To enhance aircraft reliability, dissimilar redundant actuation systems composed of a Hydraulic Actuator (HA) and an Electro-Hydrostatic Actuator (EHA) are increasingly adopted for driving aircraft control surfaces. In the active/active operational mode, due to differences in the actuation mechanisms and other factors, the outputs of the two actuators are difficult to be identical, leading to a force fight phenomenon on the control surface, which can damage the surface in severe cases. This paper proposes a Disturbance Rejection Synchronization Control (DRSC) method. The controller comprises a trajectory generator, a tracking differentiator, an Extended State Observer (ESO), and a nonlinear error feedback control law. By using the ESO to estimate the total disturbance in real time and providing feedforward compensation, the controller enables both channels to follow the command signals generated by the trajectory generator. Furthermore, a cross-coupling term is introduced to further reduce the displacement difference between the two actuation channels. The stability of the overall closed-loop system is proven by constructing a Lyapunov equation. A physical test bench is built to compare and validate the control algorithm. Experimental results show that under sinusoidal commands at different frequencies (1–8 Hz), compared with the feedforward PID algorithm, the designed DRSC reduces amplitude attenuation by an average of 34.3%, decreases phase lag by 33.3% in the low-frequency range, and reduces the displacement difference between the two channels by an average of 27.8% across the tested frequencies. The proposed DRSC algorithm effectively suppresses force fight, achieves synchronized motion states between the heterogeneous channels of the dissimilar redundant actuation system, and improves the system's tracking performance and disturbance rejection capability.

Internal leakage caused by wear in hydraulic spool valves represents a critical failure mode that threatens the performance of aircraft hydraulic systems and compromises flight safety. Due to complex operational loads and time-varying material properties, the relationship between wear state and Remaining Useful Life (RUL) is nonlinear. Consequently, accurately modeling this wear remains a significant challenge, as existing research often neglects the coupled effects of material properties, stress conditions, and dynamic lubrication parameters. To address this issue, this study proposes a novel framework integrating physical mechanisms with stochastic processes to enhance wear degradation modelling and RUL prediction. First, a Physics-of-Failure (PoF) model is developed based on Archard’s wear theory, which characterizes tribological behavior at the contact interface and accounts for the effects of lubrication and load conditions. Next, a Gamma process is introduced to model the degradation trajectory, with physical parameters guiding the specification of the time-scale function. A Bayesian expectation-maximization algorithm is employed to estimate and update the model parameters. Finally, a numerical simulation and case study on spool valves are conducted to demonstrate the effectiveness of the proposed model. The cross-validation results confirmed that the introduction of random effects effectively reduces the impact of uncertainty on physics-informed modeling. This study offers a systematic solution to RUL prediction for hydraulic systems.

The external gear pump is susceptible to coupled degradation involving multiple faults under complex operating conditions, which poses considerable challenges to fault diagnosis. However, existing methods generally lack interpretability and the capability to quantify uncertainty in compound fault modeling, which limits their reliability in practical applications. To address this issue, this paper proposes a three-stage compound fault diagnosis framework. The framework sequentially employs a Physics-Informed Neural Network (PINN) to invert key structural parameters under unlabeled conditions, a Bayesian Neural Network (BNN) to model parameter distributions and perform uncertainty reasoning, and a lightweight Multi-Layer Perceptron (MLP) based on the statistical features output by the BNN to accomplish fault classification. Using seven typical fault datasets obtained from an Electro-Hydrostatic Actuator (EHA) experimental platform, the results demonstrate that the proposed method substantially outperforms baseline models such as Transformer and 1D-CNN in terms of overall accuracy and compound fault recognition. It also maintains high robustness under small-sample and high-noise conditions, and exhibits promising generalization capability in partially labeled and unknown fault recognition tasks, providing a feasible path toward transparent and interpretable fault diagnosis under complex working conditions.

Most existing bearing-only formation control methods required that the relative bearings among neighboring agents are measured under a well-known global reference frame for each individual. To remove such constraint, this paper novelly introduces a distributed formation control scheme for quadrotors with only bearing measurement in each vehicle's local reference frame. To this end, firstly, a prescribed-time quaternion-based orientation estimator is proposed for each follower to estimate the leader's orientation without knowledge of the global reference frame. Secondly, a bearing-only formation control law is developed to achieve desired maneuvering formation using relative bearings under local reference frame, wherein a finite-time differentiator is incorporated to remove the need of bearing rate. The convergence is rigorously proven through mathematical derivations. Both comparative simulations and real-world experiments are conducted to validate the effectiveness of the proposed control scheme.

