The COVID-19 pandemic placed tremendous strain on medical resources. To assess intervention strategies and resource planning for metropolitan pandemic response, a hybrid System Dynamics–Discrete-Event Simulation (SD-DES) framework is developed, using Wuhan as a case study. With SD capturing macroscopic infection dynamics and policy intervention and DES simulating medical resource scheduling, bidirectional feedback between macro and micro levels enables a more holistic and comprehensive evaluation of policy effects. By simulating 132 representative scenarios, key factors including containment measures, medical resource capacity, mask adoption, vaccine rollout speed, and the timing of external medical support were examined. Our analysis provides a comprehensive perspective on pandemic prevention and control decision-making, an aspect often underrepresented in prior studies. The model offers an extensible computational framework for complex socioeconomic systems, such as pandemic emergencies, where dynamic and process complexities coexist. It can assist policymakers in enhancing the healthcare system’s preparedness and mitigating the spread of the pandemic. Future research could improve model robustness by incorporating advanced parameter estimation techniques, such as particle filtering. To address computational challenges, we also recommend implementing asynchronous module execution and optimized programming to improve simulation efficiency and scalability.
This paper explores an initial attempt to use the Unity ML-Agents toolkit to model the behavior of people evacuating from indoor fires. The virtual environment was created in the Unity game engine and populated with humanoid agents capable of moving autonomously within the scene. Each agent perceives information from the rendered environment, such as surfaces, directions, and line-of-sight depth and uses it to navigate toward the nearest exit. Agents were trained through reinforcement learning, using the Proximal Policy Optimization (PPO) algorithm to balance rewards and penalties for their actions. We tested five different reward schemes in single-agent simulations to observe how these affect navigation behavior. Among them, the version referred to as mark5 showed the most plausible and efficient evacuation strategy, reaching the exit quickly while avoiding collisions. The same trained agent was then used in multi-agent settings, where its performance remained stable even with groups of up to 20 individuals. These first results suggest that Unity ML-Agents can offer a practical foundation for building more realistic and adaptive evacuation models.
Simulation modeling is widely used to investigate the impact of interventions on infectious disease transmission. However, individual-based models often involve numerous parameters, making exhaustive exploration computationally prohibitive. Metamodels can address this limitation, and when coupled with interpretable machine learning, they can provide insights into intervention impacts. Moreover, multiple features of epidemic curves should be considered when evaluating interventions. Using the open-source agent-based model Covasim, we examined non-pharmaceutical interventions affecting key epidemic curve features, such as peak day, peak value, and attack rate. We employed random forests as a metamodeling technique to capture nonlinear input–output relationships. Then, we used the Boruta algorithm to identify impactful interventions. Results showed that random forest metamodels explained moderate to high variance across outputs. The Boruta method indicated that influential interventions differed by outcome. Social distancing measures, such as school closures and workplace density reductions, were mostly unimportant in the current Covasim setting, whereas intervention timing, testing, contact tracing, and mask use consistently influenced all outputs to varying degrees. Our metamodel- and variable selection-based framework can provide insights into the effectiveness of multiple interventions across various characteristics of the epidemic curve, thereby supporting public health decision-making at the onset of an epidemic.
Aerodynamic drag is a dominant factor affecting the efficiency and driving range of electric trucks, especially at high speeds. Although previous studies have mainly focused on external drag-reduction devices, internal airflow control strategies such as integrated ducts have received limited attention. This study addresses that gap by analyzing how placing an internal duct in the truck body affects drag and power efficiency. A simplified two-dimensional computational fluid dynamics (CFD) approach was employed using ANSYS Fluent to evaluate and compare a baseline truck model and a modified configuration featuring a central internal duct. The simulations assessed the drag coefficient, drag force, propulsion power, and energy consumption over a wide range of inlet velocities (50–140 km/h). The results show that the ducted configuration consistently outperforms the baseline model, achieving up to a 7.5% reduction in drag force, 7.53% savings in energy consumption per kilometer, and an 8.2% extension in driving range at 140 km/h. These findings confirm the aerodynamic and energy-saving potential of internal ducting as a passive design strategy for heavy-duty electric vehicles.
