In this paper, we are concerned with the tracking control issue for a category of constrained MIMO nonlinear cyber-physical systems (CPSs) with unidentified disturbances, which face the exposure to venomous hybrid cyber attacks. Distinct from most existing related researches which can only address the intrusion-tolerant control issue against single type of cyber attacks, the studied problem in this paper that has not been well resolved is more challenging, where multiple types of network attacks including denial-of-service (DoS) attacks and deception attacks may coexist simultaneously, increasing the design difficulty of intrusion-tolerant controller. Furthermore, by means of designing two error transformation functions (ETFs) and proposing a novel error transformation design technique for resolving the nonlinearity and MIMO coupling problem on the basis of quantifying the destruction mechanism of cyber attacks, a novel adaptive tracking control scheme is ulteriorly developed. As a result, the proposed control architecture owns inherent intrusion-tolerant capability and robustness, which is capable of settling exogenous disturbances and hybrid cyber attacks in unison. In particular, the tracking errors can be forced to convert into the preassigned constrained region at pregiven decay rate in preseted settling time. Finally, the feasibility and superiority of the proposed control framework are illustrated through an emblematic application instance.

This paper studies the robust output feedbackl1control for positive Takagi-Sugeno (T-S) fuzzy systems. A reduced-order dynamic output feedback (RODOF) controller is designed to preserve the positivity and stability of the closed-loop system while achieving anl1-induced performance. Sufficient conditions for satisfying mismatched multiple fuzzy summation inequalities and for the existence of RODOF controllers are first derived, which are formulated as a bilinear programming (BP) problem. To simultaneously address the bilinear and positivity constraints, an additional linear constraint is introduced to establish a linear approximation of the original non-convex constraint, by which we propose a sequential linear programming (SLP) algorithm to determine the desired controller parameters. Furthermore, by constructing an auxiliary stabilization problem and utilizing the convergence property of the SLP method, an initial feasible condition is systematically obtained, thus ensuring constraint satisfaction throughout the iterations. Finally, the feasibility and validity of the theoretical results are demonstrated by a tank system and a buck converter system.

Deep learning (DL) has achieved unprecedented success in precisely diagnosing dermatological lesions. However, increasing concerns over diagnostic unfairness across different demographic subgroups in DL algorithms raise issues of ethical violations and healthcare inequity. Due to limited access to the internal workings of DL algorithms, post-processing methods are widely regarded as efficient techniques for fairness enhancements in DL-based predictions. However, these methods often come at the cost of accuracy, and research exploring their application to medical images remains limited. To address these issues, we proposeFADiaFrame, an innovative post-processing framework designed to enhance both fairness and accuracy in DL-based diagnosis of dermatological lesions. Specifically, our uncertainty-aware gating calibration mechanism inFADiaFrameidentifies and calibrates untrustworthy samples, thereby enhancing diagnostic fairness, accuracy, and trustworthiness. Furthermore, we integrate this mechanism with an optimal transport method to further enhance group fairness. Extensive experiments on real-world datasets demonstrate thatFADiaFrameoutperforms existing post-processing methods in terms of both fairness and accuracy. Notably,FADiaFramepreserves diagnostic accuracy while achieving significant gains in fairness compared to pre-trained baseline models. Among all baselines, MedCLIP, withFADiaFrame, shows the most substantial improvement for the age-sensitive attribute, with accuracy increasing by 3.34% and demographic parity rising by 10.60%. Our results suggest thatFADiaFrameprovides universal applicability across diverse DL models for medical image diagnosis, ensuring both fair and accurate diagnosis across a wide range of devices and deployment contexts.

Accurate preoperative MRI diagnosis of intracranial primary tumors is critical for surgical planning and therapeutic decision-making. This study addresses two fundamental limitations of MRI-based diagnosis: the inherent class imbalance and resolution variations. To address these issues, we propose a novel self-refining framework that integrates three key innovations: (1) panoptic segmentation for unified representation of tumors and brain anatomy, (2) patch-wise cross-modality attention enabling adaptive feature fusion from multi-modal MRI, and (3) a dynamic loss function that automatically rebalances learning to prioritize rare tumor subtypes. Our method achieves state-of-the-art performance across multiple evaluation metrics, demonstrating consistent superiority over existing architectures while maintaining robustness to varying image resolutions. These technical advances can translate to clinical benefits, including improved detection of diagnostically challenging cases, anatomically plausible tumor segmentation for surgical planning, and reduced dependence on uniformly high-resolution scans, making the framework particularly valuable for real-world clinical deployment.

