Digital Twin-Assisted Semi-Federated Learning Framework for Industrial Edge Intelligence

The rapid development of emerging technologies,such as edge intelligence and digital twins,have added momentum towards the development of the Industrial Internet of Things(IIo T).However,the massive amount of data generated by the IIo T,coupled with heterogeneous computation capacity across IIo T devices,and users’data privacy concerns,have posed challenges towards achieving industrial edge intelligence(IEI).To achieve IEI,in this paper,we propose a semi-federated learning framework where a portion of the data with higher privacy is kept locally and a portion of the less private data can be potentially uploaded to the edge server.In addition,we leverage digital twins to overcome the problem of computation capacity heterogeneity of IIo T devices through the mapping of physical entities.We formulate a synchronization latency minimization problem which jointly optimizes edge association and the proportion of uploaded nonprivate data.As the joint problem is NP-hard and combinatorial and taking into account the reality of largescale device training,we develop a multi-agent hybrid action deep reinforcement learning(DRL)algorithm to find the optimal solution.Simulation results show that our proposed DRL algorithm can reduce latency and have a better convergence performance for semi-federated learning compared to benchmark algorithms.

UAV-assisted data collection for wireless sensor networks with dynamic working modes

Wireless Sensor Network(WSN)is a cornerstone of Internet of Things(IoT)and has rich application scenarios.In this work,we consider a heterogeneous WSN whose sensor nodes have a diversity in their Residual Energy(RE).In this work,to protect the sensor nodes with low RE,we investigate dynamic working modes for sensor nodes which are determined by their RE and an introduced energy threshold.Besides,we employ an Unmanned Aerial Vehicle(UAV)to collect the stored data from the heterogeneous WSN.We aim to jointly optimize the cluster head selection,energy threshold and sensor nodes’working mode to minimize the weighted sum of energy con-sumption from the WSN and UAV,subject to the data collection rate constraint.To this end,we propose an efficient search method to search for an optimal energy threshold,and develop a penalty-based successive convex approximation algorithm to select the cluster heads.Then we present a low-complexity iterative approach to solve the joint optimization problem and discuss the implementation procedure.Numerical results justify that our proposed approach is able to reduce the energy consumption of the sensor nodes with low RE significantly and also saves energy for the whole WSN.

Age of Loop Information with Flexible Transmission Enabled Communication and Control Co-Design in Industrial IoT

Precise and low-latency information transmission through communication systems is essential in the Industrial Internet of Things(IIoT).However,in an industrial system,there is always a coupling relationship between the control and communication components.To improve the system's overall performance,exploring the co-design of communication and control systems is crucial.In this work,we propose a new metric±Age of Loop Information with Flexible Transmission(AoLI-FT),which dynamically adjusts the maximum number of uplink(UL)and downlink(DL)transmission rounds,thus enhancing reliability while ensuring timeliness.Our goal is to explore the relationship between AoLI-FT,reliability,and control convergence rate,and to design optimal blocklengths for UL and DL that achieve the desired control convergence rate.To address this issue,we first derive a closed-form expression for the upper bound of AoLI-FT.Subsequently,we establish a relationship between communication reliability and control convergence rates using a Lyapunov-like function.Finally,we introduce an iterative alternating algorithm to determine the optimal communication and control parameters.The numerical results demonstrate the significant performance advantages of our proposed communication and control co-design strategy in terms of latency and control cost.

The microstructural evolution and relaxation strengthening for nano-grained Ni upon low-temperature annealing

The microstructural evolution and relaxation strengthening of nano-grained Ni annealed at a temperature range of 493–553 K were studied by in situ X-ray diffraction technique,transmission electron microscopy,and microhardness evaluation.Upon low-temperature annealing,the rather limited variations of anisotropic grain size and root-mean-square strain,conforming to an exponential relaxation model,yield a consistent activation energy of approximately 0.5 eV,which corresponds to the localized,rapid diffusion of excess vacancies on nonequilibrium surfaces/interfaces and/or defective lattice configurations.Microstructure examinations confirm the grain boundary ordering and excess defect reduction.The relaxation-induced strength enhancement can be attributed to the linear strengthening in the regime of small elastic lattice strains.This study provides an in-depth understanding of low-temperature nanostructural relaxation and its relation to strengthening.

