Reliability Simulation Testing and Verification Technologies for Intelligent Systems: Frontiers, Progress, and Challenges

doi:10.16182/j.issn1004731x.joss.25-0554

Abstract

Abstract:

Intelligent systems are extensively deployed in domains such as transportation, energy and water resources, smart healthcare, and aerospace, where their reliability is directly linked to public safety and social stability, thus requiring thorough and scientifically rigorous verification. This article conducted an in-depth exploration of the current state of reliability simulation and verification techniques for intelligent systems. It defined the concept of reliability specific to intelligent systems and identified key challenges they face in areas including mission scenario modeling, characteristic modeling and simulation, evaluation and verification, and simulation platforms. Future development requirements were proposed to guide research toward more trustworthy, precise, and efficient reliability simulation and verification technologies for intelligent systems, thereby facilitating their high-quality development.

Key words: intelligent system, reliability, mission scenario, reliability simulation testing, reliability evaluation, reliability verification

CLC Number:

TP18

Wang Zili, Gao Yuntian, Yang Dezhen, Liu Yeyang, Ren Yi. Reliability Simulation Testing and Verification Technologies for Intelligent Systems: Frontiers, Progress, and Challenges[J]. Journal of System Simulation, 2025, 37(7): 1583-1606.

Figures/Tables 13

Fig. 1

Fig. 2

Table 1

Fig. 3

Fig. 4

Fig. 5

Fig. 6

Fig. 7

Fig. 8

Fig. 9

Table 2

Fig. 10

Table 3

Comparison of core functions of RST&V platform

平台	多模态数据融合	动态场景建模	故障注入与传播分析	实时性与可扩展性	开放性与可定制化
Apollo	高精度地图与传感器融合(毫米级精度)，依赖真实硬件标定	基于规则生成交通流，极端场景依赖人工设计	支持传感器信号丢失、算法逻辑错误注入，对抗攻击漏洞覆盖不足	支持分布式部署，高精度仿真依赖本地GPU算力	开源代码，可自定义模块，架构复杂、调试门槛高
CARLA(Uber)	物理级传感器模型(LiDAR/摄像头)，数据保真度受虚幻引擎限制	动态生成天气、光照、交通流，长尾场景需脚本扩展	通过Python API注入传感器噪声、通信延迟，缺乏系统级故障传播量化	分布式并行仿真(多任务同步)，大规模场景渲染延迟显著	开源虚幻引擎插件，Python API调用灵活
Autoware(开源社区)	模块化传感器接口(ROS/ROS2)，需第三方工具提升融合精度	依赖外部场景工具(如CARLA)，具备多智能体交互建模能力	结合AVP工具包实现局部故障诊断，无系统性故障链分析	单机运行扩展性受限，模块化架构支持功能扩展	ROS节点化设计，生态丰富，硬件性能要求高
Udacity Simulator	基础传感器模拟(低保真)，无异构数据同步机制	预设简单场景(无动态交互)，无极端条件支持	不支持主动故障注入	单线程运行，实时性差，仅支持小规模测试	封闭式 API，图形界面操作，无底层修改权限
TAD Sim	数字孪生与AI交通流仿真，可进行物理级传感器噪声注入	自动生成 “鬼探头”“团雾”等极端场景，有拟人化背景车模型	全链路故障注入(传感器-计算-执行)，可量化故障传播概率(如海况影响)	云端高并发(千级并行)，支持硬件在环(HIL)实时测试	核心功能私有(如云端调度)，部分模块支持API接入