End-to-end (E2E) autonomous driving systems, which map sensory inputs directly to vehicle planning, have garnered attention for harnessing the potential of data-driven methodologies in motion planning. However, current methods face two limitations that undermine their safety performance in closed-loop driving tasks. First, the predominant imitation learning (IL) paradigm overlooks long-term safety beyond predefined planning horizons, potentially guiding the ego vehicle into hazardous states. Second, the lack of reliable online evaluation mechanisms limits real-time responses to safety risks. To overcome these challenges, we propose SoAD, a safety-oriented E2E framework that integrates long-term safety awareness into planning. SoAD is distinguished by a reinforcement learning (RL)-based value estimation module to quantify the safety of planned trajectories, and a vector world model (VWM) to generate interaction-aware future rollouts. During training, the system benefits from value-guided fine-tuning (VFT) that optimizes the planning distribution to favor safer trajectories. In closed-loop deployment, a planning rescoring (PRS) mechanism is designed to perform reliable online evaluation by combining ego-conditional predictions from the VWM with corresponding safety value estimates. Experimental results on the Bench2Drive closed-loop benchmark demonstrate the state-of-the-art (SoTA) performance of SoAD, achieving a 66.40% improvement in driving score (DS) compared to the vectorized scene representation for efficient autonomous driving (VAD) baseline, while zero-shot evaluation on the driving in occlusion simulation (DOS) benchmark further highlights its strong generalization ability.

This article addresses the bipartite output consensus problem for descriptor discrete-time multiagent systems (MASs) subject to both process faults and parameter uncertainties. The study is framed within multiagent cyber–physical systems (CPSs), aiming to achieve secure and resilient consensus in adversarial environments characterized by directed signed topologies. To achieve simultaneous estimation of system states and fault signals under discrete-time descriptor dynamics, a new distributed estimator is designed by integrating the directed signed graph topology with a discrete-time generalized algebraic Riccati equation (ARE). Furthermore, a distributed observer is constructed for each follower to track the exosystem state, enabling the bipartite tracking of reference trajectories across two opposing subgroups. This design effectively accommodates exosystem matrices whose eigenvalues lie entirely outside the unit circle, thus relaxing the conventional neutral-stability requirement. Based on the$p$-copy internal model principle from robust output regulation theory, a distributed fault-tolerant output-feedback control law is synthesized. This controller guarantees bipartite output consensus over directed signed networks in the presence of concurrent process faults and model uncertainties. Finally, a numerical simulation validates the feasibility and effectiveness of the proposed approach.

Satisfactory progress has been achieved recently in universal segmentation of CT images. Following the success of vision-language methods, there is a growing trend towards utilizing text prompts and contrastive learning to develop universal segmentation models. However, there exists a significant imbalance in information density between 3D images and text prompts. Moreover, the standard fully connected layer segmentation approach faces significant challenges with handling multiple classes and exhibits poor generalizability. To address these challenges, we propose VOxel Interacting with LAnguage method (VOILA) for universal CT image segmentation. Initially, we align voxels and language into a shared representation space and classify voxels based on cosine similarity. Subsequently, we develop the Voxel-Language Interaction framework to mitigate the impact of class imbalance caused by foreground-background discrepancies and variations in target volumes. Furthermore, a Complexity-Aware Sampling method is proposed to focus on region hard to segment, achieved by generating pseudo heatmaps from a trainable Gaussian mixture distribution. Our results indicate the proposed VOILA is capable to achieve improved performance with reduced parameters and computational cost during training. Furthermore, it demonstrates significant generalizability across diverse datasets without additional fine-tuning.