In the process of green transformation under the context of the digital economy, stakeholders such as the government, producers, and consumers find it difficult to reach an effective consensus in the game process due to differences in their interest appeals. This study constructs a tripartite stochastic evolutionary game model among the government, producers, and consumers, and analyzes the evolutionary stable strategies and evolutionary processes of the three parties. The research finds that: The existence of stochastic disturbance factors will affect the evolutionary path and slow down the speed at which the three parties evolve to stable strategies. During the strategy evolution process, supply–demand imbalance phenomena will exist. Market product circulation volume has a significant impact on strategy selection, and differences in product circulation volume under different strategy combinations will directly affect the strategy selection process. In terms of digital green transformation input costs, producers are mainly affected by their own transformation costs, while consumers are affected not only by their own transformation costs but also by producers’ transformation investments. The development of the digital economy can reduce the difficulty for producers and consumers to evolve toward digital green strategies, and sensitivity to emission reduction effects is also an important factor influencing strategy selection. Therefore, the government should formulate reasonable subsidy–penalty policy combinations, dynamically regulate market product circulation volume, implement differentiated incentives for different industries and groups, and promote supply–demand synergy toward the evolution of digital green strategies.
This research uses virtual reality (VR) to immerse human subjects in an active school shooting scenario in order to generate ecologically valid models of school shooter movement and behavior. Historically, data recovered from US school shootings has lacked the fidelity needed to model shooter movement; consequently, simulations of school shootings have had to rely on significant and unsupported assumptions about the movements of the shooter and/or victims. We asked human subjects to act as school shooters in a VR simulation. We then recorded their movements, observations, and actions. Our results show that participant shooters are statistically equivalent to historical incidents with respect to aggregate engagement metrics (shot rate, victim rate, and accuracy) across most scenarios. Moreover, empirical models trained on participant data reduced prediction error by at least 15.5% compared to heuristic baselines and 16.7% compared to models trained on pedestrian data. Overall, this work provides a reproducible framework for data-driven modeling of shooter movement, supporting controlled simulation-based evaluation of response strategies.
Reliable, performant code is essential in modeling and simulation, and integration of AI-generated code into simulation software development must not undermine the stability or integrity of models. Generative AI produces variable outputs, resulting in LLM-generated code that does not reliably satisfy standards of correctness or performance. This paper presents an LLM prompt engineering framework, Goal, Performance, Exclusion Architecture (GPE-A), that structures intent, constraints and performance criteria to guide the generation of reliable, performant simulation code. Rather than attempting reproducibility of an implementation, the framework steers generative outputs toward convergence on required behavioral metrics, consistent with requirements-driven development. The framework is evaluated using LLM-generated random number generation code as a representative simulation component across classical and quantum computing. Random number generation is a foundational simulation primitive and domain where quantum methods allow a comparable proof-of-concept implementation against classical implementations. Metric-based comparisons are made to production baselines, assessing correctness, statistical stability, and computational performance. The GPE-A is evaluated across multiple LLMs, through prompt ablation, and against modern prompting methods. Results indicate that structured LLM prompt engineering can increase predictability and quality in AI-generated simulation code, exceeding production-standard baselines for random number generation, and indicate potential extensibility toward emerging computational paradigms.
Computer simulation has become increasingly complex and widely applied across different domains. However, the reproducibility and replicability (R&R) of simulation models remain limited. Despite recent improvements, independent reproduction or replication of simulation experiments is rare in the literature. This paper provides an overview of the state of research on R&R in simulation, highlights recent developments, and discusses key concepts and future perspectives. It first examines how R&R has been viewed, approached, and evaluated and then outlines typical challenges and defining characteristics of R&R. Emerging opportunities are also discussed in light of community-driven practices, artificial intelligence, and quantum computing. Given the significant role of simulation in modern science, this paper argues that R&R studies of simulation are valuable research outputs and should be regarded as an integral and equally important part of scientific progress. R&R should be explicitly addressed and embedded into modelling and simulation practices, and supported by stronger community efforts. Researchers engaging in these efforts face substantial challenges, including those related to recognition and rewards, methodology, and scalability, many of which are under-researched.
This paper proposes a traffic modeling and a simulation methodology to assist the dimensioning of the capacity of advanced wireless systems. Knowingly, such systems accommodate a diversity of services with corresponding different bandwidth and traffic characteristics, which leads to intricate dimensioning solutions. In our methodology, we treat each type of traffic using its proper model and bundle them in a simulation tool. A fundamental contribution of this work is the proposal of a simulation tool that incorporates general traffic models to yield the performance of the various traffic flows requiring different bandwidths when sharing the same pool of resources in a loss system consisting of a link with a limited capacity. Given the generality and complexity of the model, analytical solutions are not available, and simulation has proven to be a beneficial tool for analyzing and dimensioning the system, with surprising results.
