PurposeThe purpose of this study is to propose and experimentally validate an adaptive robotic rehabilitation system for elbow joint recovery that delivers personalized assistance while accounting for patient-specific biomechanics and real-time muscle fatigue.Design/methodology/approachThe proposed system is based on a structured three-phase rehabilitation protocol consisting of passive parameter extraction, active torque identification and fatigue-aware continuous exercise. A personalized dynamic biomechanical model of the elbow joint is embedded within the control architecture to distinguish passive joint torque from voluntary active torque, enabling accurate patient-specific assistance. Muscle fatigue is estimated online using a K-nearest neighbors (K-NN) classifier trained on time-domain features extracted from surface electromyography (sEMG) signals. The estimated fatigue index is continuously used to adapt the level of robotic assistance, preventing overexertion while encouraging voluntary participation.FindingsExperimental validation was conducted with four orthopedic patients undergoing post-fracture elbow rehabilitation. The system demonstrated accurate joint torque estimation, achieving a root mean square error between 0.17 and 0.29 Nm. The proposed fatigue estimation method achieved a classification accuracy exceeding 93%. Compared to baseline conditions, passive joint torque was reduced by 39%, active torque contribution increased by 85% and elbow range of motion improved by 35%. These preliminary results suggest the feasibility of the proposed approach to provide safe, effective and adaptive rehabilitation assistance.Originality/valueThis work introduces a novel fatigue-aware robotic rehabilitation framework that combines personalized biomechanical modeling with real-time sEMG-based fatigue estimation. The proposed approach enables adaptive, patient-specific control, contributing to the development of intelligent robotic systems for musculoskeletal rehabilitation.

In this study, the problem of glucose regulation for diabetic patients is studied and a new artificial pancreas is developed. In this paper, a new fractional-order dynamic type-3 (T3) fuzzy logic model (FLM) is designed for adaptively glucose-insulin metabolism identification. Based on the identified metabolism, a predictive T3 fuzzy logic controller (T3-FLC) is designed. The designed T3-FLC is learned using a reinforcement learning scheme based on the method of Eligibility Traces (ET) algorithm. ET is an algorithm between Monte Carlo and Temporary Differences Learning. In ET to update the pair of states and agents, the whole chain of states and actions before are considered. The predictor-corrector method of Adams-Bashforth is used to predict the estimated glucose level (output of fractional-order T3-FLM) for a prediction horizon. New Lyapunov adaptations are derived such that the dynamics of the glucose regulation error are stable. Furthermore, a new adaptive parallel supplementary controller is designed to consider the medical restrictions (upper bounds of injected insulin, limitation of sensors and actuators). The comprehensive analyses are provided to show the feasibility and good efficiency of the designed controller.

This paper investigates the problem of simultaneous state, actuator-fault, sensor-fault, and unknown-input estimation for a class of tempered fractional-order Takagi–Sugeno fuzzy systems. Such systems arise in engineering applications involving memory-dependent dynamics, disturbances, and uncertain nonlinear behaviors. An augmented observer-based framework is developed to reconstruct the system states and multiple fault channels within a unified formulation. A transition-variable mechanism is introduced for actuator-fault estimation, while sensor faults and unknown disturbances are embedded into an augmented state representation. Sufficient linear matrix inequality conditions are derived to guarantee boundedness of the composite estimation errors in the general non-ideal case with bounded tempered derivatives. In addition, a special tempered-stationary case is established, ensuring tempered Mittag-Leffler convergence of all estimation errors. Numerical examples demonstrate the feasibility of the proposed design and its effectiveness in accurately tracking states, faults, and disturbances. The proposed framework can support reliable monitoring and fault diagnosis of advanced engineering systems operating under uncertainties and memory effects.

Hybrid energy storage systems (HES) based on the combination of lithium-ion battery (LII-BAT) and supercapacitor (SUP-CAP) are considered for hybrid electric vehicles (HEVs) with extended life cycle (LO-LT), decreased battery stress, and enhanced energy efficiency. Nevertheless, current studies are mainly based on traditional optimizers and do not consider the potential of hybrid approaches that can capture the complex dynamics of HESS. In addition, current methods rely on simulations that have not been experimentally validated and thus do not have real-world applicability. A common limitation of benchmarking techniques is that they are often limited to single-source or simple control-based techniques and do not benchmark to the state of the art. This paper fills in these gaps by introducing a new intelligent energy management system (IEMS) optimized using hybrid Marine Predator Algorithm (MPA) and Dynamic Floating Mechanism (DFM). The proposed MPA-DFM is a combination of adaptive refinement and global exploration to extend the system life and reduce battery power (BAT-POW) stress. System evaluation with an LII-BAT/SUP-CAP HESS prototype shows an increase of 21% of the LO-LT value compared to the previously used methods, which demonstrates a highly sustainable and efficient approach of next-generation HEVs.

