This paper investigates a robust Takagi–Sugeno (T–S) fuzzy controller for flapping-wing micro aerial vehicles (FWMAVs) subject to actuator saturation and external disturbances. To this end, we first construct a state-scheduled T–S fuzzy model to represent the longitudinal dynamics of FWMAVs, explicitly accounting for state-dependent nonlinearities. Based on this model, we propose a two-loop control architecture: an inner-loop T–S fuzzy controller that regulates the vertical position and pitch angle, and an outer-loop PD controller that achieves full position-tracking by generating a pitch reference from the desired longitudinal position. Then, we formulate the T–S fuzzy stabilization conditions as linear matrix inequalities (LMIs) that guarantee closed-loop stability. Specifically, by exploiting the mismatch between the current and subsequent fuzzy basis functions (FBFs), we introduce a relaxation method that incorporates additional slack variables into the stabilization conditions, thereby reducing conservatism. Finally, numerical comparisons and simulations for FWMAVs are presented to verify the reduced conservatism and effectiveness of the proposed method.
This work presents a comprehensive study on the formation control of multi-agent systems (MAS) utilizing proportional-derivative (PD) control within the robot operating system (ROS) framework. The control objective is to achieve precise formation maintenance and robust navigation for Mecanum wheeled drive robots (MWDRs) in various formations. The methodology involves the implementation of PD control algorithms for both leader and follower robots running in ROS for real-time control. Different formation patterns are explored, and the system’s performance is evaluated under various conditions. The leader-dependent configuration demonstrates its effectiveness in maintaining formation trajectories, with simulations and experiments confirming its capability to handle both simple and moderately complex trajectories. Finally, the corresponding results show the effectiveness of the proposed control strategy in maintaining the desired formation integrity and ensuring smooth trajectory tracking.
For the nonlinear system with states and input quantization, this paper proposes a predefined-accuracy control strategy based on the self-triggered mechanism. The traditional predefined-accuracy control methods lead to frequent parameter updates and resource wastage. To address this issue, this paper proposes a new predefined-accuracy control strategy based on quantized states and input. To address the discontinuity problem caused by states quantization, this paper establishes a mathematical relationship between quantized and unquantized signals to effectively compensate for quantization errors. To further reduce communication resource consumption, this paper introduces a self-triggered mechanism that automatically calculates trigger moments without requiring real-time monitoring. In addition, by introducing two nonlinear functions, this paper designs an adaptive controller with predefined accuracy. Based on the proposed controller, the tracking error can converge within a predefined range and all closed-loop signals remain bounded. Finally, simulation results validate the effectiveness and superiority of the proposed control scheme.
Bipedal wheel-legged robots integrate wheeled efficiency with legged adaptability; however, balance controllers designed with fixed inertial parameters can suffer severe performance degradation under payload variations and impulsive disturbances due to model mismatch. This paper presents an LQR–VMC balance-control framework with a dual-mode online inertial-parameter identification scheme for a bipedal wheel-legged robot. A sagittal-plane floating-base/virtual-leg model is derived and coupled with a virtual model control (VMC) torque-distribution layer. During dynamically excited motions, an augmented-state extended Kalman filter (EKF) estimates key inertial parameters online. During quasi-static phases with weak excitation, a static-load estimator reconstructs the ground-reaction force from measured hip torques via the VMC mapping and updates the dominant floating-base mass. The identified parameters are used to update the linearized model and schedule the LQR gains in real time. Simulations and hardware experiments involving slopes, obstacle traversal, impacts, and ramp crossing validate the proposed method. Compared with a fixed-parameter LQR baseline, the proposed framework reduces the RMS velocity-tracking error by approximately 20
This paper addresses the issue of robustly stabilizing along with attaining H_∞ control pertaining to a specific type of discretely switched nonlinear networked control systems (SNNCSs) with non-fragile controllers and sampled data. A state-feedback controller is developed through a combination of average dwell time in conjunction with multiple Lyapunov function methods. This controller guarantees robust exponential stability around the equilibrium point having a mentioned disturbance attenuation level ( γ >0 ), accounting for all potential uncertainties. Exponential stabilization conditions, expressed as linear matrix inequalities (LMIs), are derived by the construction of an appropriate Lyapunov–Krasovskii functional (LKF). These results are proven numerically.