A Systematic Approach for Making 3D-Printed Patient-Specific Implants for Craniomaxillofacial Reconstruction

Craniomaxillofacial reconstruction implants,which are extensively used in head and neck surgery,are conventionally made in standardized forms.During surgery,the implant must be bended manually to match the anatomy of the individual bones.The bending process is time-consuming,especially for inexperienced surgeons.Moreover,repetitive bending may induce undesirable internal stress concentration,resulting in fatigue under masticatory loading in v iv o and causing various complications such as implant fracture,screw loosening,and bone resorption.There have been reports on the use of patient-specific 3D-printed implants for craniomaxillofacial reconstruction,although few reports have considered implant quality.In this paper,we present a systematic approach for making 3D-printed patientspecific surgical implants for craniomaxillofacial reconstruction.The approach consists of three parts:First,an easy-to-use design module is developed using Solidworks®software,which helps surgeons to design the implants and the axillary fixtures for surgery.Design engineers can then carry out the detailed design and use finite-element modeling(FEM)to optimize the design.Second,the fabrication process is carried out in three steps:0 testing the quality of the powder;(2)setting up the appropriate process parameters and running the 3D printing process;and (3)conducting post-processing treatments(i.e.,heat and surface treatments)to ensure the quality and performance of the implant.Third,the operation begins after the final checking of the implant and sterilization.After the surgery,postoperative rehabilitation follow-up can be carried out using our patient tracking software.Following this systematic approach,we have successfully conducted a total of 41 surgical cases.3D-printed patient-specific implants have a number of advantages;in particular,their use reduces surgery time and shortens patient recovery time.Moreover,the presented approach helps to ensure implant quality.

Compound Fault Diagnosis for Rotating Machinery:State-of-the-Art,Challenges,and Opportunities

Compound fault,as a primary failure leading to unexpected downtime of rotating machinery,dramatically increases the difficulty in fault diagnosis.To deal with the difficulty encountered in implementing compound fault diagnosis(CFD),researchers and engineers from industry and academia have made numerous significant breakthroughs in recent years.Admittedly,many systematic surveys focused on fault diagnosis have been conducted by reputable researchers.Nevertheless,previous review articles paid more attention to fault diagnosis with several single or independent faults,resulting in that there is still lacking a comprehensive survey on CFD.Therefore,to fulfill the above requirements,it is necessary to provide an in-depth overview of fault diagnosis methods or algorithms for compound faults of rotating machinery and uncover potential challenges or opportunities that would guide and inspire readers to devote their efforts to promoting fault diagnosis technology more effective and practical.Specifically,the backgrounds,including the related definitions and a new taxonomy of CFD methods,are detailed according to the way of implementing compound fault recognition.Then,the stateof-the-art applications of CFD are overviewed based on relevant publications in the past decades.Finally,the challenges and opportunities associated with implementing CFD are concluded and followed by a conclusion for ending this survey.We believe that this review article can provide a systematic guideline of CFD from different aspects for potential readers and seasoned researchers.

AI-Enabled Monitoring, Diagnosis & Prognosis

The emerging and development of Artificial Intelligence(AI),especially deep learning,has stimulated its application in various engineering domains.Monitoring,diagnosis and prognosis,as the key elements of intelligence maintenance of manufacturing systems in the era of Industry 4.0,has also benefited from the advancement of AI technology.The main objective of this special issue aims at bringing scholars to show their research findings in the field of monitoring,diagnosis and prognosis driven by AI,and promote its application in intelligent maintenance of manufacturing system in China.Ten papers have been selected in this special issue after rigorous review and they represent the latest research outcomes in this active area.

A Morphological Study of Two Young Multipolar Planetary Nebulae

We carry out an optical morphological and infrared spectral study for two young planetary nebulae(PNs)Hen2-158 and Pe 1-1 to understand their complex shapes and dust properties.Hubble Space Telescope optical images reveal that these nebulae have several bipolar-lobed structures and a faint arc with a clear boundary is located at the northwestern side of Pe 1-1.The presence of this arc-shaped structure suggests that the object interacts with its nearby interstellar medium.Spitzer IRS spectroscopic observations of these young nebulae clearly show prominent unidentified infrared emission features and a weak silicate band in Pe 1-1,indicating that Hen 2-158 is a carbonrich nebula and Pe 1-1 has a mixed chemistry dust environment.Furthermore,we construct two three-dimensional models for these PNs to realize their intrinsic structures.The simulated models of the nebulae suggest that multipolar nebulae may be more numerous than we thought.Our analyses of spectral energy distributions for Hen 2-158 and Pe 1-1 show that they have low luminosities and low stellar effective temperatures,suggesting that these nebulae are young PNs.A possible correlation between typical multipolar young PNs and nested nebulae is also discussed.