This study presents a novel closed-loop bionic hand control system that integrates electromyography (EMG)-driven intent recognition with adaptive neuromuscular electrical stimulation (NMES) to mitigate muscle fatigue and improve user performance. The proposed system features a 3D-printed bionic hand actuated by five independent servomotors, a custom-built electrical stimulator, and a real-time dual-classifier architecture. Muscle fatigue is detected using a Support Vector Machine (SVM) based on frequency-domain EMG features, while handgrip state is classified using a fuzzy logic controller. Experimental trials with 10 neurologically healthy participants demonstrated a 28.6% reduction in muscle fatigue and a 22% improvement in grip force consistency under hybrid control compared to EMG-only operation. The system achieved classification accuracies of 95.4% for fatigue detection and 93% for grip estimation. These results confirm the feasibility of hybrid EMG-NMES systems in enhancing functional performance, stability, and user experience in assistive applications.

Accurate detection of user intention is a critical requirement for intelligent control systems in upper-limb rehabilitation robots. However, electromyography (EMG)-based recognition can degrade significantly under muscle fatigue. To address this limitation, we propose a hybrid EMG-electroencephalography (EEG) control framework that adaptively fuses peripheral (EMG) and central (EEG) biosignals for robust classification of elbow flexion and extension tasks. The system integrates a support vector machine (SVM)-based EMG classifier and a Common Spatial Pattern (CSP)-SVM EEG classifier, combined through a Bayesian fusion strategy whose weights are modulated in real time according to fatigue levels estimated from EMG spectral features via a k-nearest neighbors (k-NN) model. The hybrid framework was deployed on a lightweight robotic rehabilitation platform and evaluated with five healthy participants (3 females, age 26-39). Results show that adaptive fusion significantly outperformed unimodal baselines, achieving 94.5% classification accuracy (vs. 88.5% for EMG-only) with an end-to-end latency below 500 ms. Importantly, the fatigue-aware weighting preserved performance during high-fatigue conditions (91.4% vs. 83.1% for EMG-only), enhancing system robustness during prolonged sessions. These findings demonstrate the feasibility of a scalable, real-time, fatigue-adaptive control strategy with strong potential for clinical stroke rehabilitation and motor recovery.

The erratic nature of cardiac rhythms can precipitate a multitude of pathologies. Consequently, the endeavor to achieve stabilization of the human heartbeat has garnered significant scholarly interest in recent years. In this context, an adaptive nonlinear disturbance compensator (ANDC) strategy has been meticulously developed to ensure the stabilization of cardiac activity. Moreover, a double deep reinforcement learning (DDRL) algorithm has been employed to adaptively calibrate the tunable coefficients of the ANDC controller. To facilitate this, as well as to replicate authentic environmental conditions, a dynamic model of the heart has been constructed utilizing the framework of the Markov Decision Process (MDP). The proposed methodology functions in a closed-loop configuration, wherein the ANDC controller guarantees both stability and disturbance mitigation, while the DDRL agent persistently refines control parameters in accordance with the observed state of the system. Two categories of input signals, namely normal signals and MDP-based stochastic signals, are administered to assess the system’s efficacy under both standard and uncertain conditions. Furthermore, the influence of pathological neural activity is emulated through the introduction of external signals characterized by eight discrete frequency components. Quantitative assessments employing metrics such as peak amplitude, signal energy, and zero-crossing rate are performed for each state of the cardiovascular model. The findings substantiate that the ANDC-DDRL strategy effectively stabilizes cardiac rhythms across diverse conditions, surpassing the performance of conventional baseline methods.