Table 3

References 78

[1]	National Highway Traffic Safety Administration. Summary report: standing general order on crash reporting for level 2 advanced driver assistance systems[R]. US Department of Transport: Washington, DC, USA, 2022.
[2]	DMV. Autonomous Vehicle Collision Reports[EB/OL]. (2025-06-06) [2025-06-13]. .
[3]	中国新闻网. 五问小米SU7"爆燃事故"[EB/OL]. (2025-04-01) [2025-06-13]. .
[4]	新浪网. 联合国报告: 一军用无人机自主攻击人类或为史上首次[EB/OL]. (2021-06-01) [2025-06-13]. .
[5]	TIME. 'A Tragic Mistake.' Botched Drone Strike in Afghanistan Raises Concerns Over Biden's Counterterrorism Strategy[EB/OL]. (2021-09-17) [2025-06-13]. .
[6]	美国之音. 美军超级大黄蜂战机在红海上空遭"友军误击"2飞行员获救[EB/OL]. (2024-12-22) [2025-06-13]. .
[7]	Gitnux. Drone Accident Statistics[EB/OL]. [2025-06-13]. .
[8]	Commission European. Guidelines on the Definition of an Artificial Intelligence System Established by AI Act[EB/OL]. (2025-03-07) [2025-06-13]. .
[9]	Korteling J E H, van de Boer-Visschedijk G C, Blankendaal R A M, et al. Human- Versus Artificial Intelligence[J]. Frontiers in Artificial Intelligence, 2021, 4: 622364.
[10]	Jungherr Andreas. Artificial Intelligence and Democracy: A Conceptual Framework[J]. Social Media + Society, 2023, 2023(7): 20563051231186353.
[11]	国家市场监督管理总局, 国家标准化管理委员会. 汽车驾驶自动化分级: [S]. 北京: 中国标准出版社, 2021: 3-7.
	State Administration for Market Regulation, National Standardization Administration. Road Vehicles-diagnostic Communication Over Controller Area Network (DoCAN)-dictionary: [S]. Beijing: Standards Press of China, 2021: 3-7.
[12]	中国民用航空局. 民用无人驾驶航空器系统分布式操作运行等级划分: [S]. 北京: 中国民航出版社, 2022: 2-4.
	Civil Aviation Administration of China. Grading of Distributed Operation Level for Civil Unmanned Aircraft System: [S]. Beijing: Civil Aviation of China Publishing House, 2022: 2-4.
[13]	International SAE. Taxonomy and Definitions for Terms Related to Driving Automation Systems for On-Road Motor Vehicles: J3016_202104 [S]. Warrendale: SAE International, 2021: 17-26.
[14]	Afonin V L, Gavrilina L V, Yakovlev M G. An Expert System for an Intelligent Technological Complex for Complicated Surface Processing[J]. Journal of Machinery Manufacture and Reliability, 2022, 51(S1): S109-S119.
[15]	Aksjonov Andrei, Kyrki Ville. A Safety-critical Decision Making and Control Framework Combining Machine Learning and Rule-based Algorithms[EB/OL]. (2022-01-30) [2025-06-12]. .
[16]	Liu Zhijian, Amini A, Zhu Sibo, et al. Efficient and Robust LiDAR-based End-to-end Navigation[C]//2021 IEEE International Conference on Robotics and Automation (ICRA). Piscataway: IEEE, 2021: 13247-13254.
[17]	Khan Murad, Seo Junho, Kim Dongkyun. Modeling of Intelligent Sensor Duty Cycling for Smart Home Automation[J]. IEEE Transactions on Automation Science and Engineering, 2022, 19(3): 2412-2421.
[18]	He Wen, Chen Yong, Liu Tao, et al. Gaussian Pseudo-spectrum Optimization-based Fuzzy Logic Parallel Parking Trajectory Planning[J]. International Journal of Automotive Technology, 2025, 26(1): 115-128.
[19]	Liang Chen, Chen Zhenwu, Yang Liang, et al. Physical Architecture Simulation Based on System Dynamics Modelling for an Autonomous Transportation System Scenario[J]. Journal of Advanced Transportation, 2023, 2023(1): 9390468.
[20]	Su Yuanbo, Xing Lü, Li Shukai, et al. Self-adaptive Equation Embedded Neural Networks for Traffic Flow State Estimation with Sparse Data[J]. Physics of Fluids, 2024, 36(10): 104127.
[21]	Zhao Lu, Farhi Nadir, Christoforou Zoi, et al. Analysis of Driver Behavior and Intervehicular Collision: A Data-based Traffic Modeling and Simulation Approach[J]. Journal of Advanced Transportation, 2022, 2022(1): 1068311.
[22]	Jiang Yi. Intelligent Building Construction Management Based on BIM Digital Twin[J]. Computational Intelligence and Neuroscience, 2021, 2021(1): 4979249.