Modular Soft Robotic Crawlers Based on Fluidic Prestressed Composite Actuators

Soft robotic crawlers have limited payload capacity and crawling speed.This study proposes a high-performance inchworm-like modular robotic crawler based on fluidic prestressed composite(FPC)actuators.The FPC actuator is precurved and a pneumatic source is used to flatten it,requiring no energy cost to maintain the equilibrium curved shape.Pressurizing and depressurizing the actuators generate alternating stretching and bending motions of the actuators,achieving the crawling motion of the robotic crawler.Multi-modal locomotion(crawling,turning,and pipe climbing)is achieved by modular reconfiguration and gait design.An analytical kinematic model is proposed to characterize the quasi-static curvature and step size of a single-module crawler.Multiple configurations of robotic crawlers are fabricated to demonstrate the crawling ability of the proposed design.A set of systematic experiments are set up and conducted to understand how crawler responses vary as a function of FPC prestrains,input pressures,and actuation frequencies.As per the experiments,the maximum carrying load ratio(carrying load divided by robot weight)is found to be 22.32,and the highest crawling velocity is 3.02 body length(BL)per second(392 mm/s).Multi-modal capabilities are demonstrated by reconfiguring three soft crawlers,including a matrix crawler robot crawling in amphibious environments,and an inching crawler turning at an angular velocity of 2/s,as well as earthworm-like crawling robots climbing a 20 inclination slope and pipe.

A Bio-inspired Mutual-hook Strategy for the Soft Finger to Improve Load-bearing Capacity and Grasping Stability

Soft grippers have great potential applications in daily life,since they can compliantly grasp soft and delicate objects.However,the highly elastic fingers of most soft grippers are prone to separate from each other while grasping objects due to their low stiffness,thus reducing the grasping stability and load-bearing capacity.To tackle this problem,inspired from the venus flytrap plant,this work proposes a mutual-hook mechanism to restrain the separation and improve the grasping performance of soft fingers.The novel soft gripper design consists of three modules,a soft finger-cot,two Soft Hook Actuators(SHAs)and two sliding mechanisms.Here,the soft finger-cot covers on the soft finger,increasing the contact area with the target object,two SHAs are fixed to the left and right sides of the finger-cot,and the sliding mechanisms are designed to make SHAs stretch flexibly.Experiments demonstrate that the proposed design can restrain the separation of soft fingers substantially,and the soft fingers with the finger-cots can grasp objects three times heavier than the soft fingers without the proposed design.The proposed design can provide invaluable insights for soft fingers to restrain the separation while grasping,thus improving the grasping stability and the load-bearing capacity.

Attitude Regulation of Flexible Spacecrafts with Unknown Control Directions and Input Disturbances

In this paper,the authors address the attitude regulation problem of uncertain flexible spacecraft with unknown control directions and input disturbances.The major challenges of the problem include the concurrence of the unknown actuation sign and the unknown parameters in both the plant and the external disturbances,along with the impact of vibrations from flexible appendages.To overcome these challenges,the authors transform the conventional mathematical model of a flexible spacecraft to a multivariable strict-feedback normal form and adopt a systematic approach within the framework of nonlinear output regulation.To solve the attitude regulation and disturbance rejection problem,the authors introduce a nonlinear internal model candidate to convert the problem into a stabilization problem for an augmented system.Then,a Nussbaum function-based stabilizer is designed to handle unknown control directions and complete the design.Simulation results are provided to show the effectiveness of the proposed controller.

Multi-modality hierarchical fusion network for lumbar spine segmentation with magnetic resonance images

For the analysis of spinal and disc diseases,automated tissue segmentation of the lumbar spine is vital.Due to the continuous and concentrated location of the target,the abundance of edge features,and individual differences,conventional automatic segmentation methods perform poorly.Since the success of deep learning in the segmentation of medical images has been shown in the past few years,it has been applied to this task in a number of ways.The multi-scale and multi-modal features of lumbar tissues,however,are rarely explored by methodologies of deep learning.Because of the inadequacies in medical images availability,it is crucial to effectively fuse various modes of data collection for model training to alleviate the problem of insufficient samples.In this paper,we propose a novel multi-modality hierarchical fusion network(MHFN)for improving lumbar spine segmentation by learning robust feature representations from multi-modality magnetic resonance images.An adaptive group fusion module(AGFM)is introduced in this paper to fuse features from various modes to extract cross-modality features that could be valuable.Furthermore,to combine features from low to high levels of cross-modality,we design a hierarchical fusion structure based on AGFM.Compared to the other feature fusion methods,AGFM is more effective based on experimental results on multi-modality MR images of the lumbar spine.To further enhance segmentation accuracy,we compare our network with baseline fusion structures.Compared to the baseline fusion structures(input-level:76.27%,layer-level:78.10%,decision-level:79.14%),our network was able to segment fractured vertebrae more accurately(85.05%).