This article devises a centralized contouring control strategy for networked multiaxis motion control systems (NMAMSs) based on fully actuated system (FAS) approach. Firstly, a unified FAS model of the NMAMSs is established, which explicitly considers the parameter heterogeneity of each individual-axis and the influence of nonlinear lumped disturbances caused by networked time-varying delays, modeling uncertainties, and external disturbances. In contrast, existing work only considers cases with homogeneous parameters and lacks a systematic analysis of disturbances. Subsequently, an intermediate estimator (IE) is devised to estimate and compensate the lumped disturbances in a feed-forward manner. Further, a centralized contouring error-constrained model predictive control (CEMPC) strategy is developed. The contouring error is explicitly incorporated into the cost function to realize accurate individual-axis tracking and effective contouring error reduction simultaneously. Finally, the proposed method is rigorously analyzed for feasibility and stability, and the effectiveness and superiority is verified by numerical simulations and rapid control prototyping (RCP) experiments conducted on a NMAMSs experimental platform.
The complex noise environments found in industrial settings lead to a decline in the performance of deep learning-based anomaly detection models. While previous studies have primarily attempted to address this issue by increasing model complexity, this paper proposes an efficient approach that improves signal quality by incorporating the statistical characteristics of the data during the preprocessing stage. In this paper, we analyze the covariance matrix of the training dataset to establish a Global Whitening Baseline and propose a preprocessing pipeline utilizing Zero phase Component Analysis (ZCA) Whitening based on this baseline. This technique removes correlations between frequency bands in the Log-Mel Spectrogram and normalizes the variance by sphering, thereby suppressing the influence of dominant background noise and maximizing the model’s efficiency in extracting subtle anomaly patterns. Furthermore, by applying a multi-frame sliding window technique to the whitened features, we preserved temporal context and ensured robustness against local noise fluctuations. Experimental results show that the proposed preprocessing method improved anomaly detection accuracy Area Under the Curve (AUC) by more than 10
This paper proposes a robust load frequency control (LFC) strategy for dual-area thermal power systems (DATPSs) with reheat turbines. Unlike conventional sliding mode control (SMC) approaches that often suffer from severe chattering and limited transient improvements, the proposed scheme integrates an observer-based second-order SMC (SoSMC) with a PID sliding surface (PIDSS) structure and state feedback compensation (SFC). First, a PIDSS combined with pole-placement SFC is constructed using estimated state variables, thereby achieving faster transient dynamics and reduced steady-state error compared with classical sliding surfaces. Second, the closed-loop stability of the proposed design is rigorously established via Lyapunov’s theorem and linear matrix inequality analysis. Third, by employing a newly developed second-order sliding control law, the proposed method effectively suppresses oscillations and mitigates the inherent chattering phenomenon of conventional SMC schemes. Extensive simulations on DATPSs with reheat turbines, considering step load variations, matched uncertainties, and nonlinear constraints, demonstrate the superiority of the proposed strategy. Specifically, compared with recently published SMC-based controllers, the new scheme achieves significantly faster settling time, lower overshoot, and enhanced robustness, while maintaining stable tie-line power exchange. These results highlight the effectiveness and novelty of the proposed observer-based SoSMC framework as a reliable solution for frequency regulation in interconnected power systems.
Model predictive control (MPC) is characterized by its high control accuracy and superior dynamic performance, making it a prevailing trend in the field of control algorithms for permanent magnet synchronous motors (PMSMs). However, the precision of the motor model significantly influences the effectiveness of MPC. To address this issue, this paper proposes a novel model-free predictive current control algorithm that is independent of the parameters of PMSM. First, the ultra-local model is established to isolate the lumped disturbance term. Based upon the construction of an ultra-local model, a novel extended-state observer is utilized to estimate the lumped disturbances and current at the (k + 1)-th moment. Then a simplified three-vector MPC incorporating the duty cycle reconstruction technique is adopted to improve the robustness of the control system. Finally, experimental results show that under 150
Connected automated vehicles (CAVs) can improve energy efficiency through eco-driving at signalized intersections by avoiding energy-wasting stop-and-go patterns. However, when a lead vehicle’s behavior prevents safe non-stop passage, energy-optimal crossing may conflict with rear-end collision avoidance, creating a CAV-specific dilemma zone problem. This paper presents a novel control framework that couples energy-optimal trajectory planning with collision avoidance constraints. Using pontryagin’s minimum principle, we derive analytical solutions that minimize energy while satisfying both green phase passage and rear-end safety constraints. The framework provides real-time feasibility verification to establish explicit go-or-stop decision criteria, transitioning to safe fallback modes when energy-optimal intersection crossing is infeasible. Simulation results demonstrate that the proposed approach achieves energy-efficient operation while ensuring collision-free fallback mode transitions across various dilemma zone scenarios.
