ISA transactions最新文献

Hydraulic manipulators use a main control valve to distribute flow from a single pump to multiple actuators, enhancing the energy efficiency. However, this architecture causes electro-hydraulic system (EHS) performance constraints, such as flow saturation/coupling, cavitation, and valve spool stroke limitations, which can impair the coordination. To address these challenges, this paper proposes a novel framework for multi-axis coordination control. First, the non-dominated sorting genetic algorithm II is used to optimize the interpolated trajectory based on fifth-order B-splines in the joint space, which optimizes the motion reference. Second, to precisely follow the parameterized optimal trajectory, normalized guidance vector fields (GVFs) are constructed, and a convergence term is designed based on the Euclidean distance function to drive the system to globally and asymptotically converge to the target path. Lyapunov stability theory is used to prove the boundedness of the system under model uncertainties and external disturbances. The quantitative relationship between the disturbance intensity and the trajectory deviation is derived, and a singularity avoidance mechanism is proposed that ensures the uniqueness and continuity of the vector field corresponding to the end-effector trajectory control (ETC) in the global workspace solution. Finally, the performance constraints of the EHS are transformed into a real-time optimization problem, rather than relying on complex global optimizers. The GVF command is adjusted to change the velocity of the ETC in Cartesian space. The effectiveness of the proposed scheme is verified through simulations and experiments.

This article addresses the optimal cooperative output regulation problem for linear heterogeneous multi-agent systems, ensuring both transient and steady-state performance of regulated errors while minimizing the predefined cost. To achieve the explicit specifications on the regulated error even if not all agents can directly access the regulated error, we construct a hierarchical "observation-control" framework. At the observation layer, by imposing constraints on the norms of relative observation states, a distributed edge-based observer is developed to reproduce the states of the exosystem with explicitly prescribed transient performance. At the control layer, by employing a data-driven actor-critic learning algorithm and the prescribed performance control, we derive an optimal control scheme that enforces prescribed constraints on auxiliary tracking error norms and minimizes the predefined cost without solving the nonlinear Hamilton-Jacobi-Bellman equation. Through indirect specifications on observation and auxiliary tracking errors, the multi-dimensional regulated errors converge to a predefined residual set within a specified time, reducing the computational burden of decoupling. Finally, experimental results validate the effectiveness of the proposed control scheme.

In order to enhance the control performance and formation security, this paper studies the transient (overshoot and convergent speed) and steady performance constraints on formation reconfiguration of multi-aircraft system (MAS) under leader-follower structure. A fast fixed-time prescribed performance adaptive control (FFTPAC) strategy is proposed, which composes of a brain emotional learning neural network (BELNN) disturbance observer, a preset-time polynomial prescribed performance function and a fast variable power fixed-time reaching law. Firstly, BELNN is creatively designed to estimate external disturbances with simple structure and fast learning ability, which can reduce chattering and initial estimated error remarkably compared with high-order sliding mode disturbance observers. Secondly, a novel preset-time polynomial prescribed performance function is presented to ensure that formation tracking error changes within the prescribed region and converges to a specified small range at a time set in advance directly. Moreover, initial change form of the tracking error is flexible to avoid excessive initial control input value in traditional prescribed performance control. Thirdly, in order to promote efficiency of the formation reconfiguration, a novel fast variable power fixed-time reaching law is proposed to increase convergent rate of the system, where a novel nonlinear function involved system states and a variable power term are introduced based on constant power reaching law. The FFTPAC strategy can achieve excellent control performance and robustness under unknown disturbances during formation reconfiguration of MAS. Finally, the superiority of FFTPAC is verified by comparing with existing advanced methods.

Proton exchange membrane fuel cell (PEMFC) is used in several fields due to its high efficiency and environmental friendliness. Durability issues remain one of the main barriers limiting its large-scale commercialization. Although prognostic techniques can optimize operation and extend fuel cell lifetime, their accuracy is often compromised by recoverable fault behaviors. Therefore, addressing this interference has become a crucial challenge in achieving reliable PEMFC prognostics. To tackle these issues, a diagnostic-prognostic hybrid framework integrating data-driven recoverable fault recognition with model-based degradation prediction is proposed. Specifically, a recoverable fault recognition model is developed to determine the operational state of the fuel cell, where data augmentation is employed to mitigate the adverse effects of data imbalance on classification performance. Using a prognostic strategy-guided particle filter (PSG-PF) model for degradation trend prediction and remaining useful life (RUL) estimation. The proposed method is evaluated by the dynamic load aging experimental data of PEMFC. The results show that the proposed method achieves over 98% accuracy in recoverable fault recognition. Meanwhile, by effectively circumventing the impact of recoverable faults, the average relative accuracy of predicted RUL reached 86%. The proposed framework can achieve reliable recognition of recoverable faults and credible RUL prediction under complex operating conditions. It enhances the robustness of PEMFC prognostics, supporting longer lifetime and more stable operation of fuel cells.

This paper investigates the problem of designing optimal actuator and sensor attack strategies from the attacker's perspective. The first objective is to compromise system controllability and observability with minimal cost by attacking the smallest number of actuators and sensors. However, since certain actuators or sensors may be inherently difficult or costly to compromise, we further investigate the scenario in which the attacker aims to determine the minimum cardinality such that attacking any set of components of this size inevitably renders the system uncontrollable or unobservable. This essentially corresponds to disrupting the controllability and observability resilience of the system. If the system matrix A has no repeated eigenvalues, both problems can be solved in polynomial time. When the system matrix A has repeated eigenvalues, two algorithms, called the forward and reverse greedy algorithms, are proposed to address the fragility and resilience problems. Compared with the existing greedy algorithms, the proposed forward and reverse greedy algorithms initiate attacks from the empty set and the full set of actuators and sensors, thereby further reducing the computational complexity. We evaluate the proposed methods on a cruise-mode aircraft control system and identify minimum-size actuator and sensor attack sets. By attacking these selected actuators and sensors, the system becomes uncontrollable and unobservable, thus validating the effectiveness of the algorithms.

