Modern power systems face growing cyber-security threats, such as false data injection attacks (FDIA), introduced by their deepening reliance on information and communication technology, which presents substantial challenges to system security. Mitigating the impact of such low-probability, high-impact extreme events while ensuring economic operation is essential to enhancing grid reliability and resilience. This paper proposes a constrained multi-objective economic-resilience dispatch method for power systems with renewable energy sources (RES) against FDIA, called CMoEcoRes-FDIA. As the first attempt, we consider the total economic cost for FDIA defense and system security operation, the load demand loss rate, and the resilience metric function under FDIA scenarios, as the objective functions. The power balance conditions and security component operational costs are described as constraints. Then, the optimal economic-resilience dispatch problem for power systems with RES against FDIA is formulated as a constrained multi-objective optimization problem, where evolves the variables include the number of security defense components against FDIA, active power outputs and voltage magnitudes of generators, transformer tap ratios, reactive power inputs of compensation devices, and predicted power outputs of wind turbines and photovoltaic plants. The problem is solved by a elaborately-designed constrained multi-objective extremal optimization algorithm (CMOPEO) To validate the effectiveness of the proposed CMoEcoRes-FDIA method, we use IEEE 30-bus system with RES and IEEE 57-bus system with RES as the case studies. Experimental results demonstrate that CMoEcoRes-FDIA achieves satisfactory comprehensive performance on economic cost, load demand loss rate, and resilience metrics under the FDIA scenario.

Large-scale supply chain (SC) disruptions have exposed the limitations of conventional SC-oriented resilience strategies in addressing long-term component shortages. This study proposes a product change strategy with stress testing to enhance SC viability. First, a three-layer supplier-component-supplier network model is constructed to quantify change propagation effects caused by component changes and identify components requiring synchronous changes. Second, a simulation-embedded mixed integer programming stress test model is developed. Third, a two-stage solution approach is proposed, whereby the identification of changed components is prioritised, followed by the optimisation of supply portfolios and production plans. Simulation results illustrate managerial implications. The product change strategy improves customer order fulfillment rates by up to 60.28 percentage points and reduces profit losses by up to 4.88 percentage points under severe disruptions. Increasing the component design margin from 0-0.006 to 0.03-0.05 reduces the number of components requiring synchronous changes from 18 to only 2, thereby lowering the implementation costs by approximately 4.86 percentage points and improving profit levels by 8.89 percentage points. The findings demonstrate that the proposed strategy extends the SC viability theory from concept to SC recovery practices from an integrated product-SC perspective. The methods provide quantifiable decision-making tools and actionable managerial guidance for enterprises.

The dynamic-coding-based sliding mode control (SMC) problem is investigated in this article, where the encoded state is transmitted via Gilbert-Elliott (G-E) high-rate networks. To address the comprehensive challenges of parameter uncertainties, external perturbations, and network resource constraints, an SMC strategy is developed by integrating dynamic encoding and hybrid communication protocols. First, a hybrid scheduling mechanism combining the round-robin protocol (RRP) and the weighted try-once-discard protocol (WTODP) is introduced based on the static/dynamic segmentation features of the FlexRay protocol (FRP), aiming at optimized network resource allocation under high-rate networks. Then, a dynamic encoding-decoding scheme is established in conjunction with the G-E channel model, where quantization errors are mitigated through adaptive parameter adjustments. Furthermore, a multicommunication-scene-aware (MCSA) sliding mode controller is proposed, where the controller gains switch according to the communication scenes influenced by the channel mode, accessed token, and packet case. The closed-loop system under the high-rate case is reconstructed via periodic transmission analysis, whose exponentially ultimately bounded (EUB) stability and sliding domain reachability are ensured in mean square by building corresponding criteria. Finally, the effectiveness of the proposed method is verified through a numerical example.

Frequent plug-in/-out operations result in structural variations of DC microgrids (DCmGs), posing challenges to scalable control and often requiring costly redesigns to maintain stability. To address this issue, this paper proposes a scalable voltage control strategy for uncertain DCmGs, enabling plug-and-play functionality without controller redesign or system reconfiguration. A polytopic uncertain DCmG model is first formulated to simultaneously capture parameter uncertainties in distributed generation units (DGUs), power lines, and ZIP (i.e., impedance, current, and power) loads. A structured free-weight matrix technique is then developed to mitigate the adverse effects of line and load uncertainties on DGUs while yielding a more tractable linear matrix inequality formulation. The proposed scalable control method is implemented locally to ensure the dissipative voltage stability of each DGU, thereby preserving the dissipativity of the entire network. Numerical simulations validate the effectiveness of the proposed strategy in achieving faster convergence and reduced overshoot.