[23]	Zhang Hong, Zheng Zan, Yu Hailiang, et al. Analysis of Intelligent and Connected Vehicles Driving System Modeling[J]. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, 2025, 239(5): 1807-1819.
[24]	Jia Lulu, Yang Dezhen, Ren Yi, et al. A Dynamic Test Scenario Generation Method for Autonomous Vehicles Based on Conditional Generative Adversarial Imitation Learning[J]. Accident Analysis & Prevention, 2024, 194: 107279.
[25]	Wang Yunbo, Long Mingsheng, Wang Jianmin, et al. PredRNN: Recurrent Neural Networks for Predictive Learning Using Spatiotemporal LSTMs[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems. Red Hook: Curran Associates Inc., 2017: 879-888.
[26]	Chen Jialiang, Li Huizhe, Hu Zhaozheng, et al. Evaluation Index for IVIS Integration Test Under a Closed Condition Based on the Analytic Hierarchy Process[J]. Electronics, 2022, 11(22): 3830.
[27]	Wang Yiyun, Yu Rongjie, Qiu Shuhan, et al. Safety Performance Boundary Identification of Highly Automated Vehicles: A Surrogate Model-based Gradient Descent Searching Approach[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(12): 23809-23820.
[28]	Zhou Rui, Liu Yuping, Zhang Kai, et al. Genetic Algorithm-based Challenging Scenarios Generation for Autonomous Vehicle Testing[J]. IEEE Journal of Radio Frequency Identification, 2022, 6: 928-933.
[29]	Wang Pengyu, Wang Guangyu, Li Wenbo, et al. Research on Traffic Flow Generation Model Based on Improved ACGAN and Its Evaluation System[J]. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering. (2025-03-01) [2025-06-11]. .
[30]	Yang Dezhen, Liu Yeyang, Ren Yi, et al. A High-dimensional Dynamic Characteristics-based Boundary Test Scenario Generation Method for UAVs in Dynamic Target Tracking[J]. IEEE Systems Journal, 2025, 19(2): 553-564.
[31]	Zheng Chengwei, Lin Wenbin, Xu Feng. EditableNeRF: Editing Topologically Varying Neural Radiance Fields by Key Points[C]//2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway: IEEE, 2023: 8317-8327.
[32]	Wang Han, Liu Min, Shen Weiming. Industrial-generative Pre-trained Transformer for Intelligent Manufacturing Systems[J]. IET Collaborative Intelligent Manufacturing, 2023, 5(2): e12078.
[33]	Yang Lan, Yuan Jiaqi, Zhao Xiangmo, et al. SceGAN: A Method for Generating Autonomous Vehicle Cut-in Scenarios on Highways Based on Deep Learning[J]. Journal of Intelligent and Connected Vehicles, 2023, 6(4): 264-274.
[34]	Wang Xumeng, Xue Xiao, Yan Ran, et al. Sora for Intelligent Vehicles: A Step from Constraint-based Simulation to Artificiofactual Experiments Through Dynamic Visualization[J]. IEEE Transactions on Intelligent Vehicles, 2024, 9(3): 4249-4253.
[35]	王若萱, 吴建平, 徐辉. 自动驾驶汽车感知系统仿真的研究及应用综述[J]. 系统仿真学报, 2022, 34(12): 2507-2521.
	Wang Ruoxuan, Wu Jianping, Xu Hui. Overview of Research and Application on Autonomous Vehicle Oriented Perception System Simulation[J]. Journal of System Simulation, 2022, 34(12): 2507-2521.
[36]	Li Xin, Deng Weiwen, Zhang Sumin, et al. Research on Millimeter Wave Radar Simulation Model for Intelligent Vehicle[J]. International Journal of Automotive Technology, 2020, 21(2): 275-284.
[37]	Farag Wael. Real-time NMPC Path Tracker for Autonomous Vehicles[J]. Asian Journal of Control, 2021, 23(4): 1952-1965.
[38]	Gregory J, Rokonuzzaman M, Mohajer N, et al. Data-driven Vehicle Dynamic Model for Autonomous Vehicle Applications[C]//2023 IEEE International Conference on Systems, Man, and Cybernetics (SMC). Piscataway: IEEE, 2023: 2850-2855.
[39]	Mauro Da Lio, Bortoluzzi Daniele, Gastone Pietro Rosati Papini. Modelling Longitudinal Vehicle Dynamics with Neural Networks[J]. Vehicle System Dynamics, 2020, 58(11): 1675-1693.
[40]	Pan Yongjun, Nie Xiaobo, Li Zhixiong, et al. Data-driven Vehicle Modeling of Longitudinal Dynamics Based on a Multibody Model and Deep Neural Networks[J]. Measurement, 2021, 180: 109541.
