IEEE Transactions on Industrial Cyber-Physical Systems最新文献

{"title":"Self-Adjustable Anti-Attack Resilience Security Constraint Tracking Control for Underactuated Surface Vessels","authors":"Yu Zhang;Kuo Li;Xi Luo;Changchun Hua","doi":"10.1109/TICPS.2026.3722486","DOIUrl":"https://doi.org/10.1109/TICPS.2026.3722486","url":null,"abstract":"This paper considers prescribed performance tracking control for underactuated surface vehicles (USVs) under an actuator deception attack. Unlike the existing methods, which regard the attack as a bounded disturbance or rely on full state information, the attack signal is modeled as the output of a dynamic system driven by tracking error and its delay term. This modeling method only requires the attacker to obtain the output information of the USV, which has strong concealment and enforceability. For this attack, a self-adjustable prescribed performance control framework is proposed. Firstly, a self-adjustable prescribed performance constraint is designed based on the smooth tuning function, which can ensure that the tracking error is strictly constrained in the whole process under any initial conditions. Secondly, by introducing an auxiliary variable and dynamic surface control(DSC) technology, the controller design problem of USV and the computational complexity explosion problem of the backstepping method are solved, respectively. The designed control strategy is nonsingular and does not depend on the unknown parameters and delay information of the attack model. Based on the Lyapunov–Krasovskii method, the uniform boundedness of all signals in the closed-loop system is proved. Simulation results verify the effectiveness of the proposed method.","PeriodicalId":100640,"journal":{"name":"IEEE Transactions on Industrial Cyber-Physical Systems","volume":"4 ","pages":"794-803"},"PeriodicalIF":7.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148836651","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"IED-Level Sparse Recovery for Attack-Resilient PMU Data Reconstruction in Cyber Physical Power Systems","authors":"Khwrwmdao Basumatary;Anup Shukla;Francisco Gonzalez-Longatt","doi":"10.1109/TICPS.2026.3699301","DOIUrl":"https://doi.org/10.1109/TICPS.2026.3699301","url":null,"abstract":"Reliable real-time monitoring and control in industrial Cyber–Physical Power System (CPPS) rely on high-fidelity Phasor Measurement Unit (PMU) data streams; however, recent control-layer attacks—such as <italic>Stuxnet</i> and the <italic>Ukraine grid</i> incidents—have exposed their increasing vulnerability. In CPPS, Intelligent Electronic Devices (IEDs) act as the last line of defense against intrusions and the first responder to contingencies. Therefore, IED-level cybersecurity mechanisms are necessary to avoid false control actions and ensure operational reliability. This article introduces a lightweight IED-level sparse-recovery framework for real-time PMU data reconstruction, formulated as a projection-driven <inline-formula><tex-math>$ell _{p} in lbrace 0,1rbrace$</tex-math></inline-formula> optimization with an embedded Dantzig-style feasibility constraint. The approach integrates offline subspace learning with a fixed-iteration solver to ensure error-bounded reconstruction and is tractable under a strict time budget in an IED. Experimental validation on a realistic cyber-physical testbed under correlated <italic>data missing, data repetition, data resembling, bit-flipping, and MITM</i> attacks—where up to 37% of PMU signals were compromised—showed reconstruction fidelities exceeding 97% and remaining above 82% in the most corrupted cases. The IED-level implementation achieved deterministic real-time operation with a 25 ms task cycle, low % CPU utilisation, and minimal algorithmic memory footprint, confirming its suitability for resilient on-device data integrity in CPPS.","PeriodicalId":100640,"journal":{"name":"IEEE Transactions on Industrial Cyber-Physical Systems","volume":"4 ","pages":"492-503"},"PeriodicalIF":7.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148442211","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Stacked-LSTM Network for FDIA Detection With Convolutional Residual Auto-Encoder in Power Cyber-Physical Systems","authors":"Fengyong Li;Kunzhan Li;Siming Huang;Xiangjing Su","doi":"10.1109/TICPS.2026.3710165","DOIUrl":"https://doi.org/10.1109/TICPS.2026.3710165","url":null,"abstract":"False data injection attacks (FDIA) often exploit vulnerabilities in malicious data detection to manipulate state estimation results in power cyber-physical systems, leading to unstable power system operation. Traditional deep learning-based methods often fail to fully consider the spatial correlation between measurement