In this paper, the distributed bipartite output formation control problem of heterogeneous discrete-time singular multi-agent systems (MASs) is studied under the cooperative-competition network. Different from the classical formation control issue, the leader (also called exosystem) is also described by more general discrete-time singular dynamics. Based on the cooperative output regulation theory, the bipartite output formation with two opposing subgroups is realized by reconstructing bipartite tracking error and singular exosystem. To obtain the bipartite state of exosystem, a new exosystem observer associated with discrete-time singular algebraic Riccati equation (ARE) and signed directed topological matrix is designed. The observer coupling gains are introduced to enable the global error system to be included in the stability region of the unit circle. Then under the singular regulator equations and the separation principle, two classes of distributed observed-based controllers via state and output feedbacks are respectively presented for bipartite formation scenarios where the states of followers are known and unknown. Moreover, by redefining the expected relative dynamic positions, the proposed control strategies can be extended to achieve time-varying bipartite output formation. Finally, simulation results are given to verify the feasibility and effectiveness of the proposed algorithms.

Accurate prediction of pathologic complete response to neoadjuvant chemotherapy non-invasively before treatment via dynamic contrast-enhanced magnetic resonance imaging is vital for developing a personalized therapy strategy. However, the application of deep learning in this domain is characterized by its black-box nature, largely relying on post-hoc analysis to interpret final decision-making. This reliance results in a lack of self-interpretability in the operational mechanisms of feature extraction, feature fusion, and decision-making. Moreover, these models have demonstrated unsatisfactory prediction performance due to insufficient feature modeling. To address these issues, we propose a self-interpretable deep learning network that can provide the intrinsic interpretability of feature extraction, multi-scale feature fusion, and final prediction. First, the interpretable perception module is designed to extract features both effectively and interpretably. Furthermore, the interpretable adaptive multi-scale feature fusion module is proposed to fuse multi-scale features. Finally, an end-to-end self-interpretable deep learning network is presented to predict pathologic complete response with self-interpretability. Validated on a multi-center pre-treatment dynamic contrast-enhanced magnetic resonance imaging dataset, our self-interpretable deep learning network outperforms state-of-the-art methods in both prediction performance and self-interpretability, improving the area under the receiver operating characteristic curve by at least 4.81% while providing both qualitative and quantitative self-interpretability. Our study demonstrates that our proposed self-interpretable deep learning network can extract key information from pre-treatment breast dynamic contrast-enhanced magnetic resonance imaging while enhancing both the prediction performance and the transparency of the model, thereby improving its trustworthiness in clinical settings.

In this note, a novel prescribed fixed-time adaptive tracking control scheme is developed to cope with the fixed-time tracking control issue for a category of constrained MIMO nonlinear cyber-physical systems (CPSs) with exogenous perturbations, which suffer from deception attacks started in controller-actuator (C-A) channel. Distinguished from the conservative dynamic surface control (DSC) schemes with a linear filter, a novel nonlinear filter is designed in our strategy, which can tackle the intrinsic issue of explosion of computational complexity and promote the system performance. Besides, a new barrier Lyapunov function (BLF) is designed to ulteriorly enhance the tracking performance on the basis of the prescribed performance function (PPF) approach. Prominently, the proposed control strategy could accommodate the exogenous interferences and deception attacks simultaneously. Furthermore, we have substantiated that the developed approach can not only make certain that all the tracking errors of the resulting closed-loop system, including output tracking errors and virtual tracking errors, enter a prespecified small region near equilibrium point with fixed-time convergence rate, but also guarantee them obey the corresponding constraints throughout the entire control operation, where the regulation time and the tracking accuracy level keep prior known and could be prespecified arbitrarily. Finally, the validity and effectiveness of the proposed control scheme are illustrated through a representative application instance.

A new fixed-time adaptive fuzzy tracking control framework is proposed to resolve the fixed-time tracking control issue for a sort of nonlinear cyber–physical systems (CPSs), which subjugate to malignant attacks occurring in controller-actuator (C–A) channel. As a special point, the developed control architecture not only does not depend upon the accurate system model, but also can treat the exogenous perturbations and venomous attacks in unison. At the same time, distinct from the extant control studies for uncertain CPSs subject to malicious attacks which could only ensure the infinite-time or finite-time system performance, the presented control strategy owns the ability to assure the fixed-time system performance. Moreover, it has been theoretically substantiated that the developed control method can ensure that the output tracking error converges to a predefined slim region of the equilibrium point in fixed time, where both the regulation time and tracking precision degree are known beforehand and could be preconfigured by selecting the relative control design parameters suitably. In the last resort, we utilize two typical instances to illustrate the feasibility and effectiveness of our proposed approach.