The paper presents a fuzzy control strategy for nonlinear direct current (DC) microgrids, taking into consideration intermittent denial-of-service (DoS) jamming attacks, random false data injection (FDI) deception attacks, event-triggered communication mechanism (ETM), quantiser, noise, and mismatching premises. Firstly, using characteristics of attacks and the microgrid's fuzzy model, a resilient ETM is introduced to reduce the data transmission rate, which also excludes the dropout phenomenon induced by attacks. Second, considering the impacts of hybrid attacks, ETM, noise, quantiser, and mismatching premises simultaneously, a fuzzy switched system model is derived. Third, by applying the piecewise Lyapunov theory, mean-square exponentially stable criteria that satisfy performance are presented, which establish a quantitative relationship between the influencing factors and the stability of the system. In addition, sufficient conditions are presented for designing the switched fuzzy controller with mismatching premises. Finally, the effectiveness of the proposed methods is confirmed.

This article presents a plug-and-play (PnP) distributed control framework for DC microgrids (DCmGs) to address scalability and reconfiguration issues under dynamic network topologies. First, a decentralized control architecture is developed based on the small-gain theorem, enabling PnP operations of distributed generation units (DGUs). Specifically, the plug-in operation only requires local and neighboring state feedback without relying on global information exchange, while the plug-out operation can be performed without interunit communication. In this way, controller deployment is decoupled from the global network topology, thereby reducing reconfiguration overhead and improving the modular scalability and operational flexibility of DCmGs. Second, an optimization-based controller synthesis method is proposed to minimize the coupling effects among DGUs. The proposed method facilitates distributed controller design for heterogeneous DGUs without requiring structured Lyapunov functions or free-weighting matrices, thereby simplifying the synthesis procedure. Theoretical analysis establishes the asymptotic stability and prescribed performance of the closed-loop system under PnP operations. Finally, simulation studies on a six-DGU DCmG prototype are provided to validate the effectiveness of the proposed framework.

This article proposes a secure time-varying formation (TVF) control framework for heterogeneous multi-agent systems (HMASs) operating under the threats of communication eavesdropping, non-cooperative dynamic obstacles, and input saturation. The framework integrates cyber-layer and physical-layer components that collaboratively address these challenging conditions. In the cyber-layer, an intermittent privacy-preserving observer is designed, which fuses periodic activation with a threat-based dynamic triggering strategy. This design effectively mitigates eavesdropping risks while resolving the irreversibility dilemma inherent in static masking techniques. At the physical-layer, the controller is formulated as a quadratic programming problem integrating Lyapunov functions and control barrier functions. It ensures the operational safety of high-order heterogeneous agents in environments with obstacles and input saturation. The effectiveness and advantages of the proposed framework are demonstrated through simulations on a heterogeneous system consisting of uncrewed aerial vehicles (UAVs) and uncrewed ground vehicles (UGVs).

Multi-agent networks are extensively employed in unmanned autonomous systems, such as aerial swarms and ground vehicle formations, owing to their distributed intelligence and autonomous coordination capabilities, which provide a solid theoretical foundation for control system design. However, reliable distributed coordination faces critical challenges due to complex communication environments and system uncertainties. This paper addresses the distributed formation-containment tracking problem for discrete-time heterogeneous multi-agent systems (MASs) with completely unknown nonlinear dynamics under denial-of-service (DoS) attacks. First, a compact-form data-driven linear model is established for each agent. Leveraging relative output measurements, a novel dynamic event-triggered data-driven formation-containment (DET-DDFC) control scheme is proposed, which adaptively adjusts the triggering threshold to minimize communication overhead while guaranteeing closed-loop stability. Furthermore, an aperiodic attack model is introduced, integrated with a DoS-aware mechanism to bolster resilience against communication disruptions. Theoretical analysis guarantees that leaders achieve the desired formation, while followers converge to the convex hull formed by the leaders, even under unpredictable DoS attacks. Simulation results verify the effectiveness and robustness of the proposed approach.

Designing controllers directly from measurement data has attracted growing attention in recent years, as it avoids the need for accurate system modeling or explicit system identification. This paper focuses on recent advances in data-driven control for linear discrete-time systems with unknown system matrices. For noisy input-state data, an in-depth analysis is provided on several representative approaches, including data-driven control based on Willems et al.’s fundamental lemma, quadratic matrix inequalities, linear fractional transformations for combining prior knowledge with data, and integral quadratic constraints. For noisy input-output data, a concise review is presented on control methods based on quadratic matrix inequalities, along with key insights into their structure and implications. The paper concludes by outlining several challenging problems that merit further investigation in future research.