data and power grid topology, as well as the temporal characteristics of power measurement data, and thus the detection performance always is significantly weakened. To address the aforementioned challenges, an efficient FDIA detection model is proposed, which effectively captures both the temporal characteristics of measurement data and the spatial relationships between power grid topology and measurement data. A key innovation is the feature extraction module based on the stacked-LSTM network, which can efficiently and fully extract the multi-scale temporal features of measurement data. In addition, we design a convolutional attention-based residual auto-encoder to efficiently extract the complex spatial features, which can accurately capture the potential correlation between data, and dynamically aggregate key features and spatial topology information. A series of extensive experiments are performed over the IEEE14, IEEE39 and IEEE118 bus systems, and the experimental results demonstrate that the proposed detection model significantly outperforms the state-of-the-art schemes in terms of accuracy, robustness and complexity, and provides reasonable interpretability from the spatial-temporal dimensions.","PeriodicalId":100640,"journal":{"name":"IEEE Transactions on Industrial Cyber-Physical Systems","volume":"4 ","pages":"593-602"},"PeriodicalIF":7.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148479940","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Accelerated Integral Reinforcement Learning Control of Nonlinear Industrial Cyber-Physical Systems Under Asymmetric Constraints","authors":"Yu-Zhu Xiang;Chen Jiang;Zhengrong Xiang","doi":"10.1109/TICPS.2026.3710164","DOIUrl":"https://doi.org/10.1109/TICPS.2026.3710164","url":null,"abstract":"This paper proposes an accelerated integral reinforcement learning (AIRL) control algorithm for nonlinear industrial cyber-physical systems with unknown drift dynamics under asymmetric constraints. For industrial closed-loop processes subject to asymmetric constraints, a novel nonquadratic function is introduced to explicitly characterize and handle the asymmetry in the control input. Based on the transformed formulation, an online asymmetric-constraint AIRL algorithm is developed to iteratively approximate the optimal control policy directly from system data, without requiring prior knowledge of the drift dynamics. In addition, a relaxation factor is incorporated into the AIRL update law to accelerate convergence. The proposed method provides an online learning-based optimal control framework for constrained nonlinear industrial cyber-physical systems. Simulation results verify its effectiveness and corroborate the theoretical analysis.","PeriodicalId":100640,"journal":{"name":"IEEE Transactions on Industrial Cyber-Physical Systems","volume":"4 ","pages":"625-632"},"PeriodicalIF":7.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148613055","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Distributed Coordination of EV Charging and Discharging in Low-Voltage Networks","authors":"Sajad Koushkbaghi;M. Imran Azim;Alireza Barzega;Hui Song;Nameer Al Khafaf;Mahdi Jalili","doi":"10.1109/TICPS.2026.3694828","DOIUrl":"https://doi.org/10.1109/TICPS.2026.3694828","url":null,"abstract":"This paper proposes a fully distributed electric vehicle coordination framework that achieves optimality, accounts for user mobility requirements, and respects distribution network constraints without relying on a central coordinator. The method leverages peer-to-peer communication among households to achieve coordination and uses dual decomposition to enforce coupling constraints across households while integrating a surplus-based averaging mechanism to relax restrictive assumptions on the communication topology. Furthermore, a distributed voltage estimation strategy is developed that combines local smart meter measurements with the linearized distribution flow model, allowing voltage constraints to be enforced using only local information. The proposed approach ensures privacy, improves grid stability, and enables scalable coordination. Numerical simulations, carried out on a low-voltage Australian radial distribution network, demonstrate the effectiveness of the proposed scheme in mitigating voltage violations under high solar and electric vehicle penetration while meeting user mobility needs.","PeriodicalId":100640,"journal":{"name":"IEEE Transactions on Industrial Cyber-Physical Systems","volume":"4 ","pages":"356-367"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148236109","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Survey of Cyber-Security in Cyber-Physical Systems: Attack Strategies and Protection Mechanisms","authors":"Engang