, access prices, and route-aware information dissemination are treated as first-class control variables rather than secondary interfaces.

## 6. Performance trade-offs, infrastructure choices, and comparative findings

Comparative results across the literature show that charging architecture changes both operational feasibility and infrastructure economics. In [2509.07402], wireless coverage scenarios of \(20\%\), \(40\%\), and \(60\%\) produce average improvements of about \(8.84\% \pm 5.26\%\), \(15.35\% \pm 10.87\%\), and \(20.36\% \pm 12.54\%\), respectively, with secondary-objective improvements ranging from roughly \(0.7\%\) to \(35.9\%\). The paper states that \(20\%\) coverage already provides immediate benefits, while \(60\%\) coverage is the best-performing level across all benchmark instances [2509.07402]. In [1710.03726], SCS+MED is about \(2\times\), \(3\times\), and \(4\times\) better in travel time than SCS only under low, moderate, and high demand, respectively, and a \(10\)-minute dynamic charge provides about \(3\)–\(8\) kWh or \(9\)–\(23\) miles [1710.03726].

Station-based innovations also produce measurable gains. The two-mode, two-port design in [2403.06223] increases fast-charger availability by \(5\%\), improves throughput by up to \(14.28\%\) during high demand and \(15.7\%\) during low demand, and targets the fact that charging speed decreases significantly beyond \(80\%\) SoC. The ECT-Hub results in [2508.15833] show that ECT-Price achieves higher reward than OR, IPS, and DR baselines across discount levels, and that Incentive Charge cases tend to appear more at night, especially in the \(18{:}00\)–\(24{:}00\) period.

Infrastructure planning papers emphasize that charging-mode choice is inseparable from fleet design and robustness. For electric modular autonomous units, experiments show that charging at both depots and en-route fast-charging stations is necessary during operations, and in the full-day instance \(H_1\) both potential fast-charging stations are selected with the maximum allowed number of posts [2504.04408]. For metropolitan taxi fleets, increasing battery capacities has a more favorable impact on vehicle feasibility of up to \(10\) percentage points compared to increasing charging speeds, allowing for depot charging dominates both, and allowing just \(1\%\) of operational infeasibility risk lowers infrastructure costs by \(20\%\) [2209.07305]. In dockless electric micromobility, station-based charging with rider drop-off promotions yields up to \(28.8\%\) system-wide cost savings versus the better benchmark, while depot-only charging is more sensitive to truck cost and region size [2403.16029].

The literature does not identify a universal dominant option. Station-based charging is discrete, observable, and compatible with queueing control, reservation, local storage, and explicit access pricing. Route-based charging reduces detours and downtime, but its effectiveness depends on corridor coverage, schedule matching, or mobile energy availability. Hybrid designs often emerge when operational realism is introduced: depot charging plus en-route fast charging, fixed stations plus route-aware assignment, or static stations combined with arc-level wireless gain [2504.04408], [1903.06388], [2509.07402].

Source: https://www.emergentmind.com/topics/station-based-or-route-based-charging-options