Growing maintenance demand for high-voltage transmission lines, together with the inefficiency and safety risks of manual climbing, has created a strong need for uncrewed systems that can perform precise physical interventions. However, strong wind, line vibration, and floating-base coupling still hinder reliable dual-arm operation in real environments. In this article, we propose a cooperative disturbance-resilient dual-arm control framework for transmission line bolt tightening under persistent small-amplitude disturbances and occasional large-amplitude vibrations induced by wind gusts and line motion. A task-space cooperative impedance outer layer is established to explicitly regulate inter-arm consistency and compliant interaction during operations, while a robust tracking and disturbance estimation inner layer is integrated to suppress model uncertainties, line vibration, and wind disturbances. The proposed controller is deployed on a self-developed aerial transmission-line operation robot. Simulations and laboratory disturbance experiments validate the proposed approach, demonstrating both improved robustness and coordination, as well as the reliable completion of the bolt-tightening mission.

Due to inherent compliance and safety, cable-driven robots have been widely used in the field of physical human-robot interaction. Scenarios that involve physical human-robot interaction impose stringent requirements on the transient performance of robots, which has become a key focus of current research. To tackle this problem, this paper designs a novel control strategy based the second-order Lyapunov stability criterion to predefine the damping ratio, thereby achieving an adjustable transient response. To alleviate the adverse impacts of unknown nonlinearities and external disturbances, neural networks trained via the gradient descent method are adopted in this paper for the online compensation of these nonlinearities. Furthermore, a new filter called dynamic command filter is utilized to resolve the "explosion of complexity" issue inherent in the traditional backstepping design approach, which arises from repeated differentiation of virtual signals. Based on the second-order Lyapunov stability criterion, this paper rigorously proves the practical stability of the closed-loop system and its ability to preserve the predefined damping ratio. Finally, the superiority of the proposed method is validated through comparative simulation experiments.

Welded joints are critical components in engineering structures, yet accurate fatigue life prediction remains challenging due to multiaxial loading complexity and material nonlinearity. Conventional physics-based models often fail to capture intricate load-material interactions, while data-driven approaches demand extensive datasets and lack physical interpretability. To address these limitations, this study introduces CINAS-PINN, a causal inference-based neural architecture search integrated with physics-informed neural networks for welded joint fatigue life prediction. By constructing equivalent tensors, multiaxial load paths are converted into scalar strain energy densities, aiming to capture the physical characteristics of multiaxial loading and provide input support for neural networks. We integrate causal inference with neural architecture search (NAS) in physicsinformed neural networks (PINNs). Based on this, we implemented the PINN structure and optimized the model parameters, addressing the challenge of accurate fatigue life prediction. To overcome issues related to poor model interpretability and low accuracy, we employed a causal graph-constrained architecture, enabling the model to focus on key physical factors. Additionally, a dynamic loss function, adjusted through Granger causality analysis, prioritizes key physical constraints during training, improving model efficiency and physical consistency. Case studies on AISI316L, GH4169, and TC4 alloys demonstrate that CINAS-PINN achieves superior accuracy, reducing prediction errors by more than 30% compared with benchmark methods. The proposed framework offers enhanced physical consistency, robustness, and generalization for fatigue life prediction under complex service conditions.

This paper proposes a coordinated control method for the master cylinder and wheel cylinders of the integrated electro-hydraulic braking system in intelligent vehicles, while also considering fault-tolerant control design in the presence of system faults. First, typical faults of the integrated electro-hydraulic braking system are presented, and a control-oriented model is established. Based on this, control strategies are designed for both the master cylinder and wheel cylinders. The coordinated fault-tolerant control of the master cylinder pressure is realized by using a finite-time disturbance observer and backstepping control method, which enables precise pressure control in the presence of master cylinder faults and external disturbances as well as coordinated control with the wheel cylinder pressure. Adaptive sliding mode control is employed for fault-tolerant control of the wheel cylinder pressure to effectively estimate the fault parameters and realize the precise pressure control of the wheel cylinder with faults. Comparative simulations validate the superiority of the proposed model-based coordinated fault-tolerant control method, controller-in-the-loop and hardware-in-the-loop experiments further confirm the engineering effectiveness of the method.