In-wheel motor vehicles, characterized by their unique feature of independent torque control on all axles, high energy conversion efficiency, rapid motor response times, and the ability to notably decrease curb weight and streamline overall vehicle architecture, represent a significant area of advancement within the electric vehicle domain. This study, with a primary focus on improving vehicle stability management, delves deeply into the mechanics of vehicle instability, deriving the correlation between yaw rate, lateral displacement of the vehicle's center of gravity, and overall handling stability. Based on these findings, a specialized phase plane method for assessing instability was formulated. Utilizing an enhanced version of the dynamic surface sliding mode variable structure control technique, a direct yaw moment controller was designed for the in-wheel motor vehicles, considering the equality constraint posed by longitudinal tire forces, alongside inequality constraints related to torque limitations and road surface frictional conditions. Subsequently, an optimal torque allocator was engineered to distribute the direct yaw moment under these constraints, employing a quadratic programming approach for solving the optimal allocation problem. A comprehensive simulation model was built using MATLAB/Simulink to verify the functionality and benefits of the presented design. Moreover, practical validation was conducted through experimental testing on an in-wheel motor vehicles prototype developed by our research team, thereby affirming the practical feasibility and efficacy of the proposed solution.

This article addresses cluster lag synchronization (CLSyn) and control problems for multiplex and directed network systems (MDNS). Outer matrices (OMs) of the presented model can be directed, with competitive elements, and even not connected. Additionally, inner matrices containing negative elements are used for CLSyn, which significantly improves existing results. First, a concept of CLSyn is proposed encompassing both cluster complete synchronization and cluster anticipated synchronization. Then, we study CLSyn and cluster anticipated synchronization for MDNS under a coupling agreement with time lags by employing a rearranging variable order technique. Moreover, synchronization rules under pinning control and adaptive strength are derived. Furthermore, CLSyn and pinning control conditions are established for multiplex and directed reaction-diffusion network systems (MDRDNS) as an application. Finally, simulations of MDRDNS are provided using a two step Richtmyer algorithm with the Crank-Nicolson scheme, to verify the validity of these obtained results.

This paper proposes a novel impact time control guidance scheme with bounded normal acceleration. A look-angle profile, formulated as a hyperbolic tangent function of range-to-go, enables accurate impact timing using measurable quantities. A numerical algorithm estimates time-to-go based on this profile, adjusting a single parameter online for impact time regulation. The profile is further updated using the predicted intercept point to compensate for disturbances. A second-order sliding mode controller with bounded integral action ensures accurate tracking of the desired profile under input constraints. The approach enables effective impact time control. Simulations validate the method's effectiveness against both stationary and moving targets, demonstrating high terminal accuracy and suitability for cooperative interception missions.

Remaining useful life (RUL) prediction for rolling bearings is critical for maintaining the reliability and safety of rotating machinery. Under complex operating conditions, cross-domain distribution drift and nonlinear degradation dynamics pose considerable challenges for existing methods in achieving both reliable degradation stage division and accurate RUL regression. To address these issues, this paper proposes an integrated online prediction framework based on transfer clustering and Kolmogorov-Arnold Networks (KAN), termed TC-BiKAN. For first prediction time identification, a domain-adaptive fuzzy clustering algorithm with boundary identification (DAFC-BI) is developed to attenuate domain shift and quantify boundary ambiguity through membership differences, thereby enabling robust detection of degradation onset across varying operating conditions. To reduce supervision bias arising from stage misalignment in cross-domain scenarios, a trend-adaptive multistage labeling (TAML) strategy is introduced to reconstruct stage-wise RUL labels by estimating physical degradation rates. To enhance nonlinear regression during accelerated late-stage degradation, a BiKAN predictor is proposed that replaces the conventional fully connected regression head of bidirectional long short-term memory networks with KAN-based learnable spline functions. Experiments on the PHM2012 and XJTU-SY datasets demonstrate that TC-BiKAN achieves competitive prediction accuracy while providing, on average, more conservative RUL estimates than representative baselines, thereby supporting practical deployment for health management under non-stationary operating conditions.

Flexible beveled puncture needles may undergo lateral deflection in soft tissue due to bevel-induced tip forces, which deteriorates targeting accuracy. To accurately characterize the needle motion during insertion, a coupled ordinary and partial differential equations (ODE-PDE) dynamic model is developed based on the variable length Euler-Bernoulli beam (VL-EB) theory. For the axial insertion subsystem, a nonlinear extended state observer (NESO) with a backstepping controller is developed to regulate insertion in the presence of unknown disturbances. To handle boundary uncertainties caused by the needle's unique structure and other factors, a boundary disturbance observer is proposed for the PDE subsystem. An improved boundary sliding mode controller (SMC) is then constructed for the coupled system, where an auxiliary system is employed to accommodate saturation-induced performance degradation. Rigorous Lyapunov analysis establishes the stability and convergence of the closed-loop system. Finally, simulation results demonstrate that the proposed method effectively suppresses needle bending and ensures robust and effective closed-loop performance under disturbances and input saturation.