arXiv subjects

Jakob Puchinger

Publications and source records attributed to Jakob Puchinger.

14 recordsLinked to original sources

arXiv2026

Improving critical buildings energy resilience via shared autonomous electric vehicles -- A sequential optimization framework

The interdependence between electric power systems and transportation systems is rapidly increasing due to the high adoption of Electric Vehicles (EVs) and their charging infrastructures. Electric vehicles can represent additional load for the power system, but can also bring new opportunities for contributing to the efficient and resilient operations of the power grid. This is mainly because of their ability to provide back power to the system when it is not used for transportation, essentially serving as a moving battery source for the power grid. This so-called Vehicle-to-Grid (V2G), Vehicle-to-Building (V2B), or, generally, Vehicle-to-X (V2X) capability of EVs has been extensively studied in the literature. However, the upcoming development of autonomous driving systems and their integration within sharing mobility services can significantly add to the possibilities of interaction between the transportation system and the power grid. This paradigm is studied to a much lesser extent in the existing literature. Shared Autonomous Electric Vehicles (SAEVs) could allow for more control of the actions of the fleet allowing for large-scale coordinated responses both to mobility and energy demands. This coordinated response can be particularly useful for providing emergency power services in case of power loss while maintaining a high level of transportation service. Thus, improving the overall resilience of the system. In this work, we develop a dynamic optimization framework to evaluate the potential contribution of the SAEV fleet for improving critical buildings' energy resilience via V2B services. The model considers passengers' pick-up and transportation, relocation of vehicles, and battery charging and discharging. Power outage scenarios for critical buildings are considered and the potential of the SAEV fleet to fully or partially respond to the emergency power outage is studied. In addition, sensitivity analysis for key parameters such as the outage parameters is introduced. The results of the case study for the Ile-de-France region in France shows that the SAEV fleet has the potential to provide V2B service for critical building at an acceptable loss of passenger total waiting time on the transportation side. Furthermore, it is shown that it is capable of satisfying the emergency power load at a lower cost compared to investing in extra backup generators unless the outage occurs at significantly high frequencie

math.OC↗

arXiv2026

New electric vehicle charging rate design : an MPEC assessment

A high penetration of electric vehicles (EVs) will deeply impact the management of electric power systems. To avoid costly grid reinforcements and the risk of load curtailment due to EV charging, indirect load control via adapted economic signals is a solution proposed by many utilities. Charging costs can be reduced with a domestic tariff applied only to EV charging using a dedicated load measurement method while enhancing the flexibility offered by EVs. We develop a game-theoretical model expressed and treated as a mathematical programme with equilibrium constraints (MPEC) to capture the interaction between a national regulatory authority (NRA) designing these tariffs and heterogeneous agents. First, we analyse the conditions in which EV-only tariffs can be applied for domestic slow charging sessions by comparing different energy profiles. Second, we study the impact of EV charging on different tariff structures to identify the most efficient way of recovering network costs. Submetering with a pure volumetric tariff can bring yearly gains varying from $64 to $110 compared to a flat rate. This depends on the share of investment in grid reinforcement that remains to be made. Finally, we derive policy implications from the results and earmark more sophisticated tariff designs for further investigation.

math.OC↗

arXiv2026

Large Language Model-Driven Full-Component Evolution of Adaptive Large Neighborhood Search