[41]	Zhou Zhisong, Wang Yafei, Zhou Guofeng, et al. Vehicle Lateral Dynamics-inspired Hybrid Model Using Neural Network for Parameter Identification and Error Characterization[J]. IEEE Transactions on Vehicular Technology, 2024, 73(11): 16173-16186.
[42]	Gao Jiyang, Sun Chen, Zhao Hang, et al. VectorNet: Encoding HD Maps and Agent Dynamics from Vectorized Representation[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway: IEEE, 2020: 11522-11530.
[43]	dSPACE. ADASIS[EB/OL]. [2025-06-09]. ,future%20standards%20for%20access%20to%20predictive%20road%20data.
[44]	Dai Yuqi, Sun Jian, Li Shengbo, et al. Hierarchical and Decoupled BEV Perception Learning Framework for Autonomous Driving[EB/OL]. (2024-07-25) [2025-06-10]. .
[45]	Fu Haoyu, Zhang Diankun, Zhao Zongchuang, et al. Orion: A Holistic End-to-end Autonomous Driving Framework by Vision-language Instructed Action Generation[EB/OL]. (2025-03-25) [2025-06-10]. .
[46]	Jia Xiaosong, You Junqi, Zhang Zhiyuan, et al. DriveTransformer: Unified Transformer for Scalable End-to-end Autonomous Driving[EB/OL]. (2025-03-07) [2025-06-11]. .
[47]	Gao Hao, Chen Shaoyu, Jiang Bo, et al. RAD: Training an End-to-end Driving Policy via Large-scale 3DGS-based Reinforcement Learning[EB/OL]. (2025-02-18) [2025-06-11]. .
[48]	Zhang Yang, Yang Shixin, Bai Chenjia, et al. Towards Efficient LLM Grounding for Embodied Multi-agent Collaboration[EB/OL]. (2024-05-26) [2025-06-12]. .
[49]	贾露露. 融合认知和行为的无人车自主系统可靠性仿真评估方法[D]. 北京: 北京航空航天大学, 2025.
	Jia Lulu. A Reliability Simulation Evaluation Method of Autonomous Vehicles Integrating Cognition and Behavior[D]. Beijing: Beihang University, 2025.
[50]	Jha S, Tsai T, Hari S, et al. Kayotee: A Fault Injection-based System to Assess the Safety and Reliability of Autonomous Vehicles to Faults and Errors[EB/OL]. (2019-07-01) [2025-06-12]. .
[51]	Jha S, Banerjee S, Tsai T, et al. ML-based Fault Injection for Autonomous Vehicles: A Case for Bayesian Fault Injection[C]//2019 49th Annual IEEE/IFIP International Conference on Dependable Systems and Networks (DSN). Piscataway: IEEE, 2019: 112-124.
[52]	Zhou Yu, Chen He, Ren Hongping, et al. Event-driven Fault Injection Testing Framework[C]//2024 4th International Conference on Intelligent Technology and Embedded Systems (ICITES). Piscataway: IEEE, 2024: 40-46.
[53]	Liu Ting, Fu Yuzhuo, Xu Xiaotong, et al. A Cross-layer Fault Propagation Analysis Method for Edge Intelligence Systems Deployed with DNNs[J]. Journal of Systems Architecture, 2021, 116: 102057.
[54]	Saeed Awadh Bin-Nashwan, Li Zhanbiao, Jiang Haichang, et al. Does AI Adoption Redefine Financial Reporting Accuracy, Auditing Efficiency, and Information Asymmetry? An Integrated Model of TOE-TAM-RDT and Big Data Governance[J]. Computers in Human Behavior Reports, 2025, 17: 100572.
[55]	Gao Maosheng, Yu Juan, Kamel S, et al. An Intelligent Operational Reliability Assessment Approach Considering Sample Imbalance[C]//2023 IEEE 7th Conference on Energy Internet and Energy System Integration (EI2). Piscataway: IEEE, 2023: 4729-4734.
[56]	Ribeiro M T, Singh S, Guestrin C. "Why Should I Trust You?": Explaining the Predictions of Any Classifier[C]//Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining. New York: ACM, 2016: 1135-1144.
[57]	Pradeep Aneesh, Bakoev Mironshokh, Akhroljonova Nazokat. A Reliability Analysis of Self-driving Vehicles: Evaluating the Safety and Performance of Autonomous Driving Systems[C]//2023 15th International Conference on Electronics, Computers and Artificial Intelligence (ECAI). Piscataway: IEEE, 2023: 1-5.
[58]	Friston K. The Free-energy Principle: A Unified Brain Theory?[J]. Nature Reviews Neuroscience, 2010, 11(2): 127-138.
[59]	Bahl E, Chatterjee S, Mukherjee U, et al. Using Deep Learning to Quantify Neuronal Activation from Single-cell and Spatial Transcriptomic Data[J]. Nature Communications, 2024, 15(1): 779.