The problem of enhancing the robust performance of nonlinear fault estimation (FE) is addressed by proposing a novel real-time gain-scheduling mechanism for discrete-time Takagi-Sugeno fuzzy systems. The real-time status of the operating point for the considered nonlinear plant is characterized by using these available normalized fuzzy weighting functions at both the current and the past instants of time. To achieve this, the developed fuzzy real-time gain-scheduling mechanism produces different switching modes by introducing key tunable parameters. Thus, a pair of exclusive FE gain matrices is designed for each switching mode on the strength of time-varying balanced matrices developed in this study, respectively. Since the implementation of more FE gain matrices can be scheduled according to the real-time status of the operating point at each sampling instant, the robust performance of nonlinear FE will be enhanced over the previous methods to a great extent. Finally, considerable numerical comparisons are implemented in order to illustrate that the proposed method is much superior to those existing ones reported in the literature.

This article studies the distributed fault-tolerant bipartite output synchronization problem of discrete-time linear multiagent systems (MASs) with process faults under a general directed signed graph. The reference signal is generated by an autonomous exosystem, which can also be seen as a leader. All followers are divided into two subgroups with antagonistic interactions, and the followers in each subgroup are cooperative. We aim to solve the bipartite fault-tolerant control (FTC) problem via the output regulation theory such that bipartite output synchronization can be achieved in the presence of process faults, that is, the outputs of followers with different subgroups can approach the output of exosystem with the same magnitude and the opposite sign regardless of process faults. To estimate the states and the faults of each follower, a simultaneous state and fault estimator based on the neighboring signed output estimation error and the standard discrete-time algebraic Riccati equation (ARE) is designed. Besides, a new exosystem observer with two classes of convergence conditions relying on the respective solutions of standard and modified AREs is provided. All eigenvalues of the exosystem matrix can lie completely outside the unit circle. Based on these estimations, we present a distributed fault-tolerant output feedback controller, which can overcome the no-loops constraint. Finally, simulation results are given to demonstrate the analytic results.

A novel fixed-time adaptive fuzzy tracking control strategy is developed to resolve the fixed-time tracking control issue for a category of state-constrained high-order uncertain nonlinear cyber-physical systems (CPSs), which suffer from malicious attacks launched in controller-actuator (C-A) channel. The proposed control strategy is independent of the exact system model and can accommodate external disturbances and malicious attacks. Meanwhile, distinguished from the existing control methods for state-constrained uncertain CPSs under malicious attacks, which can only ensure the system performance when time leans toward infinity or in finite time, the presented control strategy can guarantee the system performance in fixed time. It has been substantiated that the developed control strategy can make sure that the output tracking error converges to a predefined small neighborhood of the origin, and all signals of the resulting closed-loop system satisfy the corresponding time-varying state constraints during the total operation. In particular, the settling time and tracking precision are known and can be preconfigured by selecting the design parameters appropriately. More significantly, all the initial states of the system are independent of the developed fixed-time adaptive fuzzy tracking control law and can be arbitrarily chosen in the constrained region specified by the corresponding time-varying state constraints conditions. Ultimately, a representative simulation is supplied to illustrate the effectiveness and superiority of the proposed control scheme.

This article studies the$\mathcal{l}_{1}$filtering problem for positive Takagi–Sugeno fuzzy systems. A pair of filters is sought to assure the stability and positivity of filtering error systems with guaranteed$\mathcal{l}_{1}$performance indices. For this purpose, a pair of linear copositive Lyapunov functions that capture the positivity characteristics of filtering error systems is adopted to derive sufficient conditions for the existence of filters, which boils down to the bilinear programming problem. To solve this dilemma, we propose a convex approximation strategy, and thus, a sequence of convex surrogates is developed to reasonably replace the nonconvex constraints, based on which a successive linear programming algorithm is constructed. Therefore, the filters can be achieved by solving a battery of linear programming. Finally, the efficiency of the designed filtering scheme is justified by simulation validations.