Adaptive Large Neighborhood Search (ALNS) is a prominent metaheuristic and a widely adopted approach for production and logistics optimization. However, it has long relied on hand-crafted components built on expert experience, which makes development slow and costly to adapt to new problems. This paper proposes a closed-loop, large-language-model-driven evolutionary framework that decouples ALNS and automatically rebuilds all of its components. We break ALNS into seven key modules: destroy, repair, operator selection, weight update, initial solution construction, acceptance rule, and destroy-rate control, and evolve each module through a dedicated task. By incorporating the Multi-dimensional Archive of Phenotypic Elites mechanism, the framework maintains a multi-dimensional elite archive to simultaneously drive the evolution of solution quality and strategic diversity. In addition, we design multiple mechanisms, including parallel and sequential multi-module evolution as well as single-expert-driven and multi-expert-driven evolution, to systematically evaluate the impact of different evolutionary paradigms on algorithm generation performance. Evaluations on Traveling Salesman Problem and Capacitated Vehicle Routing Problem benchmarks demonstrate that evolved algorithms consistently outperform optimized classic ALNS baselines under both fixed-iteration and fixed-time limits. The framework also shows a degree of generalizability and cross-problem transferability. Code analysis also uncovers several counterintuitive yet meaningful design patterns that emerged naturally during evolution, offering practical and theoretical insights for future ALNS design. Finally, comparisons across multiple language models highlight clear differences in their ability to support evolutionary algorithm design, helping guide model selection for real-world engineering use.

cs.NE↗

arXiv2026

New Adaptive Mechanism for Large Neighborhood Search using Dual Actor-Critic

Adaptive Large Neighborhood Search (ALNS) is a widely used heuristic method for solving combinatorial optimization problems. ALNS explores the solution space by iteratively using destroy and repair operators with probabilities, which are adjusted by an adaptive mechanism to find optimal solutions. However, the classic ALNS adaptive mechanism does not consider the interaction between destroy and repair operators when selecting them. To overcome this limitation, this study proposes a novel adaptive mechanism. This mechanism enhances the adaptability of the algorithm through a Dual Actor-Critic (DAC) model, which fully considers the fact that the quality of new solutions is jointly determined by the destroy and repair operators. It effectively utilizes the interaction between these operators during the weight adjustment process, greatly improving the adaptability of the ALNS algorithm. In this mechanism, the destroy and repair processes are modeled as independent Markov Decision Processes to guide the selection of operators more accurately. Furthermore, we use Graph Neural Networks to extract key features from problem instances and perform effective aggregation and normalization to enhance the algorithm's transferability to different sizes and characteristics of problems. Through a series of experiments, we demonstrate that the proposed DAC-ALNS algorithm significantly improves solution efficiency and exhibits excellent transferability.

cs.GT↗

arXiv2026

The Dial-a-Ride Problem with Synchronized Visits

The limited capacity of drones and future one- or two-seat modular vehicles requires multiple units to serve a single large customer (i.e., a customer whose demand exceeds a single vehicle's capacity) simultaneously, whereas small customers (i.e., those whose demand can be served by a single vehicle) can be consolidated in one trip. This motivates the Dial-a-Ride Problem with Synchronized Visits, where a fleet of drones must be routed and scheduled to transport orders at minimum cost. We propose four formulations: arc-based, event-based, time-space event-based (TSEF), and time-space fragment-based (TSFrag). An event is defined as a tuple of a location and a set of onboard customers, while a fragment represents a partial path. For TSEF and TSFrag, we also employ the dynamic discretization discovery (DDD) algorithm, which iteratively refines an initial low-resolution time-space network to obtain a continuous-time optimal solution. Computational results show that the event-based formulation performs best under low request intensity (few customers per unit time), whereas TSFrag with DDD excels with high request intensity; both substantially outperform the arc-based formulation. When implemented with DDD, TSFrag also requires less time and fewer iterations than TSEF. We also apply our methods to the classical dial-a-ride problem, where we find that that TSFrag with DDD can replace callbacks in case of high request intensity, and that using DDD is more beneficial to this problem than to the pickup-and-delivery problem with time windows.

econ.TH↗

arXiv2026

New Formulations and Discretization Insights for the Electric Autonomous Dial-a-Ride Problem