This paper explores the combination of agent-based social simulation (ABSS) models. Model combination facilitates the efficient development of more complex models through reuse, enabling a more comprehensive understanding of phenomena and outcomes that individual models cannot provide on their own. Through a narrative literature review of model combination in other simulation paradigms, six different approaches were identified: ensemble techniques, meta-analysis, model merging, models as modules, model integration and model chains . For each approach, examples and relevant literature are presented, and current challenges are identified. To illustrate the different approaches, a number of models of disease spread are then implemented and combined according to each approach. Through this, the paper aims to both provide inspiration to modelers and to identify paths for future research for the combination of ABSS models and model results.
The scheduling of offshore wind farm (OWF) installations is a complex, weather-driven, and resource-constrained problem. Mixed-integer linear programming (MILP) models excel at cost and makespan optimization, but remain difficult to verify and validate beyond feasibility checks. Similarly, Petri nets (PNs) provide behavioral transparency and simulation fidelity, yet lack prescriptive optimality. This article presents frameworks that utilize complementary application strategies to address two key needs, thereby combining the strengths of both approaches. First, an iterative verification–validation framework is employed that derives local PN models from MILP solutions and compares them with knowledge-based PN representations using conformance and reachability checks. Implemented in a simulation environment for the installation of OWF, the framework enables consistent model verification, cross-model validation, and context-aware scheduling. Second, a cascading decision-support framework that selects or blends MILP, heuristic, and PN-based scheduling methods through standardized descriptors and context signals. Numerical experiments demonstrate that optimization minimizes offshore time and cost, but requires high computational efforts. In contrast, Heuristics deliver short-term plans rapidly, while PN-based schedules offer intermediate costs and longer planning horizons, demonstrating that hybrid, context-sensitive orchestration outperforms any single method.
Primary healthcare (PHC) systems face unsustainable pressure from population aging and the burden of chronic diseases, creating critical workflow inefficiencies and affecting patient care. This challenge particularly affects chronic patient follow-up, where case managers are often overwhelmed by excessive workloads and the complexities of care coordination. In this context, discrete-event simulation (DES) is essential for evaluating the impact of new health policies before implementation. However, simulating these processes at scale entails high computational costs, limiting the feasibility of standard DES approaches for exploring large-scale PHC system scenarios. To address this challenge, we propose a parallel discrete event simulation (PDES) framework designed explicitly for PHC systems. Our proposal implements an approximate parallel simulation protocol based on the bulk synchronous parallel (BSP) model, enhancing efficiency by simplifying event ordering and eliminating rollback. Validation was conducted by the assessment of healthcare experts. We show that this approach achieves significant scalability (up to a 6.95-fold speedup) while maintaining high fidelity. More than 90% of the tested parallel configurations produced results identical to sequential models, with all outcomes remaining clinically accurate. This validates the protocol’s use for strategic public health decision-making.
The rapid growth of connected devices and associated data has increased pressure on cloud-edge infrastructures, where cloud servers (CSs) saturate as devices scale. Most existing studies minimize execution time but overlook system scalability, while many simulation tools are unable to model how performance bottlenecks evolve as systems grow. In this work, we propose a server capacity-driven methodology using the VisualSim simulator that improves scalability by reducing the utilization of CSs and edge servers (ESs), enabling additional devices to be supported without infrastructure upgrades. The methodology progresses by saturating the CSs with computations to evaluate their maximum capacity, offloading computations to ESs to assess their limits, and distributing computations across CSs and ESs to achieve additional headroom for computations. Scalability is then measured by integrating new device data into the system for processing. It is evaluated using three heterogeneous systems: System-1, System-2, and System-3 with 20, 30, and 40 devices, and compared against Greedy Earliest Finish Time (EFT), Genetic Algorithm (GA), and Particle Swarm Optimization (PSO). The experimental results show that our method enables System-1, System-2, and System-3 to process 66.67%, 62.50%, and 60% more device data, respectively, while reducing execution time by up to 48.40% and energy consumption by up to 12.50% for System-3. The proposed method consistently outperforms Greedy EFT, GA, and PSO, achieving up to 40% faster execution, 16% energy savings, and 7% lower CS utilization. This methodology offers a generalizable approach for analyzing the scalability of complex systems used in defense, healthcare, and other fields.