In this paper, a security defense issue is investigated for networked control systems susceptible to stochastic denial of service (DoS) attacks by using the sliding mode control method. To utilize network communication resources more effectively, a novel adaptive event-triggered (AET) mechanism is introduced, whose triggering coefficient can be adaptively adjusted according to the evolution trend of system states. Differing from existing event-triggered (ET) mechanisms, the proposed one demonstrates exceptional relevance and flexibility. It is closely related to attack probability, and its triggering coefficient dynamically adjusts depending on the presence or absence of an attack. To leverage attacker information more effectively, a switching-like sliding mode security controller is designed, which can autonomously select different controller gains based on the sliding function representing the attack situation. Sufficient conditions for the existence of the switching-like sliding mode secure controller are presented to ensure the stochastic stability of the system and the reachability of the sliding surface. Compared with existing time-invariant control strategies within the triggered interval, more resilient defense performance can be expected since the correlation with attack information is established in both the proposed AET scheme and the control strategy. Finally, a simulation example is conducted to verify the effectiveness and feasibility of the proposed security control method.

Complex industrial processes are influenced by various factors dur-ing operation,such as variations in operational conditions and pro-cess drift,which lead to variations in process characteristics and behaviors over time.

We study derivative-dependent control of nonlinear systems with external disturbances, where the output with a time-varying and unknown bias that may be induced by a sensor error is available for the measurements. The derivatives are approximated via finite-difference resulting in a delay-dependent feedback that is dependent on the measurement and the estimation of the unknown bias as well as their past. To address the estimation of the unknown bias, we design a high-order adaptive law through the $\sigma $ -modification method. We next apply the Lyapunov-Krasovskii approach to establish sufficient conditions for input-to-state stability (ISS), which are formulated in terms of linear matrix inequalities (LMIs). Then criterion for designing the gains of the delay-dependent feedback and adaptive law is presented leading to a minimization problem with solvable LMIs rather than the linear constraints. Moreover, we suggest its sampled-data implementation by using the consecutively sampled output and bias's estimation, where the design criterion for finding the gains of the sampled-data feedback and adaptive law is presented. The proposed method is validated by two simulation examples including a practical inverted pendulum on a cart.

This paper proposes a data-driven detection framework for false data injection (FDI) attacks in cyber-physical DC microgrids. First, a novel data-driven characterization framework for cyber-physical DC microgrids is presented, eliminating the need for restrictive assumptions such as the invertibility of stacked data matrices and quasi-stationary linear approximations. Subsequently, a data-driven ${\mathcal L-infinity estimator is designed for real-time monitoring of the states of distributed generation units (DGUs), capable of adapting to noise-corrupted measurements. Meanwhile, a two-stage model-free scalable (TSMFS) attack detection framework is proposed. In the offline stage, historical data are leveraged to construct a set of linear inequalities whose solution yields a residual generator. The online stage generates detection residuals exclusively from received measurements, without relying on coupling information from neighboring DGUs, thus endowing the scheme with inherent scalability. Finally, experiments validate the performance of the data-driven estimator and the effectiveness of the proposed TSMFS detection framework.

This paper proposes a formation control strategy based on noncooperative game framework for unmanned surface vehicles (USVs) in the presence of eavesdroppers. Firstly, the interactions among USVs during formation control are modeled as a noncooperative game. A privacy-protected estimator is then developed for each USV by employing a twin-network structure, so as to effectively estimate the actions of other USVs while preventing information leakage incurred by eavesdropping attacks. Additionally, a novel neural predictor is designed using an accelerated learning-boosted echo state network (ALESN) to approximate unknown system parameters with enhanced convergence rate and prediction accuracy. On the basis of the estimator and predictor, a distributed formation control method is subsequently devised via seeking the Nash equilibrium of the formulated game. The implementation of the control method only requires local information exchanged between neighboring USVs, and thus the desired computational efficiency and scalability can be achieved. A simulation example is finally provided to validate the effectiveness of the reported privacy-protected formation control approach.