Efficient path planning is challenging for optimizing the trajectory of uncrewed marine vehicles navigating complex environments. However, when the global optimum is zero, path planning optimization encounters a significant challenge, a major shortcoming of the grey wolf optimizer (GWO). This study intentionally integrates multiple approaches to present a comprehensive methodology called fractal-enhanced chaotic GWO (FECGWO) in conjunction with differential evolution (DE) to fill this research gap. This method uses DE to strengthen the local search or exploitation phases, chaotic maps to improve the exploration phase, and fractals to fine-tune the transition between the two phases. In addition to testing against 46 sophisticated benchmark maps, this study carries out practical experimentation over commonly utilized meta-heuristic algorithms to comprehensively evaluate the proposed hybrid model's performance (FECGWO-DE). This thorough evaluation demonstrates notable advancements in unmanned marine vehicle path planning. The evaluation criteria include path length, consistency, time complexity, and success rate. These metrics illustrate the statistical significance of the novel methodology's improvements. The study demonstrates that FECGWO can precisely identify the best routes in given test maps, offering insightful information for developing path planning optimization-especially concerning unmanned marine vehicles.

This paper presents a novel cognitive few-shot learning (CFSL) for the diagnosis of cleft lip and palate and Parkinson’s diseases. The proposed method utilizes computational analysis of paralinguistic features to expedite the diagnostic process. Unlike other methods that rely on complex and fragmented representations, CFSL trains itself to recognize patterns that are easily interpretable by humans. Rather than learning a single, unstructured metric space, CFSL combines the outputs of individual landmark (LM) learners by mapping LMs into semi-formation spaces. In order to assess the effectiveness of CFSL, we conducted a comparative analysis with seven distinct FSL-based models, including momentum contrastive learning for FSL (MCFSL), self-updating FSL (SUFSL), mutual info multi-attention FSL (MAMIFSL), dual class representation FSL (DCRFSL), Improved FSL (IFSL), meta-knowledge for FSL (MKFSL), and prototypical networks (ProtoNet), using three popular datasets, namely GPRS, CIEMPIESS, and PC-GITA. The findings indicate that CFSL demonstrated superior performance compared to the highest-performing baseline frameworks for the 5-shot (5-sh) and 1-shot (1-sh), having an average enhancement of 4.39% and 4.49%, respectively. CFSL demonstrated better performance than the ProtoNet baseline in both 1-sh and 5-sh across all datasets, with an improvement of 12.966% and 11.033%, respectively. In addition, we performed ablation tests to assess the effects of variables such as the density of LMs, the structure of the network, the distance measure used, and the positioning of LMs. The CFSL approach, if adopted in hospitals, has the potential to enhance the precision and efficiency of diagnosis for cleft lip and palate as well as Parkinson’s disease.

This paper introduces a novel approach to enhancing the architecture of deep convolutional neural networks, addressing issues of self-design. The proposed strategy leverages the grey wolf optimizer and the multi-scale fractal chaotic map search scheme as fundamental components to enhance exploration and exploitation, thereby improving the classification task. Several experiments validate the method, demonstrating an impressive 87.37% accuracy across 95 random trials, outperforming 23 state-of-the-art classifiers in the study’s nine datasets. This work underscores the potential of chaotic/fractal and bio-inspired paradigms in advancing neural architecture.

This study presents an AI-enhanced hybrid rehabilitation system that integrates a dual-arm robotic platform with electromyography (EMG)-guided neuromuscular electrical stimulation (NMES) to support upper-limb motor recovery in stroke survivors. The system features a symmetrical robotic arm with real-time anatomical adaptation for bilateral therapy and incorporates a Support Vector Machine (SVM)-based model for continuous muscle fatigue detection using time-frequency features extracted from EMG signals. A ROS2-based architecture enables real-time signal processing, adaptive control, and remote supervision by clinicians. The system dynamically adjusts stimulation parameters based on fatigue classification results, allowing personalized and responsive therapy. Preliminary clinical validation with three post-stroke patients demonstrated a 44% increase in range of motion, 45% enhancement in active torque, and 36% reduction in passive torque. The SVM model achieved a 95% accuracy in fatigue detection, and initial patient results suggest the feasibility and potential benefits of this intelligent, closed-loop rehabilitation approach.

This paper examines finite-time (F-T) observer-based control of fractional-order nonlinear systems with time delay, employing the Takagi-Sugeno fuzzy (T-SF) approach. This study utilizes the conformable fractional derivative (CFD) as its foundational framework. Sufficient conditions are formulated to guarantee that the augmented vector, consisting of the state estimates and the state estimation errors, stays within a ball centered at the origin. This applies not only to specified initial conditions but also to predefined disturbances. These conditions are then reformulated as linear matrix inequalities (LMIs) conditions, which can be efficiently solved using the LMI toolbox. An illustrative example is given to illustrate the proposed result.