The structural reliability of the aviation hydraulic rotary joint was studied. The failure behavior of the aviation hydraulic rotary joint was analyzed by the finite element method, and the failure mode of the rotary joint was analyzed. The finite element calculation results were verified by comparison with experimental results. The structural reliability calculation method of aviation hydraulic rotary joint was established. The Kriging model, support vector machine, and extreme learning machine model were used to predict the stress of rotary joint seals, and the influence of different parameters on the reliability of aviation hydraulic rotary joints was investigated. The research results show that Stress failure is the structural failure mode of the sealing assembly of aviation hydraulic rotary joint. The structural reliability of aviation hydraulic rotary joints increases with the increase of material strength and temperature, and decreases with the increase of working pressure and fit clearance, but there is no monotonic relationship between it and the lap size.

This paper investigates the bearing-based fault-tolerant formation control problem for fixed-wing unmanned aerial vehicle (UAV) swarms. To ensure safe and reliable formation maneuvering under loss-of-effectiveness and bias actuator faults as well as multiple system uncertainties, a cascaded control framework with prescribed performance is proposed. First, nonholonomic constraints and lumped uncertainties are incorporated into the UAV dynamics, and a prescribed-time observer is designed to estimate the lumped uncertainties within a user-defined time interval. Second, a hierarchical error transformation is applied to enforce user-specified transient and steady-state performance bounds on the tracking errors. Third, based on the uncertainty estimation and transformed errors, a distributed bearing-based fault-tolerant controller is developed to achieve the desired formation maneuvering behavior. The proposed method is rigorously analyzed for stability, and comparative simulations demonstrate its effectiveness and scalability under challenging conditions, achieving significantly improved performance metrics compared to existing approaches.

This paper presents PRED, a Probabilistic Routing Expert Diffusion framework for automated high-level design of Sigma-Delta Modulators ($\Sigma \Delta$ Ms). The inverse mapping from target specifications to circuit variables is inherently one-to-many and topology-dependent. PRED addresses this by combining (i) an XGBoost-based probabilistic router that retains plausible topology hypotheses and (ii) topology-dedicated conditional diffusion experts that generate diverse candidate design vectors. Candidate solutions are evaluated by SIMSIDES and reranked using Schreier’s figure of merit (FOM $_{\mathrm{s}}$). Experiments on three Switched-Capacitor (SC) $\Sigma \Delta \mathbf{M}$ topologies show that PRED consistently improves normalized FOM $_{\mathrm{s}}$ deviation and targetattainment rate over an ANN point-estimate baseline under the same best-of-10 simulation budget. On the most challenging 12-variable cascade, PRED improves the mean normalized fom ${ }_{\mathrm{s}}$ deviation from -15.84% to +1.32%, highlighting the value of probabilistic topology retention and stochastic design-vector generation.1

Stochastic process models can capture the stochastic dynamics, and they have been widely applied in component degradation modeling. However, the model uncertainty may lead to inaccurate and unreliable prediction of the remaining useful life. In addition, time-varying working conditions and maintenance should be incorporated into the degradation model to accurately determine the component degradation level. To solve the aforementioned problem, a generalized degradation model, based on the nonlinear Tweedie exponential dispersion process (TEDP) considering random effects, covariates, and imperfect maintenance, is developed. First, the nonlinear TEDP model with random parameters is proposed to construct the component degradation model. Considering the time-varying operating environment, the effect of the continuous and discrete covariates on the component degradation is modeled based on the proportional hazards model. The effect of the imperfect maintenance is also incorporated into the component degradation model. Then, mean-field variational inference, stochastic gradient variational inference, and expectation-maximum algorithms are used together to estimate the unknown parameters in order to improve the mathematical interpretability of the developed models. Finally, simulation studies and two real cases on gallium arsenide (GaAs) laser and hydraulic pump are used to demonstrate the effectiveness and validity of the proposed model and algorithm.