This article considers the control synthesis for positive nonlinear systems modeled by Takagi–Sugeno (T–S) fuzzy characterizations. A fuzzy static output feedback (SOF) controller is sought to assure the closed-loop stability and positivity with a guaranteed$\mathcal {l}_{1}$performance level. The favorable positivity enables us to adopt a linear co-positive function as the Lyapunov candidate, and we construct the essential conditions for the existence of fuzzy SOF controllers in the form of bilinear programming (BP) problem. To provide a direct and generic algorithm for solving the BP problem, we propose a vital approximation strategy by which the general bilinear constraints are replaced with a sequence of convex surrogate ones. Recognizing the fact that a fundamental promise of the proposed successive linear programming algorithm is a feasible initial solution, we also develop an iterative procedure to calculate an initial controller gain. Finally, we clarify the applicability of the developed scheme with two examples.

This article focuses on developing a static output feedback (SOF) control scheme for Takagi–Sugeno fuzzy systems. The proposed SOF controller does not share the same premise membership functions with the model, which permits enhancing the flexibility in controller design and implementation. By contrast with the state feedback case, the SOF control generally results in nonconvex design conditions. To circumvent this problem, we develop a successive convex optimization algorithm, which is based on solving a sequence of more tractable convex optimization problems obtained by approximating the nonconvex constraints with some convex ones. As a heuristic algorithm, the validity of the developed successive convex optimization algorithm is highly affected by initial conditions, and, recognizing this, we put forward an iterative procedure for determining the feasible initial condition. Finally, two illustrative examples are presented to validate the efficiency of the proposed algorithms.

Realistic and diverse simulation scenarios with reactive and feasible agent behaviors can be used for validation and verification of self-driving system performance without relying on expensive and time-consuming real-world testing. Existing simulators rely on heuristic-based behavior models for background vehicles, which cannot capture the complex interactive behaviors in real-world scenarios. Meanwhile, existing learning-based methods are typical insufficient, yielding behaviors of traffic participants that frequently collide or drive off the road especially for a long horizon. To address this issue, we propose TrajGen, a two-stage trajectory generation framework, which can capture more realistic and diverse behaviors directly from human demonstration. In particular, TrajGen consists of the multi-modal trajectory prediction stage and the reinforcement learning based trajectory modification stage. In the first stage, we propose a novel auxiliary RouteLoss for the trajectory prediction model to generate multi-modal diverse trajectories in the drivable area. In the second stage, reinforcement learning is used to track the predicted trajectories while avoiding collisions, which can improve the reactivity and feasibility of generated trajectories. In addition, we develop a simulator I-Sim that can provide support for data-driven agent behavior simulation and train reinforcement learning models in parallel based on naturalistic driving data. The vehicle model in I-Sim can guarantee that the generated trajectories by TrajGen satisfy vehicle kinematic constraints. Finally, we give comprehensive metrics to evaluate generated trajectories for simulation scenarios, which shows that TrajGen outperforms either trajectory prediction or inverse reinforcement learning in terms of fidelity, reactivity, feasibility, and diversity.

This article discusses$H_{\infty }$fuzzy control for the nonlinear fourth-order parabolic equation with input delay via collocated actuator/sensor pairs. We suggest that the interval [0, 1] is divided into$M$subdomains, where sensors provide spatially averaged/point discrete-time state measurements. The control design strategy is proposed based on output measurements. We derive constructive conditions ensuring that the resulting closed-loop system is internally exponentially stable and has$H_{\infty }$performance by means of the Lyapunov approach.

This article is concerned with the design of robust resilient control strategy for nonlinear systems under Denial-of-Service (DoS) attacks. First, Takagi–Sugeno fuzzy model is employed to approximate the nonlinear dynamics, and an improved sensor system is constructed by fuzzy observer and fuzzy predictor. By applying periodic event-triggered control strategy, the relationship between the estimated state and predicted state is obtained by event-triggering mechanism, which can reduce transmission attempts significantly in the sensor-to-controller channel. Second, caused by DoS attacks, the transmission attempts may be denied over the communication network. Under the proposed transmission policy with a bounded update period, input-to-state stability (ISS) of closed-loop systems can be guaranteed. Moreover, the maximum frequency and duration of DoS attacks are calculated. Finally, simulation results are provided to show the effectiveness of the proposed method.