Dynamic constrained many-objective optimization problems (DCMaOPs) are prevalent in engineering systems, such as portfolio management, energy scheduling, and mechatronic design. In these scenarios, fluctuated market or operational demands give rise to dynamic changes in many objectives and constraints. Additionally, those many objectives are conflicted and need to be trade-off, resulting in a high-dimensional objective space. Consequently, it poses challenges in collaboratively seeking for and tracking balanced many objectives under limited feasible regions, aggravating diversity loss, premature convergence, and local optimal trapping. To address these issues, we propose a Q-learning driven prediction for dynamic constrained many-objective optimization (QPMaOEA). In QPMaOEA, a Q-learning driven prediction strategy (QP) is tailored to make an estimation of the changing optima. It includes four response actions, re-initialization, elite perturbation, historical reuse, and a hybrid strategy, adapted to various environmental states. Subsequently, a leader-guided correction mechanism is designed to increase the feasibility of the population, based on the constrained two-archive evolutionary algorithm (CTAEA). In the experimental part, a new test suite, DC_DTLZs, is derived from existing ones. QPMaOEA is compared with six other algorithms on both test instances and an application. To make a fair comparison, we embed QP into peer algorithms and conduct further comparisons by replacing QP with two other prediction algorithms. Experimental results demonstrate the superiority of QPMaOEA in convergence, diversity, and feasibility.

This article proposes a semantic-driven integrated try-once-discard (TOD) scheduling protocol for networked control systems (NCSs) with autoencoders and limited bandwidth. A multinode integrated TOD scheduling protocol is first proposed to address transmission inefficiencies due to node-level resource competition, relaxing traditional single-node activation constraints. To support this integrated scheduling, a semantic communication scheme is developed based on an autoencoder structure, which incorporates semantic encoding and decoding processes. Specifically, the raw data from multiple activated nodes are encoded into a semantic representation, with the information subsequently decoded for reconstruction. This collaboration between the scheduling protocol and semantic communication enables simultaneous updates across multiple nodes, optimizing overall transmission efficiency while guaranteeing reliable communication. In fact, due to relaxed transmission accuracy, the decoded information cannot perfectly match the original data, leading to inevitable decoding errors. Accounting for delays, scheduling errors, decoding errors, and bounded disturbances, an input-to-state stability (ISS) analysis is well performed, with a dedicated controller design to ensure the system’s robustness and responsiveness in dynamic environments. Moreover, a gradient-descent-based pre-training algorithm and a parameter refinement algorithm are designed to optimize the semantic encoder-decoder configuration, enhancing the encoder-decoder synergy for continuous-time NCSs. Finally, the effectiveness of the proposed method is verified through an illustrative example.

This paper proposes a novel data-driven point-to-point finite-iteration learning control scheme for unknown nonlinear discrete-time repetitive systems operating under the coupled interference of physical-layer false data injection attacks and channel fading. A unified system framework is first established by integrating a ramp-type false data injection attack model and a channel fading model. To mitigate signal distortion induced by channel fading, an unbiased correction mechanism is introduced to reconstruct reliable data reflecting the physical-layer output corrupted by false data injection attacks. A dual-update strategy is further developed to simultaneously update the control input and parameter estimation. Convergence analysis demonstrates that once the number of iterations exceeds a predefined threshold, the corrected tracking error at specified tracking points will satisfy the prescribed accuracy requirements, thereby achieving the pointto-point finite-iteration tracking objective. Finally, simulation studies validate the effectiveness of the proposed method. The results show that the proposed scheme drives the tracking mean square error below 0.01 within 20 iterations for nonlinear discrete-time systems, outperforming existing methods in convergence speed. For the single-link robotic system, it realizes stable angular velocity tracking within 100 iterations and exhibits strong robustness over the attack coefficient range of 3 to 12.

With the rapid development of renewable energy and high-voltage direct current (HVDC) delivery technologies, the demand for AC/DC power flow calculations incorporating multiple control modes is increasingly growing. However, during the solution process of AC/DC systems, oscillation and divergence issues often arise due to improper handling of DC control mode switching or inappropriate initial condition settings during iterative calculations. To address this challenge, this paper proposes an adaptive initialization method for AC/DC power flow based on control operating region prediction (CORP). This method utilizes control mode switching conditions to delineate control operating regions, transforming the discrete control mode switching problem into the identification and matching of continuous control operating regions, thereby predicting the control mode and feasible solution range in which the system should operate. Based on this, by analyzing the influence of control commands and system strength on control operating regions, an initialization pre-adjustment strategy is provided for subsequent AC/DC solution processes. This approach transforms the traditional initialization process relying on historical data into a guided initialization based on system physical characteristics, avoiding convergence failures caused by control logic mismatch or excessive initial value deviation. Test results based on the IEEE 118-bus system and the CSEE-RAS system incorporating renewable energy demonstrate that the proposed strategy effectively ensures the convergence of AC/DC solutions under multiple control modes.