Tian;Zhihua Wu;Yu Huang;Hongtian Chen","doi":"10.1109/TICPS.2026.3696239","DOIUrl":"https://doi.org/10.1109/TICPS.2026.3696239","url":null,"abstract":"Cyber-physical systems (CPSs) are a new generation of engineering systems that deeply integrate embedded computing, communication, and control technology with physical systems. They have the characteristics of real-time capability, adaptability, and intelligence. Therefore, they have been widely applied in fields such as smart healthcare, smart energy, and smart manufacturing, effectively improving the production efficiency and quality of human life. The proliferation of open communication networks within CPSs significantly amplifies their susceptibility to malicious attacks. A successful intrusion by an adversary can lead to severe consequences, underscoring the critical need for robust cyber-security measures. This issue has garnered substantial research interest. The main purpose of this review is to comprehensively summarize the recent progress of cyber-security research in CPSs from the perspectives of both attackers and defenders. First, the principles and mathematical models of classical attacks are presented based on the cyber-physical attack space. Second, various attack strategies are introduced, that is, how attackers exploit less energy and information to achieve maximum destruction while avoiding detection. Third, protection mechanisms such as attack detection, secure estimation, and control are discussed. Finally, the remaining challenges and potential research directions regarding the security problems of CPSs are provided for future research.","PeriodicalId":100640,"journal":{"name":"IEEE Transactions on Industrial Cyber-Physical Systems","volume":"4 ","pages":"392-404"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148236111","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"SAAP-GRID: Context-Aware Digital Assistant for Situational Awareness in Wildfire-Impacted Power Grid Operation Control Center","authors":"Shradha Shrestha;Cosmas I. Nwakanma;Anurag Srivastava","doi":"10.1109/TICPS.2026.3725618","DOIUrl":"https://doi.org/10.1109/TICPS.2026.3725618","url":null,"abstract":"Emergency operations in power grids require rapid interpretation of heterogeneous, time-varying information under dynamic and high-stakes conditions. Large Language Model (LLM) based digital assistants show promise for supporting situational awareness, but evaluating them safely under realistic emergency conditions remains challenging. We present Situationally Aware Assistant for Power Grid (SAAP-GRID), a simulation-grounded and context-aware orchestrator, along with an evaluation framework specifically in wildfire-related emergency scenarios relevant to power-grid operations. The developed framework uses physics-based simulations as controlled proxies for emergency conditions and benchmarks models through grounded question-answering tasks that require reasoning over time-varying system states. We instantiate the framework using a fire simulation engine and evaluate seven LLMs across factual retrieval, computational reasoning, trend analysis, comparative reasoning, temporal reasoning, and robustness tasks. Importantly, wildfire simulation outputs are treated as a dynamic hazard layer rather than as the final analytical target; the objective is to assess whether LLMs can translate evolving hazard conditions into power-grid-relevant situational awareness. We observe consistent gaps across computational, comparative, and temporal queries, indicating limitations in multi-step and spatio-temporal reasoning. While robustness is high for certain models, reliability remains inconsistent under complex scenarios. We also compare two external-grounding strategies, tool calling and Retrieval-Augmented Generation (RAG) to assess their benefits in the studied emergency scenarios. Quantitatively, the best-performing model achieved perfect accuracy on factual, comparative, and temporal retrieval tasks, while computational reasoning remained the most challenging category (80%). External grounding through tool calling matched the strong direct-prompting baseline accuracy (93%) while improving robustness from 92% to 97%, whereas chain-of-thought prompting reduced overall accuracy to 85%, highlighting that structured grounding rather than unstructured reasoning improves LLM reliability in emergency power-grid scenarios. Overall, SAAP-GRID provides a scalable pre-deployment evaluation framework for assessing the performance and robustness of LLM-based assistants in safety-critical power-grid contexts.","PeriodicalId":100640,"journal":{"name":"IEEE Transactions on Industrial Cyber-Physical Systems","volume":"4 ","pages":"833-847"},"PeriodicalIF":7.