[60]	Wan Wenfei, Meng Ya, Shang Bin, et al. Reliability Assessment Scheme for Intelligent Autonomous System[C]//2022 13th International Conference on Reliability, Maintainability, and Safety (ICRMS). Piscataway: IEEE, 2022: 285-289.
[61]	Yue Kun. Multi-sensor Data Fusion for Autonomous Flight of Unmanned Aerial Vehicles in Complex Flight Environments[J]. Drone Systems and Applications, 2024, 12: 1-12.
[62]	Yang Jiachen, Liu Shan, Su Hansong, et al. Driving Assistance System Based on Data Fusion of Multisource Sensors for Autonomous Unmanned Ground Vehicles[J]. Computer Networks, 2021, 192: 108053.
[63]	Toroghi A, Guo W, Abdollah Pour M M, et al. Right for Right Reasons: Large Language Models for Verifiable Commonsense Knowledge Graph Question Answering[C]//Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing. Miami: Association for Computational Linguistics, 2024: 6601-6633.
[64]	Latapie H. Towards A Litmus Test for Common Sense[EB/OL]. (2025-01-17) [2025-06-13]. .
[65]	Dong Liu. Research on Intelligent Level Evaluation System of Intelligent System[C]//2009 Fifth International Conference on Natural Computation. Piscataway: IEEE, 2009: 274-278.
[66]	Min Jie, Hong Yili, King C B. Reliability Analysis of Artificial Intelligence Systems Using Recurrent Events Data from Autonomous Vehicles[J]. Journal of the Royal Statistical Society Series C: Applied Statistics, 2022, 71(4): 987-1013.
[67]	Wang Lizhi, Zhao Xuejiao, Zhang Yuan, et al. Unmanned Aerial Vehicle Swarm Mission Reliability Modeling and Evaluation Method Oriented to Systematic and Networked Mission[J]. Chinese Journal of Aeronautics, 2021, 34(2): 466-478.
[68]	Xing Liudong, Johnson B W. Reliability Theory and Practice for Unmanned Aerial Vehicles[J]. IEEE Internet of Things Journal, 2023, 10(4): 3548-3566.
[69]	Hendrycks D, Mazeika M, Dietterich T. Deep Anomaly Detection with Outlier Exposure[EB/OL]. (2019-01-28) [2025-06-12]. .
[70]	Huang Xinyu, Wang Peng, Cheng Xinjin, et al. The Apolloscape Dataset for Autonomous Driving[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops. 2018: 954-960.
[71]	Luo Xiaolong, Wei Zhiyuan, Zhang Guoqing, et al. High-risk Powered Two-wheelers Scenarios Generation for Autonomous Vehicle Testing Using WGAN[J]. Traffic Injury Prevention, 2025, 26(2): 243-251.
[72]	Tian Haoxiang, Wu Guoquan, Yan Jiren, et al. Generating Critical Test Scenarios for Autonomous Driving Systems via Influential Behavior Patterns[C]//Proceedings of the 37th IEEE/ACM International Conference on Automated Software Engineering. 2022: 1-12.
[73]	Dosovitskiy A, Ros G, Codevilla F, et al. CARLA: An Open Urban Driving Simulator[C]//Proceedings of the 1st Annual Conference on Robot Learning. Chia Laguna Resort: PMLR, 2017: 1-16.
[74]	Pawelczyk M, Bielawski S, Johannes van den Heuvel, et al. CARLA: A Python Library to Benchmark Algorithmic Recourse and Counterfactual Explanation Algorithms[EB/OL]. (2021-08-02) [2025-06-13]. .
[75]	Hossain J. Autonomous Driving with Deep Reinforcement Learning in CARLA Simulation[EB/OL]. (2023-06-20) [2025-06-13]. .
[76]	丁善婷, 王淼, 董正琼, 等. 基于多智能体的舰船装备健康状态仿真评估方法[J]. 中国机械工程, 2022, 33(10): 1169-1177.
	Ding Shanting, Wang Miao, Dong Zhengqiong, et al. A Multi-agent-based Simulation Method for Health State Assessments of Naval Equipment[J]. China Mechanical Engineering, 2022, 33(10): 1169-1177.
[77]	Sinha V, Barman A. Neural Network Modelling, Testing, and Simulation Study of an Autonomous Vehicle Using the Udacity Simulator[C]//2023 4th International Conference on Intelligent Technologies (CONIT). IEEE, 2024: 1-11.
[78]	Niu Yihan, Zhu Feixiang, Wei Moxuan, et al. A Multi-ship Collision Avoidance Algorithm Using Data-driven Multi-agent Deep Reinforcement Learning[J]. Journal of Marine Science and Engineering, 2023, 11(11): 2101.