This paper proposes a high-level design method for Sigma-Delta modulators ($\Sigma \Delta \mathbf{M s}$) based on fine-tuning large language models (LLMs) with physics-informed chain-of-thought (PI-CoT). First, high-quality training samples are extracted from raw simulation data through physical feasibility filtering and design space deconfliction based on Schreier figure of merit (FOM). Second, the XGBoost-based architecture prediction confidence combined with SHapley Additive exPlanations (SHAP) attribution analysis is used to construct architecture confirmation inference texts. Subsequently, a three-step physical causal chain is integrated to build the PI-CoT dataset covering four typical $\Sigma \Delta \mathbf{M}$ architectures. Furthermore, a $\Sigma \Delta \mathbf{M}$-domain LLM based on the PI-CoT dataset is fine-tuned. Behavioral-level simulation verification across four architectural test samples demonstrates that the proposed method achieves an average FOM satisfaction rate of 78.75%, exhibiting a significant advantage over generalpurpose LLMs and machine learning methods. In addition, the generated design parameters possess genuine physical significance and exhibit strong interpretability.

The Internet of Things (IoT) technology can improve the efficiency of human-machine collaboration (HMC) systems. However, the presence of collaboration mechanisms introduces dynamic human-machine interaction for the mission reliability modeling. This paper develops a cloud edge-based dynamic mission reliability prediction method for HMC systems. Firstly, a dynamic human reliability analysis method is proposed based on the IoT-driven common performance condition and attention curve. Secondly, based on hierarchy and dependency of remotely operated vehicle (ROV) components, a hierarchical dependency Markov process is proposed for component reliability modeling. In addition, IoT-driven proactive maintenance for ROV components is developed. Afterwards, the dynamic human reliability and ROV reliability is corrected based on the impact of collaboration mechanisms, and the mission reliability model of the IoT-enabled HMC system is developed. Finally, a case study of the IoT-enabled HMC system for an underwater search and rescue mission is conducted to validate the proposed method.

This work proposes a prescribed adaptive finite-time control approach that takes the impacts of oil temperature into account in order to address the problem of low tracking control accuracy in oil-immersed electro-hydrostatic actuators (OI-EHAs) under large temperature ranges. Firstly, a dynamic thermal coupling model is established to estimate the unmeasurable oil temperature in real-time, integrating temperature dynamics into the controller architecture to reduce system complexity. Secondly, a finite-time observer (FTO) is designed to effectively suppress composite disturbances arising from temperature prediction bias, parameter perturbations, and external disturbances. By feeding back the estimated disturbance values, the system error converges within a finite time. The tracking error and its dynamic properties are constrained by the introduction of a prescribed performance function (PPF), which guarantees that the error stays within specified bounds. Finally, the backstepping method is employed to integrate the FTO and the PPF, forming a composite controller with temperature-sensitive parameters and mechanisms for compensating uncertainty disturbances. This method ensures the stability of the closed-loop system. Experimental results demonstrate that the proposed controller exhibits excellent performance under various operating conditions.

With the increasing number of individuals with disabilities and an aging population, power wheelchairs have become essential. However, users with upper limb impairments struggle with conventional controls, and customized solutions remain costly. This paper proposes a low-cost, highly functional, and universally applicable intelligent assistive control system for power wheelchairs, tailored for individuals with limbs disabilities. The system integrates multi-sensor information fusion to enhance mobility and environmental awareness. An inertial measurement unit (IMU) on the user's head captures motion states, using an extended Kalman filter (EKF) to generate precise control commands. LiDAR, an odometer, and ultrasonic sensors provide environmental perception, processed via the particle filter Gmapping algorithm and Bayesian theory. Dempster-Shafer (D-S) evidence theory fuses motion intention with environmental data, enabling autonomous path planning. Prototype experiments show that the system ensures accurate wheelchair control while adapting to environmental constraints. The head motion-based directional control achieves 98% average accuracy, with the EKF reducing false triggers by 84%. Steering response time remains below 1.3 s, and obstacle detection reaches 100%. Additionally, the system dynamically optimizes paths, ensuring safe and efficient operation. This research offers a costeffective mobility solution for individuals with limb disabilities, including those with high-level paralysis, significantly improving independence and quality of life.