Bioinspired Closed-loop CPG-based Control of a Robot Fish for Obstacle Avoidance and Direction Tracking

This paper presents a study on bioinspired closed-loop Central Pattern Generator(CPG)based control of a robot fish for obstacle avoidance and direction tracking.The biomimetic robot fish is made of a rigid head with a pair of pectoral fins,a wire-driven active body covered with soft skin,and a compliant tail.The CPG model consists of four input parameters:the flapping amplitude,the flapping angular velocity,the flapping offset,and the time ratio between the beat phase and the restore phase in flapping.The robot fish is equipped with three infrared sensors mounted on the left,front and right of the robot fish,as well as an inertial measurement unit,from which the surrounding obstacles and moving direction can be sensed.Based on these sensor signals,the closed-loop CPG-based control can drive the robot fish to avoid obstacles and to track designated directions.Four sets of experiments are presented,including avoiding a static obstacle,avoiding a moving obstacle,tracking a designated direction and tracking a designated direction with an obstacle in the path.The experiment results indicated that the presented control strategy worked well and the robot fish can accomplish the obstacle avoidance and direction tracking effectively.

Trajectory Tracking Control of Euler-Lagrange Systems with ISS-Like Robustness to Actuator Noises

This paper studies global robust tracking of uncertain Euler-Lagrange systems with input disturbances.The authors develop a robust regulation-based approach for the problem.Specifically,by introducing a novel nonlinear internal model,the authors solve global asymptotic trajectory tracking with disturbance rejection of multiple step/sinusoidal signals with unknown amplitudes,frequencies,and phases.Moreover,the authors show that a robustness property to actuator noises can be guaranteed in a sense of strong integral input-to-state stability(iISS).That is,the closed-loop system is not only i ISS but also input-to-state stable(ISS)to small magnitude actuator noises.Furthermore,the authors explore a by-product overparametrized linear regression estimation,coming up with robust estimation of the unknown parameters.Finally,the authors present several numerical examples to illustrate the theoretical results.

Research on Transfer Learning of Vision-based Gesture Recognition

Gesture recognition has been widely used for human-robot interaction.At present,a problem in gesture recognition is that the researchers did not use the learned knowledge in existing domains to discover and recognize gestures in new domains.For each new domain,it is required to collect and annotate a large amount of data,and the training of the algorithm does not benefit from prior knowledge,leading to redundant calculation workload and excessive time investment.To address this problem,the paper proposes a method that could transfer gesture data in different domains.We use a red-green-blue(RGB)Camera to collect images of the gestures,and use Leap Motion to collect the coordinates of 21 joint points of the human hand.Then,we extract a set of novel feature descriptors from two different distributions of data for the study of transfer learning.This paper compares the effects of three classification algorithms,i.e.,support vector machine(SVM),broad learning system(BLS)and deep learning(DL).We also compare learning performances with and without using the joint distribution adaptation(JDA)algorithm.The experimental results show that the proposed method could effectively solve the transfer problem between RGB Camera and Leap Motion.In addition,we found that when using DL to classify the data,excessive training on the source domain may reduce the accuracy of recognition in the target domain.

Superhydrophobic Purple Orchid Leaves:Variation in Surface Morphology During the Vegetation Stages Leading to Diversity in Wettability

Learning hydrophobic phenomena from nature is always a promising approach to design the superhydrophobic surface.Purple orchid leaf which processes superhydrophobicity is an ideal plant model,and through mimicking its structure,the surface with excellent hydrophobicity is able to be obtained.However,the unclear of the diversity in wettability during the different vegetation stages and the absence of its relation to the surface morphology limits the further enhancement of the inspired structure.Here,we analyze the wettability difference as the leaf grows from tender to mature and then to senescent.Combining with the variation of surface morphology and chemical composition,the well-developed micro-scale basic unit bumps with dense nano-scale waxy layer on the surface are proven to be responsible for the best hydrophobicity of the mature leaf.The presence of the undeveloped or damaged micro-nano hierarchical structure reduces the formation of air pockets at the interface,leading to the decrease of the wettability for leaves at other stages.Moreover,by fabricating artificial leaves,the nano-waxy layer is proved to be more effective than that of the micro-bumps on the surface wettability.The results of study are of a great significance for guiding the design and fabrication of plant-inspired bionic superhydrophobic surface.

