Table of Content

30 September 2025, Volume 13 Issue 3

Previous Issue   

Preface: Special Issue on Simulation Optimization—A Powerful Tool in Operations Research and Management Science

L. Jeff Hong, Jun Luo, Guang-Xin Jiang, Wei-Wei Fan

2025, 13(3):  685-687.  doi:10.1007/s40305-025-00616-w

Asbtract ( 25 )   PDF  

Related Articles | Metrics

Review of Large-Scale Simulation Optimization

Wei-Wei Fan, L. Jeff Hong, Guang-Xin Jiang, Jun Luo

2025, 13(3):  688-722.  doi:10.1007/s40305-025-00599-8

Asbtract ( 25 )   PDF  

References | Related Articles | Metrics

Large-scale simulation optimization (SO) problems encompass both large-scale ranking-and-selection problems and high-dimensional discrete or continuous SO problems, presenting significant challenges to existing SO theories and algorithms. This paper begins by providing illustrative examples that highlight the differences between large-scale SO problems and those of a more moderate scale. Subsequently, it reviews several widely employed techniques for addressing large-scale SO problems, such as divide-and-conquer, dimension reduction, and gradient-based algorithms. Additionally, the paper examines parallelization techniques leveraging widely accessible parallel computing environments to facilitate the resolution of large-scale SO problems.

Blackbox Simulation Optimization

Hao Cao, Jian-Qiang Hu, Teng Lian

2025, 13(3):  723-749.  doi:10.1007/s40305-024-00549-w

Asbtract ( 18 )   PDF  

References | Related Articles | Metrics

Simulation optimization is a widely used tool in the analysis and optimization of complex stochastic systems. The majority of the previous works on simulation optimization rely heavily on detailed system information, such as the distributions of input variables and the system dynamics. However, it is usually costly and even unrealistic to obtain such detailed information in practice. To overcome this difficulty, new methods are needed to solve stochastic optimization problems based on simulation outputs or real-world data and fairly limited knowledge of the underlying system, i.e., in a blackbox setting. In this paper, we provide an up-to-date overview of the subject of blackbox simulation optimization, which has only been attracting attention from the simulation community in recent years. We discuss some of the new stochastic approximation algorithms developed recently for blackbox stochastic optimization. We also discuss challenges and potential future research opportunities.

Black-Box Rare-Event Simulation for Safety Testing of AI Agents: An Overview

Yuan-Lu Bai, Zhi-Yuan Huang, Henry Lam, Ding Zhao

2025, 13(3):  750-774.  doi:10.1007/s40305-025-00585-0

Asbtract ( 28 )   PDF  

References | Related Articles | Metrics

This paper provides an overview of black-box rare-event simulation methods applicable to the safety testing of artificial intelligence agents. We explore the challenges and efficiency criteria in black-box simulation, especially emphasizing the subtle occurrence and control of underestimation errors. The paper reviews various adaptive methods, such as the cross-entropy method and adaptive multilevel splitting, highlighting both their empirical effectiveness and theoretical limitations. Additionally, it offers a comparative analysis of different confidence interval constructions for crude Monte Carlo methods, aiming to mitigate underestimation errors through effective uncertainty quantification. The paper concludes with a certifiable deep importance sampling approach, using deep neural networks to develop conservative estimators that address underestimation issues.

Simulation Budget Allocation for Improving Scheduling and Routing of Automated Guided Vehicles in Warehouse Management

Gong-Bo Zhang, Hao-Bin Li, Xiao-Tian Liu, Yi-Jie Peng

2025, 13(3):  775-809.  doi:10.1007/s40305-024-00553-0

Asbtract ( 19 )   PDF  

References | Related Articles | Metrics

Simulation budget allocation is a widely used technique for evaluating and optimizing dynamic discrete event stochastic system via efficient sampling. In warehouse management, the scheduling and routing algorithms of automated guided vehicles aim to optimize order assignments and travel paths under certain constraints to achieve specific objectives. However, these algorithms often rely on deterministic optimization methods, neglecting the dynamic and stochastic nature of warehouse management systems. In this work, we propose an efficient method that integrates simulation budget allocation methods with deterministic scheduling and routing algorithms to enhance overall system performance. The proposed method leverages the benefits of both simulation optimization and deterministic optimization techniques, accounting for the inherent uncertainty of the system. We adopt a discrete event simulation model used in the Case Study Competition of the 2022 Winter Simulation Conference, where the objective is to minimize the adjusted average order cycle time (ACT) in a given warehouse simulation scenario. Numerical examples demonstrate that the use of simulation budget allocation methods can significantly further reduce the ACT, thereby highlighting the effectiveness of our proposed method.

Simulation Optimization for Queues with Heavy-Tailed Service Times

Gui-Yu Hong, Xin-Yun Chen

2025, 13(3):  810-836.  doi:10.1007/s40305-024-00562-z

Asbtract ( 20 )   PDF  

References | Related Articles | Metrics

We develop a gradient-based simulation optimization algorithm, dabbed KWiQ-H, for joint pricing and staffing problems in single-server queues with heavy-tailed service time distributions. Our algorithm is designed based on the well-known Kiefer-Wolfowitz algorithm so that it is applicable tomore general and practical settings where customer’s behavior is unknown to service providers in prior. We first establish a convergence result for KWiQ-H when the service times have a finite fifth moment. Then, we show that under a stronger condition with a finite seventh moment, KWiQ-H could achieve sample complexity with the same asymptotic order as in the case when service times are light-tailed in Chen et al. (Oper Res, 2023). Complementing the theoretic results, we carry out comprehensive numerical experiments to test the efficiency and robustness of KWiQ-H in a variety of model settings.

Improved Convergence Rate of Nested Simulation with LSE on Sieve

Ruo-xue Liu, Liang Ding, Wen-jia Wang, Lu Zou

2025, 13(3):  837-872.  doi:10.1007/s40305-025-00609-9

Asbtract ( 18 )   PDF  

References | Related Articles | Metrics

Nested simulation encompasses the estimation of functionals linked to conditional expectations through simulation techniques. In this paper, we treat conditional expectation as a function of the multidimensional conditioning variable and provide asymptotic analyses of general nonparametric least squared estimators on sieve, without imposing specific assumptions on the function’s form. Our study explores scenarios in which the convergence rate surpasses that of the standard Monte Carlo method and the one recently proposed based on kernel ridge regression. We use kernel ridge regression with inducing points and neural networks as examples to illustrate our theorems. Numerical experiments are conducted to support our statements.