Electro-hydraulic braking (EHB) systems are safety-critical subsystems in intelligent and electrified vehicles, where actuator and sensor faults can significantly degrade braking performance and pose serious safety risks. To address such problem, this paper proposes a novel fault-tolerant control framework for EHB system. First, the orifice flow principle and Takagi–Sugeno fuzzy modeling techniques are applied to construct an EHB dynamic model with dual faults. Then, by integrating adaptive laws with H_∞ robustness theory, a dual-fault adaptive observer is developed to simultaneously estimate system states and reconstruct actuator and sensor fault in the presence of uncertainties and disturbances. On this basis, a prescribed performance control strategy is designed by barrier Lyapunov functions, to guarantee that the braking pressure tracking behavior meets the predefined transient and steady-state performance constraints. Simulation results under both slowly varying and suddenly fault scenarios demonstrate that the proposed method achieves effective fault estimation and satisfactory pressure tracking performance. Comparative results further confirm significant improvements in robustness and fault-tolerance capabilities of the proposed method over other existing ones.

Maintenance management based on industrial Internet of Things (IoT) can significantly increase the efficiency of complex system maintenance and enable a transformation from reactive response to proactive prevention. However, conventional condition-based maintenance (CBM) tends to be single-component independent maintenance, which cannot perform collaborative optimization of multi-component maintenance and resource scheduling. In addition, due to shared resources and sequential workflows, the dependencies of the maintenance processes between components makes the existing CBM models not applicable. To solve the aforementioned problem, this paper proposes an industrial IoT-driven condition-based maintenance plus (CBM+) method allowing to perform collaborative optimization of proactive maintenance activities for complex systems. Firstly, the real-time remaining useful life of components is predicted based on degradation data monitored by IoT sensors. Secondly, considering the economic-functional-maintenance process dependencies, a multi-component opportunistic maintenance strategy, based on hierarchical Bayesian networks and multi-layer maintenance process networks, is proposed to increase the collaborative maintenance capacity. Afterwards, leveraging an industrial IoT-enhanced field management approach, the maintenance elements (human, equipment, material, method, and environment) are systematically managed to optimize the maintenance efficiency. Furthermore, an industrial IoT-driven CBM+ optimization model considering multiple dependencies and maintenance elements is developed. Finally, a case study of an industrial IoT-driven hydraulic system is conducted to demonstrate the proposed maintenance strategy.

Proportional solenoids are widely used as key power conversion devices in advanced servo systems. The proportional solenoids are highly prone to interturn short circuit faults after prolonged operation due to thermal stress and insulation degradation, leading to gradual performance degradation or even catastrophic system failure. However, predicting the degradation of proportional solenoids within embedded packages remains a significant technical challenge due to limited accessibility. In this article, a novel data-model interactive degradation prediction approach is proposed, which eliminates the requirement for additional sensors or signal injection. A physics-based degradation model is developed to characterize the internal degradation process under thermal stress. To this end, the particle filtering (PF) method is first employed to estimate the unmeasurable states of the solenoid using indirect sensor measurements from the solenoid valve. Then, the expectation-maximization (EM) method is applied to identify the degradation-related hidden parameters in the physics-based degradation model, thereby enabling accurate degradation prediction. Experimental validations are conducted on a proportional solenoid test rig under diverse operating conditions. The experimental results demonstrate that the proposed method significantly improves the degradation prediction accuracy compared with the state-of-the-art algorithms.

Reliability management is crucial for ensuring stable operation of mechatronics components, as well as reducing the downtime and the operating costs. However, the existing degradation models based on Markov properties are not applicable because of the long-term memory of the components. In addition, the degradation of many components in their life cycle exhibits multi-stages, and dependencies exist between different degradation stages. Therefore, this paper proposes a stage-dependent Markov-switching fractional Brownian motion (FBM) model allowing to better capture the characteristics of nonlinearity, randomness, unit-to-unit variability, long-term memory, and dependency of multi-stage degradation. More precisely, the long-term memory of degradation is represented by the FBM process, and random effects are used to describe the unit-to-unit variability. Moreover, a stage-dependent Markov-switching process is proposed for describing the state transitions of multi-stage degradation processes. The working conditions of the different degradation stages are then used to describe the stage impact levels. Furthermore, the unknown parameters of the Markov-switching process and the nonlinear degradation model with FBM are determined based on the two-stage parameter estimation method. Finally, a simulation study and a real case on hydraulic pumps are conducted to demonstrate the high performance of the proposed model.