Precision attitude regulation of uncertain flexible spacecraft

Consider the precision attitude regulation with vibration suppression for an uncertain and disturbed flexible spacecraft.The disturbance at issue is typically any finite superposition of sinusoidal signals with unknown frequencies and step signals of unknown amplitudes.First we show that the conventional mathematical model for flexible spacecrafts is transformable to a multi-input multi-output(MIMO)strict-feedback nonlinear normal form.Particularly it is strongly minimum-phase and has a well-defined uniform vector relative degree.Then it enables us to develop an adaptive internal model-based controller in the framework of adaptive output regulation to solve the problem.It is proved that asymptotic stability can be guaranteed for the attitude regulation task and the vibration of flexible appendages vanishes asymptotically.Hence,the present study explores a new idea for control of flexible spacecraft in virtue of its system structures.

Stretchable and durable HD-sEMG electrodes for accurate recognition of swallowing activities on complex epidermal surfaces

Surface electromyography(sEMG)is widely used in monitoring human health.Nonetheless,it is challenging to capture high-fidelity sEMG recordings in regions with intricate curved surfaces such as the larynx,because regular sEMG electrodes have stiff structures.In this study,we developed a stretchable,high-density sEMG electrode array via layerby-layer printing and lamination.The electrode offered a series of excellent human‒machine interface features,including conformal adhesion to the skin,high electron-to-ion conductivity(and thus lower contact impedance),prolonged environmental adaptability to resist water evaporation,and epidermal biocompatibility.This made the electrode more appropriate than commercial electrodes for long-term wearable,high-fidelity sEMG recording devices at complicated skin interfaces.Systematic in vivo studies were used to investigate its ability to classify swallowing activities,which was accomplished with high accuracy by decoding the sEMG signals from the chin via integration with an ear-mounted wearable system and machine learning algorithms.The results demonstrated the clinical feasibility of the system for noninvasive and comfortable recognition of swallowing motions for comfortable dysphagia rehabilitation.

Multi-Objective Optimization of EV Charging and Discharging for Different Stakeholders

As a dynamic energy storage system,electric vehicles(EV)play important roles in future power grids.In this paper,a model for EV aggregator participation in the electricity market has been built with a focus on the feasibility issue of the model arising from economic interest inconsistencies between different stakeholders:EV owners and aggregator.In the model,the EV aggregator attends day-ahead energy and reserve markets for profit maximization by scheduling charging and discharging behaviors of EVs.This issue exists since different stakeholders have different interests which are not necessarily consistent,e.g.profit maximization leads to increasing EV owners'charging fee.To investigate the economic relationship between the two stakeholders,two multi-objective optimization methods(weighted sum and$\varepsilon$-constraint methods)are proposed to take the aggregator profit and EV owners'charging fee into account in the model.A sensitivity analysis is applied to examine the aggregator profit under different price scenarios,which reveals the internal relationship between EV owners'charging fees and aggregator profit.The proposed EV charging and discharging strategy in this paper could be used to determine the settlement price between the aggregator and owners to ensure the feasibility of participation from both EV owners and stakeholders in electricity markets.

Digital twin-assisted gearbox dynamic model updating toward fault diagnosis

One of the core challenges of intelligent fault diagnosis is that the diagnosis model requires numerous labeled training datasets to achieve satisfactory performance.Generating training data using a virtual model is a potential solution for addressing such a problem,and the construction of a high-fidelity virtual model is fundamental and critical for data generation.In this study,a digital twin-assisted dynamic model updating method for fault diagnosis is thus proposed to improve the fidelity and reliability of a virtual model,which can enhance the generated data quality.First,a virtual model is established to mirror the vibration response of a physical entity using a dynamic modeling method.Second,the modeling method is validated through a frequency analysis of the generated signal.Then,based on the signal similarity indicator,a physical–virtual signal interaction method is proposed to dynamically update the virtual model in which parameter sensitivity analysis,surrogate technique,and optimization algorithm are applied to increase the efficiency during the model updating.Finally,the proposed method is successfully applied to the dynamic model updating of a single-stage helical gearbox;the virtual data generated by this model can be used for gear fault diagnosis.