The Electric Autonomous Dial-a-Ride Problem (E-ADARP) involves routing and scheduling electric autonomous vehicles under battery capacity and partial recharging constraints, aiming to minimize total travel cost and excess ride time. In practice, operational data for time and state-of-charge (SoC) are often available only at a coarse granularity. This raises a natural question: can discretization be exploited to improve computational performance by enabling alternative formulation structures? To investigate this question, we develop three formulations reflecting different levels of discretization. The first is an improved event-based formulation (IEBF) with arc-flow SoC variables for the continuous-parameter E-ADARP, serving as a strengthened baseline. The latter two are fragment-based formulations designed for discretized inputs. The second is a time-space fragment-based formulation with continuous SoC arc-flow variables (TSFFCS), which discretizes time while keeping SoC continuous. The third is a battery-time-space fragment-based formulation (BTSFF), which discretizes both time and SoC. Here, an event denotes a tuple consisting of a location and a set of onboard customers, while a fragment denotes a partial path. Computational results show that IEBF improves upon the existing event-based formulation for the original E-ADARP. Under discretized settings, TSFFCS tends to outperform IEBF, particularly when recharging is frequent and time discretization is relatively coarse, indicating that time discretization can improve computational performance across a wide range of settings. In contrast, BTSFF rarely outperforms TSFFCS unless the number of reachable SoC levels is limited, suggesting that explicit SoC discretization is beneficial only in relatively restricted settings.

eess.SY↗

arXiv2025

The Bi-objective Electric Autonomous Dial-a-Ride Problem

The electric autonomous dial-a-ride problem (E-ADARP) introduces electric, autonomously driving vehicles and their unique requirements into the classic dial-a-ride problem, where people are transported between pickup and drop-off locations. Next to an electric autonomous vehicle fleet, in the literature, a weighted-sum objective function, which combines the classic routing cost-oriented objective with a user-oriented objective function, has usually been considered. The user-oriented objective function minimizes the total excess user ride time. In this work, we treat them as two separate objective functions, which are optimized concurrently. In order to address the resulting bi-objective E-ADARP, we develop a novel exact framework (called fragment-based checker), whose core part is a smart ``select-and-check" algorithm that iteratively constructs feasible solutions using fragments. Several enhancements are proposed to enforce the computational efficiency of the proposed method. In the computational experiments, we evaluate several variants of our checker algorithm by leveraging a previously developed branch-and-price algorithm. We benchmark the checker-based framework against state-of-the-art criterion space frameworks as well as a generalized branch-and-price algorithm. Numerical results on both bi-objective DARP and E-ADARP instances demonstrate the effectiveness of the proposed framework. With our proposed approaches, 21 out of 38 instances are solved optimally, where small-to-medium-sized instances are solved within seconds. On larger-scale instances, especially those requiring high battery end levels are computationally challenging to solve, our approaches provide high-quality approximations of the Pareto frontiers. Efficient solutions with varying energy restrictions are compared and we obtain valuable managerial insights for different kinds of service providers.

math.OC↗

arXiv2025

A review on reinforcement learning methods for mobility on demand systems

Mobility on Demand (MoD) refers to mobility systems that operate on the basis of immediate travel demand. Typically, such a system consists of a fleet of vehicles that can be booked by customers when needed. The operation of these services consists of two main tasks: deciding how vehicles are assigned to requests (vehicle assignment); and deciding where vehicles move (including charging stations) when they are not serving a request (rebalancing). A field of research is emerging around the design of operation strategies for MoD services, and an increasingly popular trend is the use of learning based (most often Reinforcement Learning) approaches. We review, in this work, the literature on algorithms for operation strategies of MoD systems that use approaches based on Reinforcement Learning with a focus on the types of algorithms being used. The novelty of our review stands in three aspects: First, the algorithmic details are discussed and the approaches classified in a unified framework for sequential decision-making. Second, the use cases on which approaches are tested and their features are taken into account. Finally, validation methods that can be found across the literature are discussed. The review aims at advancing the state of the art by identifying similarities and differences between approaches and highlighting current research directions.