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148925403","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"New Results on Secure Sliding Mode Control for Stochastic 2-D Cyber-Physical Systems Under Denial-of-Service Attacks With Application to Industrial Thermal Processes","authors":"Ziyi Wang;Heting Zhang;Yang Liu;Wenqiang Ji;Meng Wang","doi":"10.1109/TICPS.2026.3731128","DOIUrl":"https://doi.org/10.1109/TICPS.2026.3731128","url":null,"abstract":"This paper investigates the secure sliding mode control for discrete-time two-dimensional (2-D) Markov jump systems (MJSs) subject to denial-of-service (DoS) attacks, which disrupt signal transmission within communication networks and result in system state information loss. To account for DoS attacks, a suitable sliding surface is firstly designed, and the stochastic stability of the sliding motion is analyzed by constructing a singular Lyapunov function which introduces additional degrees of freedom. Subsequently, a secure sliding mode control law is constructed to guarantee that the system trajectories can be driven into the vicinity of the prescribed sliding surface. The effectiveness of the proposed secure control strategy is substantiated through simulations on an industrial thermal process.","PeriodicalId":100640,"journal":{"name":"IEEE Transactions on Industrial Cyber-Physical Systems","volume":"4 ","pages":"945-954"},"PeriodicalIF":7.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149033409","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimal Sensor Scheduling Against Deception Attacks Under Random Access Protocol","authors":"Wenzhe Tao;Simone Baldi;Di Liu;Xiang Yu;Dong Zhao","doi":"10.1109/TICPS.2026.3739698","DOIUrl":"https://doi.org/10.1109/TICPS.2026.3739698","url":null,"abstract":"State estimation from multiple remote wireless sensors requires scheduling protocols to orchestrate the limited bandwidth and prevent data collisions. A critical aspect overlooked in state-of-the-art protocols is handling the exposure of the wireless channel to cyber attacks. We investigate how, under a random access protocol, optimal sensor scheduling strategies can mitigate the detrimental effects of deception attacks. To this purpose, we quantify the contribution of each sensor to the overall estimation performance and formulate a corresponding scheduling metric. We then model the strategic interaction between the attacker and the defender as a zero-sum game and analytically derive its Nash equilibrium. We further show that the proposed optimal scheduling strategy is flexible enough to incorporate sensor power constraints. Numerical simulations validate the theoretical results and demonstrate the effectiveness of the proposed scheduling strategies in enhancing state estimation resilience against deception attacks.","PeriodicalId":100640,"journal":{"name":"IEEE Transactions on Industrial Cyber-Physical Systems","volume":"4 ","pages":"1101-1110"},"PeriodicalIF":7.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149188341","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"GAIN-Pro2: A Dual Progressive-Transfer Network With Generative Adversarial Imputation for Industrial Soft Sensing","authors":"Yuting Lyu;Maochao Ye;Le Zhou","doi":"10.1109/TICPS.2026.3685583","DOIUrl":"https://doi.org/10.1109/TICPS.2026.3685583","url":null,"abstract":"Deep learning models have been widely applied in industrial soft sensing to predict key quality variables that are difficult to measure directly. However, these models typically assume that multiple variables are uniformly sampled, whereas in practice, data are often collected at different sampling rates due to technical limitations or economic considerations. To address this issue, this paper proposes a novel dual progressive-transfer network with generative adversarial imputation (GAIN-PRO²) for industrial soft sensing. First, a progressive transfer learning strategy is applied across multiple variable chunks to pretrain the generator of GAIN-Pro², which effectively utilizes the information embedded in the multirate data. Second, the pretrained GAIN-Pro² is adversarially trained and applied to perform targeted imputation on the multirate data. Third, the imputed data is rechunked and used to train the soft sensor model via progressive transfer learning again. Finally, the effectiveness of the proposed method is validated through a penicillin fermentation process and a real ammonia synthesis process.","PeriodicalId":100640,"journal":{"name":"IEEE Transactions on Industrial Cyber-Physical Systems","volume":"4 ","pages":"298-305"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147937667","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}