The current landscape of medical diagnostics grapples with a critical challenge posed by the limitations of existing meta-learning techniques in interpreting complex representations from limited labeled data, particularly evident in COVID-19 datasets. In response to this gap, our paper introduces a groundbreaking Turning Point (TP) methodology designed to enhance the interpretability of machine learning diagnostics, specifically addressing the shortcomings highlighted in conventional meta-learning approaches. Our Turning Point-based Few-Shot Learning (TPFSL) model goes beyond traditional FSL methods by embracing structured knowledge representation, departing from unstructured metric spaces. We found that our TPFSL model outperformed the state-of-the-art models by an average of 4.50% in 1-shot learning and 4.43% in 5-shot learning after conducting extensive benchmarking on the COVCT, SARSCOV2, and SIRM datasets. Across all datasets studied, TPFSL outperforms the ProtoNet benchmark in 1-shot classification by 12.966% and in 5-shot classification by 11.033%. The importance of TP density, the structure of the network, and placement in improving model performance have been shown by a thorough set of ablation investigations; with its revolutionary TPFSL model, which tackles the shortcomings of current meta-learning methods head-on, COVID-19 diagnostic procedures can be made more accurate and valid in clinical situations.

Recent sensor, communication, and computing technological advancements facilitate smart grid use. The heavy reliance on developed data and communication technology increases the exposure of smart grids to cyberattacks. Existing mitigation in the electricity grid focuses on protecting primary or redundant measurements. These approaches make certain assumptions regarding false data injection (FDI) attacks, which are inadequate and restrictive to cope with cyberattacks. The reliance on communication technology has emphasized the exposure of power systems to FDI assaults that can bypass the current bad data detection (BDD) mechanism. The current study on unobservable FDI attacks (FDIA) reveals the severe threat of secured system operation because these attacks can avoid the BDD method. Thus, a Data-driven learning-based approach helps detect unobservable FDIAs in distribution systems to mitigate these risks. This study presents a new Hybrid Metaheuristics-based Dimensionality Reduction with Deep Learning for FDIA (HMDR-DLFDIA) Detection technique for Enhanced Network Security. The primary objective of the HMDR-DLFDIA technique is to recognize and classify FDIA attacks in the distribution systems. In the HMDR-DLFDIA technique, the min-max scalar is primarily used for the data normalization process. Besides, a hybrid Harris Hawks optimizer with a sine cosine algorithm (hybrid HHO-SCA) is applied for feature selection. For FDIA detection, the HMDR-DLFDIA technique utilizes the stacked autoencoder (SAE) method. To improve the detection outcomes of the SAE model, the gazelle optimization algorithm (GOA) is exploited. A complete set of experiments was organized to highlight the supremacy of the HMDR-DLFDIA method. The comprehensive result analysis stated that the HMDR-DLFDIA technique performed better than existing DL models.

The dynamic and multimodal nature of photovoltaic (PV) systems makes it challenging to examine all solar photovoltaic characteristics. Consequently, this study recommends a recently developed optimization method called the marine predator algorithm (MPA) for developing reliable PV models. In the traditional MPA, the two main search processes are L & eacute;vy flight (LF) and Brownian walk (BW), and the switch across them is unpredictable. This is while the transition between these two mechanisms is naturally continuous and dynamic. To rectify the limitation mentioned above, this research paper presents an innovative, dynamic shift function that effectively modulates the interplay that exists between the BW and LF procedures. By enhancing the changeover pattern between the primary phases of MPA, the suggested dynamic walk substantially boosts the performance of MPA. The dynamic L & eacute;vy-Brownian MPA (DLBMPA) is also made to be resilient in dealing with the parameterization limitations of PV Modeling approaches by using a constraint handling technique. The performance of DLBMPA is tested using ten popular optimization methods. Employing the DLBMPA achieved an average RMSE of 9.7 x 10(- 4) in the parameter estimation across a number of multiple PV models, including the SDM, DDM, and TDM, where out of the ten optimization algorithms experimented, this was statistically significant (p < 0.05) better. In terms of averaged computation time, DLBMPA was 13 ms and still showed high accuracy in dealing with different irradiance and temperature levels. These improvements allow for MBPA to be credited as having a high efficiency when estimating the PV parameters since its speed of convergence and accuracy level surpass the previous techniques used.