cs.MA↗

arXiv2024

A Branch-and-Price Algorithm for the Electric Autonomous Dial-A-Ride Problem

The Electric Autonomous Dial-A-Ride Problem (E-ADARP) consists in scheduling a fleet of electric autonomous vehicles to provide ride-sharing services for customers that specify their origins and destinations. The E-ADARP differs from the classical DARP in two aspects: (i) a weighted-sum objective that minimizes both total travel time and total excess user ride time; (ii) the employment of electric autonomous vehicles and a partial recharging policy. This paper presents a highly-efficient labeling algorithm, which is integrated into Branch-and-Price (B&P) algorithms to solve the E-ADARP. To handle (i), we introduce a fragment-based representation of paths. A novel approach is invoked to abstract fragments to arcs while ensuring excess-user-ride-time optimality. We then construct a new graph that preserves all feasible routes of the original graph by enumerating all feasible fragments, abstracting them to arcs, and connecting them with each other, depots, and recharging stations in a feasible way. On the new graph, partial recharging (ii) is tackled exactly by tailored Resource Extension Functions (REFs). We apply strong dominance rules and constant-time feasibility checks to compute the shortest paths efficiently. These methods construct the first labeling algorithm that can deal with minimizing (excess) user ride time. In the computational experiments, the B&P algorithm achieves optimality in 71 out of 84 instances. Remarkably, among these instances, 50 were solved optimally at the root node without branching. We identify 26 new best solutions, improve 30 previously reported lower bounds, and provide 17 new lower bounds for large-scale instances with up to 8 vehicles and 96 requests. In total 42 new best solutions are generated on previously solved and unsolved instances.

math.OC↗

arXiv2023

A Deterministic Annealing Local Search for the Electric Autonomous Dial-A-Ride Problem

This paper investigates the Electric Autonomous Dial-A-Ride Problem (E-ADARP), which consists in designing a set of minimum-cost routes that accommodates all customer requests for a fleet of Electric Autonomous Vehicles (EAVs). Problem-specific features of the E-ADARP include: (i) the employment of EAVs and a partial recharging policy; (ii) the weighted-sum objective function that minimizes the total travel time and the total excess user ride time. In this work, we propose a Deterministic Annealing (DA) algorithm and provide the first heuristic results for the static E-ADARP. Partial recharging (i) is handled by an exact route evaluation scheme of linear time complexity. To tackle (ii), we propose a new method that allows effective computations of minimum excess user ride time by introducing a fragment-based representation of paths. These two methods compose an exact and efficient optimization of excess user ride time for a generated E-ADARP route. To validate the performance of the DA algorithm, we compare our algorithm results to the best-reported Branch-and-Cut (B\&C) algorithm results on existing instances. Our algorithm provides 25 new best solutions and 45 equal solutions on 84 existing instances. To test the algorithm performance on larger-sized instances, we establish new instances with up to 8 vehicles and 96 requests, and we provide 19 new solutions for these instances. Our final investigation extends the state-of-the-art model and explores the effect of allowing multiple visits to recharging stations. This relaxation can efficiently improve the solution's feasibility and quality.

math.OC↗

arXiv2022

The bi-objective multimodal car-sharing problem

The aim of the bi-objective multimodal car-sharing problem (BiO-MMCP) is to determine the optimal mode of transport assignment for trips and to schedule the routes of available cars and users whilst minimizing cost and maximizing user satisfaction. We investigate the BiO-MMCP from a user-centred point of view. As user satisfaction is a crucial aspect in shared mobility systems, we consider user preferences in a second objective. Users may choose and rank their preferred modes of transport for different times of the day. In this way we account for, e.g., different traffic conditions throughout the planning horizon. We study different variants of the problem. In the base problem, the sequence of tasks a user has to fulfill is fixed in advance and travel times as well as preferences are constant over the planning horizon. In variant 2, time-dependent travel times and preferences are introduced. In variant 3, we examine the challenges when allowing additional routing decisions. Variant 4 integrates variants 2 and 3. For this last variant, we develop a branch-and-cut algorithm which is embedded in two bi-objective frameworks, namely the $ε$-constraint method and a weighting binary search method. Computational experiments show that the branch-and cut algorithm outperforms the MIP formulation and we discuss changing solutions along the Pareto frontier.