This paper considers delayed fuzzy control of one-dimensional (1-D) reaction-diffusion equation under distributed in-domain point actuation and measurements. The delay may be uncertain, but bounded by a known upper bound. We propose an observer-based controller that employs the averaged values of the observer. Sufficient conditions ensuring exponential stability of the closed-loop system are established in terms of linear matrix inequalities by using the Lyapunov–Krasovskii method and the descriptor method. A numerical example demonstrates the efficiency of the results.

This paper is concerned with the dissipativity-preserving model reduction problem for Takagi–Sugeno (T–S) fuzzy systems. The principal goal is to approximate the high-order T–S model with a dissipative reduced-order T–S model. The number of fuzzy rules and the membership functions of the reduced-order T–S model are chosen freely to enhance design flexibility. To this end, an$H_{\infty }$performance index is used to describe the approximation error. Meanwhile, dissipativity of the reduced-order model is guaranteed by satisfying a dissipation inequality. With the aid of fuzzy-basis-dependent Lyapunov functions and slack variable techniques, less conservative design conditions for reduced-order models are derived. An algorithm is proposed to calculate a desired reduced-order model. A rail traction control system is given to illustrate the effectiveness of the proposed method and the advantages over the existing methods.

This paper presents a model-free solution to the robust stabilization problem of discrete-time linear dynamical systems with bounded and mismatched uncertainty. An optimal controller design method is derived to solve the robust control problem, which results in solving an algebraic Riccati equation (ARE). It is shown that the optimal controller obtained by solving the ARE can robustly stabilize the uncertain system. To develop a model-free solution to the translated ARE, off-policy reinforcement learning (RL) is employed to solve the problem in hand without the requirement of system dynamics. In addition, the comparisons between on- and off-policy RL methods are presented regarding the robustness to probing noise and the dependence on system dynamics. Finally, a simulation example is carried out to validate the efficacy of the presented off-policy RL approach.

This paper considers the model-reduction problem for continuous-time Takagi–Sugeno (T–S) fuzzy systems. Different from existing full-frequency methods, a finite-frequency model-reduction method is proposed in this paper. The proposed method can get a better approximation performance when input signals belong to a finite-frequency domain. To this end, a finite-frequency$H_\infty$performance index is first defined. Then, a sufficient finite-frequency performance analysis condition is derived by the aid of Parseval's theorem and quadratic functions. Based on this condition and projection lemma, three model-reduction algorithms for T–S fuzzy systems with input signals in low-frequency, middle-frequency, and high-frequency domain are obtained, respectively. Finally, an example is given to illustrate the effectiveness of the proposed method.

This paper proposes relaxed stabilization conditions of discrete-time nonlinear systems in the Takagi–Sugeno (T–S) fuzzy form. By using the algebraic property of fuzzy membership functions, a novel nonparallel distributed compensation (non-PDC) control scheme is proposed based on a new class of fuzzy Lyapunov functions. Thus, relaxed stabilization conditions for the underlying closed-loop fuzzy system are developed by applying a new slack variable technique. In particular, some existing fuzzy Lyapunov functions and non-PDC control schemes are special cases of the new Lyapunov function and fuzzy control scheme, respectively. Finally, two numerical examples are provided to illustrate the effectiveness of the proposed method.

This paper is concerned with the problem of fuzzy-filter design for discrete-time nonlinear systems in the Takagi–Sugeno (T–S) form. Different from existing fuzzy filters, the proposed ones are designed in finite-frequency domain. First, a so-called finite-frequency$l_2$gain is defined that extends the standard$l_2$gain. Then, a sufficient condition for the filtering-error system with a finite-frequency$l_2$gain is derived. Based on the obtained condition, three fuzzy filters are designed to deal with noises in the low-, middle-, and high-frequency domain, respectively. The proposed fuzzy-filtering method can get a better noise-attenuation performance when frequency ranges of noises are known beforehand. An example about a tunnel-diode circuit is given to illustrate its effectiveness.