This paper addresses the problem of event-triggered secure control for networked positive systems operating under hybrid network attacks. It is assumed that the communication network is subject to both random deception attacks and denial-of-service attacks. To conserve limited network bandwidth, a memory-based dynamic event-triggering scheme (MDETS) is proposed for scheduling data transmissions. Unlike conventional dynamic event-triggering strategies, the proposed mechanism integrates historical triggering information into the threshold condition enabling more flexible adjustment of the triggering threshold. This effectively reduces redundant data transmissions and enhances network resource utilization efficiency. Based on the MDETS, an event-triggered model for networked positive systems under hybrid attacks is established. Moreover, sufficient conditions ensuring the positivity and stability of the closed-loop system are derived, along with a corresponding output feedback controller design. Finally, two examples are provided to demonstrate the effectiveness and feasibility of the proposed approach in enhancing system security, reducing communication load, and improving control performance.
This paper proposes a periodic event-triggered formation (PETF) protocol based on adaptive fuzzy control for nonlinear multi-agent systems (NMASs) under directed topology. By introducing second-order auxiliary dynamics, the formation problem is transformed into a stabilization problem for the constructed error system. Fuzzy logic systems (FLSs) are employed to handle uncertain nonlinear functions. To reduce the usage of communication resources (UCRs), an adaptive fuzzy PETF strategy is developed, in which the event-triggering mechanism (ETM) is monitored only at sampling instants. Moreover, the upper bound of the sampling period is explicitly derived. It is proven that the formation errors converge to a small neighborhood of the origin. Finally, a practical example is provided to verify the effectiveness of the proposed scheme.
This paper proposes an autonomous navigation framework for orchard weeding robots based on the ZTR (Zero Turn Radius) platform. Unlike conventional agricultural robots, our system utilizes multiple sensors, including GNSS, IMU, 2D LiDAR, and cameras, to achieve reliable object detection and accurate position estimation based on the Extended Kalman Filter (EKF). For path tracking, we implemented a pure-pursuit algorithm with a look-ahead distance strategy, which is particularly effective for ensuring stable navigation at low speeds. To evaluate the path tracking performance, experiments were conducted in a structured orchard environment. During autonomous navigation for an average of 8.54 min, the proposed framework achieved an average trajectory error of 11.66 cm. For localization performance evaluation, the proposed algorithm was compared with a baseline method that fuses raw GPS and IMU data. The proposed method achieved a mean error of 0.0078 m, representing a 78.3
This paper investigates the problem of protocol-based I-2 - I-infinity output feedback tracking control for discrete-time nonlinear systems subject to the fading channel. First, an interval type-2 (IT2) Takagi-Sugeno (T-S) fuzzy model is employed to characterize the nonlinear dynamics of the plant. Then an improved event-triggered stochastic protocol is introduced to reduce the data transmission frequency and schedule the transmission sequences. On this basis, an IT2 fuzzy output feedback tracking controller is designed to ensure the mean-square asymptotic stability and I-2 - I-infinity tracking performance of the closed-loop system. Furthermore, all design parameters can be obtained by solving a set of linear matrix inequalities. Finally, two simulation examples are provided to verify the effectiveness of the proposed control approach.
As advances in missile technology compress engagement windows and complicate trajectories, modern theater missile defense systems are required to respond rapidly and adaptively. We present a unified architecture that couples a modular simulator built on the Discrete Event System Specification (DEVS) formalism with an AI-based weapon–target assignment (WTA) algorithm. The proposed architecture supports rapid prototyping across diverse engagement scenarios and enables quantitative evaluation of WTA strategies. In our case study, we evaluate the proposed AI-based algorithm against heuristic baselines under varied theater defense conditions. The results demonstrate that the AI-based policy exhibits non-myopic behavior, conserving interceptors for anticipated future threats. Under resource constraints, this leads to more efficient use of defensive assets. We also find that this non-myopic tendency can be tuned via hyperparameters, indicating that the AI-based WTA flexibly covers a broader range of engagement strategies.
This study presents a variable-augmented free-weighting matrix framework for analyzing the exponential stability of time-delayed neural network systems. Significant emphasis has been placed on the reduced conservatism and simplicity of contemporary stability criteria in the systematic evaluation of global exponential stability. Time-varying delayed neural networks are addressed using an enhanced free-weighting matrix method with variable augmentation to effectively resolve this issue. This approach reduces superfluous decision variables and averts the formation of higher-order delay terms. This leads to the development of new-brand global exponential stability conditions in the form of linear matrix inequalities, offering improved computational efficiency and reduced conservatism. The practicality and effectiveness of the proposed approach are finally demonstrated through two numerical examples.
