simulation-guided optimization

This paper systematically examines the structural role and evolutionary trajectory of simulation methods across the statistical lifecycle. Addressing the current fragmentation and conceptual ambiguity in simulation practice, the study introduces, for the first time, a comprehensive functional taxonomy—spanning model specification, diagnostic checking, validation, and inference—and proposes a “simulation-driven” paradigm for statistical practice, prioritizing computational scalability. Methodologically, it integrates Monte Carlo simulation, approximate Bayesian computation (ABC), simulation-based calibration, and posterior predictive checking, implemented via high-performance computing frameworks to enable large-scale empirical analysis. Key contributions are: (1) establishing simulation as foundational statistical infrastructure; (2) providing an actionable roadmap for algorithm design, statistical software development, and pedagogical reform; and (3) advancing a paradigm shift in statistical practice—from model-centric to simulation-augmented inference.

Optimization model developers face significant adoption barriers, low stakeholder trust, and poor communication in real-world domains such as healthcare and logistics. Method: We conducted a qualitative empirical study involving semi-structured interviews with 15 cross-domain practitioners to investigate optimization practices in situ. Contribution/Results: Our analysis reveals a highly iterative, six-stage optimization practice pattern, identifying continuous data processing and sustained stakeholder dialogue as the core mechanisms driving optimization decisions. Moving beyond the traditional “algorithm-centric” paradigm, we propose a novel “data-and-dialogue co-driven” framework that reconceptualizes optimization as a sociotechnical process—not merely a technical one. This framework provides empirically grounded design principles for developing more transparent, interpretable, and human-centered optimization support tools, thereby bridging critical gaps between technical modeling and organizational implementation.

Optimizing complex stochastic systems with prohibitively high sampling costs remains a fundamental challenge. Method: This paper proposes the first large language model (LLM)-integrated, two-stage automated stochastic optimization (SO) framework. In Stage I, an LLM parses system structure, performs causal discovery, and constructs an ensemble of digital twins. In Stage II, a surrogate-assisted, performance-trajectory-driven meta-optimization generates adaptive hybrid optimization policies capable of runtime dynamic evolution. Contribution/Results: The framework innovatively embeds LLMs into both causal inference and the meta-optimization feedback loop, enabling algorithm-level customization. Experiments demonstrate substantial reductions in sample complexity, along with improved optimization efficiency and robustness across diverse stochastic system domains.

This study addresses the challenges of control design in complex industrial processes characterized by multivariable coupled dynamics by proposing an automated control strategy generation framework that integrates large language models (LLMs) with Bayesian optimization. The approach decomposes control design into structured code generation steps, ensuring physical consistency through execution-based validation and feedback-driven repair. It pioneers the automatic synthesis of decentralized PI controller architectures and their tuning environments directly from dynamic process models. Evaluated on a nonlinear gas preheater benchmark, the generated control schemes—subsequently refined via Bayesian optimization—achieve a 26.5% improvement in closed-loop performance and significantly enhance the transient response of pressure loops, thereby demonstrating the method’s effectiveness and novelty.

To address the challenges of algorithm integration, prolonged development cycles, and high engineering deployment costs in multi-objective simulation optimization (MOSO), this paper proposes and implements ParMOO—a parallel multi-objective simulation optimization framework. ParMOO is the first framework to systematically integrate heterogeneous MOSO algorithms with domain-adaptive mechanisms, leveraging surrogate modeling, asynchronous parallel computation, adaptive sampling, and modular problem interfaces to achieve environment independence, scalability, and low-code customization. It supports black-box simulation and optimization under multiple conflicting objectives. Evaluated on two real-world industrial case studies, ParMOO enabled solver customization and deployment within two weeks, improved Pareto front acquisition efficiency by 3–5×, and significantly reduced integration complexity and implementation cost.

This work proposes a novel paradigm that integrates discrete-event simulation with a large language model (Gemini-1.5-Pro) to overcome the limitations of traditional simulation-based optimization, which treats simulators as black boxes and offers little insight into policy failure. By leveraging event-level trajectory replay, the method automatically identifies bottlenecks from low-scoring simulation runs and generates interpretable, traceable, code-level policy revisions in parallel. It pioneers the use of simulation trajectories to guide the LLM in targeted heuristic rule modification, combined with rolling evaluation and an elite retention mechanism for iterative policy improvement. Evaluated on dynamic production and AGV scheduling tasks, the approach achieves an average policy score of 77.51 (out of 100), improving the best run from 62.49 to 78.61, and significantly outperforms MILP, handcrafted rules, and metaheuristic baselines across 100 random seeds and fault perturbations.

This work proposes a novel approach to black-box testing of Functional Mock-up Units (FMUs) by integrating large language models (LLMs) with a human-in-the-loop mechanism. Addressing the inefficiency and poor interpretability of traditional FMU-based dynamic simulation testing—which relies on manually crafted scenarios—the method automatically generates structured Given-When-Then test objectives from FMU interface and functional specifications, and constructs complete test plans comprising input sequences and assertion oracles. Upon simulation execution, the framework produces visualizable logs and statistical evaluation metrics. The approach significantly enhances test design efficiency and result interpretability, facilitates test asset reuse, and demonstrates effectiveness on a lubricating oil cooling system by autonomously generating executable test scenarios and delivering objective-level pass-rate analysis.

This work addresses the limitations of conventional large language models in generating simulation code—namely, their lack of verifiability, reproducibility, and accessibility to non-programmers—by introducing Sketch2DES, a structured multi-stage workflow that translates queuing network diagrams into discrete-event simulation models. The approach integrates multimodal large language models, a reflective validation loop, JSON schema validation, and a deterministic code adapter to enhance transparency, verifiability, and reproducibility while lowering the programming barrier for users. Evaluated on eight queuing networks of varying complexity, the generated simulations exhibit no statistically significant differences from both hand-coded implementations and analytical benchmarks, demonstrating high reliability.

This study addresses the challenge in bioprocess development of simultaneously optimizing performance, constraint satisfaction, and operational robustness—a task traditionally reliant on expert judgment. To this end, the authors propose a human-in-the-loop multi-objective Bayesian optimization framework that explicitly incorporates the probability of constraint satisfaction and robustness under input perturbations into the Pareto optimization objectives. The approach integrates Gaussian process surrogate models, Monte Carlo–based robustness evaluation, and Pareto-guided sampling, complemented by an interactive four-dimensional visualization interface to support dynamic expert decision-making. Demonstrated on an eight-dimensional fed-batch CHO cell culture simulation, the method efficiently identifies high-performing, feasible, and robust operating conditions, substantially improving experimental resource efficiency and enabling more intelligent termination criteria for iterative optimization.