cs.AI↗

arXiv2019

Shared Autonomous Vehicle Simulation and Service Design

Today, driverless cars, as a new technology that allows a more accessible, dynamic and intelligent form of Shared Mobility, are expected to revolutionize urban transportation. One of the conceivable mobility services based on driverless cars is shared autonomous vehicles (SAVs). This service could merge cabs, carsharing, and ridesharing systems into a singular transportation mode. However, the success and competitiveness of future SAV services depend on their operational models, which are linked intrinsically to the service configuration and fleet specification. In addition, any change in operational models will result in a different demand. Using a comprehensive framework of SAV simulation in a multi-modal dynamic demand system with integrated SAV user taste variation, this study evaluates the performance of various SAV fleets and vehicle capacities serving travelers across the Rouen Normandie metropolitan area in France. Also, the impact of ridesharing and rebalancing strategies on service performance is investigated.Research results suggest that the performance of SAV is strongly correlated with the fleet size and the strategy of individual or shared rides. Further analysis indicates that for the pricing scheme proposed in this study (i.e., 20% lower for ridesharing scenario), the standard 4-seats car with shared ride remains the best option among all scenarios. The results also underline that enabling vehicle-rebalancing strategies may have an important effect on both user and service-related metrics. The estimated SAV average and maximum driven distance prove the importance of vehicle range and charging station deployment.

cs.MA↗

arXiv2019

Robo-Taxi service fleet sizing: assessing the impact of user trust and willingness-to-use

The first commercial fleets of Robo-Taxis will be on the road soon. Today important efforts are made to anticipate future Robo-Taxi services. Fleet size is one of the key parameters considered in the planning phase of service design and configuration. Based on multi-agent approaches, the fleet size can be explored using dynamic demand response simulations. Time and cost are the most common variables considered in such simulation approaches. However, personal taste variation can affect the demand and consequently the required fleet size. In this paper, we explore the impact of user trust and willingness-to-use on the Robo-Taxi fleet size. This research is based upon simulating the transportation system of the Rouen-Normandie metropolitan area in France using MATSim, a multi-agent activity-based simulator. A local survey is made in order to explore the variation of user trust and their willingness-to-use future Robo-Taxis according to the sociodemographic attributes. Integrating survey data in the model shows the significant importance of traveler trust and willingness-to-use varying the Robo-Taxi use and the required fleet size.

cs.MA↗

arXiv2019

Shared Autonomous Electric Vehicle Service Performance: Assessing the Impact of Charging Infrastructure and Battery Capacity

Shared autonomous vehicles (SAVs) are the next major evolution in urban mobility. This technology has attracted much interest of car manufacturers aiming at playing a role as transportation network companies (TNCs) in order to gain benefits per kilometer and per ride. The majority of future SAVs will most probably be electric. It is therefore important to understand how limited vehicle range and the configuration of charging infrastructure will affect the performance of shared autonomous electric vehicle (SAEV) services. We aim to explore the impacts of charging station placement, charging type (rapid charging, battery swapping) as well as vehicle range onto service efficiency and customer experience in terms of service availability and response time. We perform an agent-based simulation of SAEVs across the Rouen Normandie metropolitan area in France. The simulation process features impact assessment by considering dynamic demand responsive to the network and traffic. Research results suggest that the performance of SAEVs is strongly correlated to the charging infrastructure. Importantly, faster charging infrastructure and optimized placement of charging locations in order to minimize distances between demand hubs and charging stations result in a higher performance. Further analysis indicates the importance of dispersing charging stations across the service area and how this affects service effectiveness. The results also underline that SAEV battery capacity has to be carefully selected to avoid the overlaps between demand and charging peak times. Finally, the simulation results show that by providing battery swapping infrastructure the performance indicators of SAEV service are significantly improved.

cs.MA↗