This paper focuses on addressing the synchronization issue of inertial neural networks (INNs) with heterogeneous time-varying delays (HTVDs) under hybrid cyber attacks. First, hybrid cyber-attacks in network communications are considered, highlighting the variety of attack strategies and the uncertainties inherent in their occurrence and impact. Second, an aperiodic quantized sampled-data (AQSD) controller is proposed to address the limited transmission capacity in network communications. In addition, an asymmetric Lyapunov–Krasovskii functional (LKF) method and looped-functional terms are introduced to further relax the initial constraints requirement. Next, less conservative synchronization criteria for the drive–response systems are deduced. Finally, numerical simulation is used to demonstrate the effectiveness and superiority of the proposed method.
Urban autonomous driving requires route-level planners that connect start and goal locations through a road network while respecting traffic regulations encoded in high-definition (HD) semantic maps. Existing route-level planners with fixed hand-crafted costs can mis-rank maneuvers in dense urban layouts, whereas more learning-based approaches often improve adaptivity but do not retain an explicit, auditable route-selection objective or classical heuristic-search properties. To address these limitations, this paper presents a semantic lane-segment road-graph planner that retains an explicit, nonnegative nominal edge cost (geometric distance plus interpretable rule/penalty terms) and learns bounded multiplicative edge-wise cost-shaping factors with an edge-aware graph neural network (GNN). The bounded shaping factors preserve the admissibility and consistency of a scaled geometry-based heuristic, enabling standard A* search on the shaped objective while retaining predictable and auditable behavior. To reduce reliance on any single hand-tuned cost design, the GNN is trained by distilling routing decisions from multiple diverse A*-based teachers and by using a calibration regularizer that discourages unnecessary deviation from nominal costs. Experiments in CARLA across 7 towns and 2100 trials (ego-only rollouts with static-map inputs) show improved route completion and substantially reduced map-based collision and traffic-rule-violation scores compared with classical and learning-based baselines, with only modest increases in path length and runtime.
In the industrial fuel ethanol fermentation process, accurate prediction of out-of-tank ethanol concentration, i.e., its main performance, is essential for its control and optimization. However, ethanol fermentation is a long-period batch process, whose performance is affected by multiple static variables, i.e., initial tank entry information, and dynamic variables, including time-series operational variables and compositional variables sampled at different time points. Existing prediction models are usually based on static variables and dynamic variables. However, dynamic variables generally require sampling over a period of 0–40 h, resulting in delays in the timely prediction of fermentation performance. To address this challenge, this paper proposes a novel model called the Temporal Fusion Enhanced Network (TFEN). First, the enhancement module extracts temporally fused features from historical data, including static variables at the beginning of production, low-frequency dynamic variables sampled at different hours, and high-frequency dynamic variables. Second, the reconstruction module learns the hidden relationship between temporally fused features and static variables. Third, static variables from the dataset are passed through the reconstruction module to generate new features, which are used to train the model. Finally, TFEN can accurately predict the ethanol fermentation performance only based on static variables at the beginning of production. Comparative experiments indicate that the predictive performance of TFEN not only surpasses that of other algorithms, but also exceeds the results of the model employing static variables and dynamic variables at 8 h and 24 h as the input variables. Moreover, its accuracy is close to the model with other dynamic variables at 40 h. Also, the more comprehensive the historical data is, the more accurate TFEN is. The proposed TFEN achieves the best prediction accuracy with an RMSE of 0.0526 (normalized), corresponding to 0.339 g/100 mL in physical units.
This paper deals with the input-to-state stability (ISS) of stochastic delayed systems with packet losses by event-triggered delayed impulsive control (ETDIC). Based on the Lyapunov function method, the designed event-triggered mechanism (ETM) only depends on the information from the system state and external input, without requiring the existence of an upper bound for the triggering interval. It can achieve the desired control performance and avoid Zeno behavior. Our results reveal the tradeoff among triggered parameters, impulsive intensity and impulsive delay at a certain packet loss rate (PLR). Especially, in the average sense, we not only relax the requirement on impulsive strength but also remove the restriction on the magnitude of impulsive delay. Finally, several examples are provided to illustrate the validity of our conclusions.