Chronic diseases, such as diabetes, cause serious challenges worldwide due to their long-lasting effect on health and quality of life. Diabetes, considered a high glucose level, requires continuous management to avoid complications like kidney failure, low vision, and heart disease. Leveraging state-of-the-art technology, especially IoT devices, shows great potential for optimizing the recognition and management of diabetes. These devices, including smart glucose monitors and wearable sensors, provide real-time data on crucial health metrics, enabling early intervention and proactive monitoring. Furthermore, incorporating deep learning (DL) techniques improves data analysis and recognizes risk factors and subtle patterns related to diabetes. By integrating IoT technology with DL techniques, the healthcare system can empower patients with tools for self-management and develop more conventional approaches for earlier diagnosis and personalized treatment plans, ultimately improving long-term health outcomes and reducing the burden of diabetes. This study develops a new IoT-enabled Driven Early Detection of Chronic Disease using the DL and Dimensionality Reduction (EDCD-DLDR) approach. The EDCD-DLDR method aims to enable IoT devices to collect patient medical data and employ the DL model for earlier diagnosis of diabetes. In the EDCD-DLDR technique, the IoT-based data acquisition process is initially involved, and the collected data gets normalized using Z-score normalization. The EDCD-DLDR technique uses an artificial rabbit optimizer-based feature selection (ARO-FS) approach for dimensionality reduction. In addition, the detection of diabetes is achieved by the attention bidirectional gated recurrent unit (ABiGRU) model. The pelican optimization algorithm (POA) based hyperparameter selection model is included to improve the performance of the ABiGRU network. A comprehensive set of simulations is made to highlight the performance of the EDCD-DLDR method on the Kaggle dataset. The experimental validation of the EDCD-DLDR method underscored a superior accuracy value of 97.14% over existing techniques.

The Robotics Editorial Office retracts the article, “A Newly Designed Wearable Robotic Hand Exoskeleton Controlled by EMG Signals and ROS Embedded Systems” [...]

The utilization of robotic systems in upper limb rehabilitation has shown promising results in aiding individuals with motor impairments. This research introduces an innovative approach to enhance the efficiency and adaptability of upper limb exoskeleton robot-assisted rehabilitation through the development of an optimized stimulation control system (OSCS). The proposed OSCS integrates a fuzzy logic-based pain detection approach designed to accurately assess and respond to the patient’s pain threshold during rehabilitation sessions. By employing fuzzy logic algorithms, the system dynamically adjusts the stimulation levels and control parameters of the exoskeleton, ensuring personalized and optimized rehabilitation protocols. This research conducts comprehensive evaluations, including simulation studies and clinical trials, to validate the OSCS’s efficacy in improving rehabilitation outcomes while prioritizing patient comfort and safety. The findings demonstrate the potential of the OSCS to revolutionize upper limb exoskeleton-assisted rehabilitation by offering a customizable and adaptive framework tailored to individual patient needs, thereby advancing the field of robotic-assisted rehabilitation.

In the midst of a global health crisis, it is of utmost importance for healthcare technologies to possess the capability to regulate and monitor the physiological variables of patients remotely and automatically. The effective control of mean arterial pressure (MAP) in a closed-loop manner is particularly critical for individuals who are critically ill or in the process of recovering from surgical procedures. Within the framework of the present research, an adaptive closed-loop structure has been formulated with the objective of controlling a patient’s MAP through governed administration of the medication sodium nitroprusside (SNP), to attain the desired MAP levels under varying conditions. The proposed closed-loop technique incorporates an intelligent controller known as the active disturbance rejection control (ADRC) with the intention of tracking the desired MAP value, alongside the utilization of continuous action policy gradient (CAPG) for the optimization of the controller’s coefficients. Under the DRL strategy, an actor is responsible for generating policy requests, while a critic assesses the efficacy of the actor’s policy directives. This approach uses gradient descent to train the weight values of both actor and critic networks, and it is dependent on the reward return linked to the MAP fault. Upon comparing the outcomes of the recommended structure with conventional models, numerical simulation results demonstrate the superiority of the proposed system in coping with varying working conditions, key-value fluctuations, and uncertainties, while effectively maintaining the desired mean arterial pressure and drug administration rate.