As discrete-event systems grow in complexity across domains such as manufacturing, communication, and cyber-physical systems, their modeling and simulation face increasing challenges. Key issues include model reusability, interoperability across heterogeneous simulation platforms, and maintaining consistency of execution semantics. Traditional approaches often rely on simulators bound to specific programming languages, which restricts portability and hinders collaboration across different tools. To overcome these limitations, this paper introduces FPDEVSML, a formal specification language for parallel-DEVS (PDEVS) models. FPDEVSML provides a structured and unambiguous framework for defining system behavior independently of any particular simulator. This language is based on formal foundations in mathematical logic, algebra, and set theory that ensure only well-formed models can be expressed and validated. From a given specification, models are automatically parsed, semantically analyzed, and transformed into a platform-independent representation, which is then used to generate executable code for multiple simulation environments following model-driven engineering (MDE) approach. We demonstrate this approach through the ft4devs (Formal Tool for DEVS) framework, which we developed, providing code generation for both Artis*/C++ and PyPDEVS/Python. Case studies highlight the FPDEVSML language’s expressiveness in modeling parallel and modular systems. The results confirm the feasibility of the approach, bridging the gap between formal modeling and practical simulation for PDEVS modeling formalism, and paving the way for future automatic integration with formal verification.
Drilling is an expensive but necessary part of mining, geothermal and oil operations. The cost of drilling can be reduced using simulation models to optimize parameters and perform fatigue prognoses, expanding the useful life of downhole equipment. However, current state-of-the-art simulation models have limitations in modeling drill string dynamics in the build section and generating stress history for drilling and tripping operations. This paper presents a model with an efficient contact algorithm, reconfigurable stiction/friction submodels, accurate boundary conditions, expanding drill string length as the bit advances and readily available stress history for any component. The model predicts the onset and location of wellbore contact, as well as axial, lateral and torsional vibrations in any well section. Multiple well trajectory simulations are presented, showing responses such as whirling and snaking. Stress histories and predicted fatigue life of components are compared, showing how the model can be used for selecting the best well trajectory and drilling parameters during well planning. These case studies also show that dogleg severity and kick-off point of the deviated section influence the fatigue life of the downhole components.
The deployment of Digital Twins (DTs) for decision-making support is rapidly expanding in different industries. DTs are cross-disciplinary hybrid systems that combine elements from various engineering domains. Their diverse services require hybrid modeling and simulation techniques to replicate complex system behaviors. Therefore, their inherent complexity requires comprehensive verification and validation frameworks. In the state-of-the-art, verification and validation of DTs are mostly narrowed down to simulation models. However, it needs to extend beyond that and encompass a broader scope. Thus, it is crucial to accurately identify the scope of verification and validation that encompasses data integration, model fidelity, and service functionality. The main contribution of this paper is to propose a verification and validation framework that synergistically integrates white-box and black-box testing techniques to systematically verify and validate data, models, and services of DTs throughout the development life cycle. It leverages an agile methodology to enable iterative and continuous development, verification, and validation. It employs white-box testing using formal methods and black-box testing using software testing techniques, respectively, for verification and validation. A case study is presented that utilizes the framework to develop, verify, and validate a DT. Ultimately, a set of defined evaluation metrics is tracked throughout the development life cycle in the case study to demonstrate the effectiveness of the agile V&V framework proposed.
Rapid advancements in the text generation capabilities of large language models (LLMs) have expanded their applicability across various domains, notably in scenario generation. This paper introduces a dynamic framework that utilizes the advanced text generation of LLMs to enhance scenario generation and decision-making in the wargame scenario. Our framework, fine-tuned on domain-specific data, automates the generation of complex and adaptive scenarios, responding to user interactions. It integrates a wargame domain ontology to ensure scenario accuracy and employs event–condition–action (ECA) rules for decision-making to improve the realism and explainability of scenarios. In addition, through direct preference optimization (DPO), the framework continually refines scenarios based on user feedback, thereby ensuring highly sophisticated simulations. This approach not only diminishes the time and resources needed for wargame scenario generation but also significantly boosts the effectiveness of training. Our findings highlight the transformative potential of LLMs in automating scenario generation and decision support for wargame simulations, thereby broadening the applicability of scenario generation research to various fields.
This paper studies production and distribution planning in a multi-level supply chain for industrial gas. The proposed solution uses a simulation-driven, agent-guided biased multi-objective particle swarm optimization (MOPSO) framework combined with discrete-event simulation (DES). This approach balances competing goals like cost, work time, and service performance under uncertain demand and production variability. The model considers mode-dependent production with different operating modes for each facility and jointly examines routing and scheduling choices. Biasing through balancing, routing, distribution, and priority agents influences initialization and search. An adaptive evaluation scheme focuses simulation efforts on promising areas of the Pareto front. The framework was tested on a real-world case in the process industry with realistic uncertain parameters. Compared to an unbiased particle swarm baseline, the proposed method achieves faster results and better Pareto-front quality within a similar simulation budget. This leads to effective production and distribution schedule that enhance operational performance and lowering cost while still meeting service targets. The findings indicate that the biased approach outperforms others in finding the best trade-offs between time, cost, and simulation efficiency, showing its potential to improve decision-making in complex and unpredictable supply chain environments.