分级	名称	自动驾驶车辆动作控制	无人机飞行任务执行
0	应急辅助	驾驶员	操作人员
1	部分驾驶辅助	驾驶员与系统	系统+操作人员
2	结合驾驶辅助	系统	系统
3	有条件自动驾驶	系统	系统
4	高度自动驾驶	系统	系统
5	完全自主	系统	系统

指标	内涵	主要应用领域	侧重点
准确率	正确预测与总预测次数的比率	金融领域的信用评分模型；医疗领域的疾病判断模型	评价模型预测能力^[54]
精确率	正确阳性预测与预测阳性的比例	医疗领域的疾病判断模型	在分类任务中评估模型的阳性预测性能
召回率	正确阳性预测与实际阳性的比例	医疗领域的疾病判断模型	识别出所有阳性样本的能力
F1值	精确度与召回率的调和平均值	金融领域，信用评分和欺诈检测模型；舆情领域，文本分析模型的分类效果	综合考虑误报与漏报，提供对模型分类性能的整体评估维度^[55]
鲁棒性	评估在复杂且动态环境下是否维持稳定性能	在自动驾驶领域，需确保在复杂路况和不同天气下仍能稳定运行，避免安全风险	面对输入数据分布变化、噪声干扰或其他外部环境因素时，能够持续输出一致且准确的预测结果^[54]