Search

The research team has developed an information system based on clinical blood cell analysis and designed and implemented highly innovative algorithms. A neural network model was created based on these feature data of the blood cell population. Artificial intelligence algorithms can label susceptible populations for digestive tract cancer with an accuracy rate of over 80 %. A multi universe optimized BP neural network model was implemented based on TCGA data of common immune antigens in clinical laboratories. The working mechanism of this model is to assign values to the parameters of the BP neural network by using the process of searching for the best fitness in multiple universes. This model can predict the five-year survival rate of patients based on immunohistochemical data. Based on these data, an AI algorithm was used to develop a clinical prognostic model with an accuracy rate of over 99 %. The research team used single-cell sequencing data to locate cell subtypes in the features of immunohistochemical data, providing a biological basis for artificial intelligence models. The research team explored the potential biological mechanisms of cancer progression and occurrence based on gastrointestinal neuroendocrine products, and these algorithms have contributed to the prediction of cancer survival and incidence,team invented a simple and efficient algorithm.

In response to the rapid growth of future business and data traffic, the widespread deployment of small base stations (BSs) in 5G networks has emerged as a promising solution, albeit intensifying the network’s energy consumption. Additionally, traditional BSs lack adaptive adjustment of parameter information, posing challenges in delivering satisfactory quality of service (QoS), particularly in the context of highly uneven business distribution. Reducing energy consumption while ensuring that QoS represents a critical challenge. To address this issue, this paper first comprehensively considers the interests of both supply and demand, proposing a service utility measurement method in communication networks to achieve a balance between energy consumption and QoS. Furthermore, this paper integrates cell zooming and sleeping strategies for small BSs, designing a dynamic game algorithm aimed at optimizing service utility in a two-tier heterogeneous network. Through ten distinct scenario simulations, our proposed algorithm demonstrates significant enhancements in service utility while achieving near-optimal optimization results more expeditiously compared to the genetic algorithm (GA).

The dense deployment of Femto Base Stations (FBS) assisting Macro Base Stations (MBS) in a Heterogeneous Network (HetNet) resolves the coverage issue of 5G signal transmission. However, the imprudent layout of FBSs results in extensive energy consumption and increased signal interference among base stations. Regulating the transmission power of each base station in the HetNets through the main controller or MBS is essential to maximize the power efficiency of the entire HetNets while adhering to the constraints of basic signal throughput and fairness. To address this challenge, this paper proposes an Adaptive Acceleration Particle Swarm Optimization (AA-PSO) algorithm. This algorithm dynamically determines the inertia weight based on each particle’s optimal position and the global optimal position, and introduces the concept of time-varying parameters to control the learning rate, thus managing the search range and convergence speed of the particle swarm. The results demonstrate that the AA-PSO algorithm can efficiently determine the optimal transmission power of each base station in the HetNets, reduce interference between MBS and FBSs, as well as among FBSs, and ultimately improve the service efficacy of the entire network.

The ladle shroud is a vital component in the continuous casting process of steelmaking as it supports the production of a wide range of specialty steels. However, it is plagued by a persistent issue of clogging, which not only diminishes the lifetime of the Alumina-Carbon ladle shroud but also severely disrupts the smooth operation of the continuous casting process. Utilizing a numerical simulation technique, the research adopted the method of particle-liquid two-phases coupling, to probe into the clogging mechanism in the ladle shroud from behavior of heat transfer, flow and inclusions. The study indicated that the temperature distribution and flow behavior have a positive effect on the formation of clogging, especially the inclusions. Various factors synergistically contribute on the inner wall, promoting severe clogging at the outlet of the ladle shroud. Finally, a simulation-based clogging mechanism is concluded coupling above factors.

Flexible strain sensors have attracted significant attention due to their critical applications in wearable devices, biological detection, and artificial intelligence. However, achieving both a wide strain range and high sensitivity remains a major challenge in current research. This study aims to develop a novel composite material with a synergistic conductive network to construct high-performance flexible strain sensors. Thermoplastic polyurethane (TPU) nanofiber membranes were first prepared using electrospinning technology, and their surface was modified with polydopamine (PDA) via in-situ polymerization, which significantly enhanced the fibers’ adsorption capacity for conductive materials. Subsequently, carbon nanotubes (CNTs) and carbon black (CB) were coated onto the PDA-modified TPU fibers through ultrasonic anchoring, forming a CB/CNT/PDA/TPU composite with a synergistic conductive network. The results demonstrated that the flexible strain sensor fabricated from this composite material (with a CB-to-CNT mass ratio of 7:3) achieved ultrahigh sensitivity (gauge factor, GF, up to 1063) over a wide strain range (up to 300%), along with a low detection limit (1% strain), fast response and recovery times (137 ms), and exceptional stability and durability. Further evaluations confirmed that this sensor reliably captured biological signals from various joint movements, highlighting its broad application potential in human motion monitoring, human–machine interaction, and soft robotics.

The interpretability of neural networks has garnered significant attention. In the domain of computer vision, gradient-based feature attribution techniques like RectGrad have been proposed to utilize saliency maps to demonstrate feature contributions to predictions. Despite advancements, RectGrad falls short in category discrimination, producing similar saliency maps across categories. This paper pinpoints the ineffectiveness of threshold-based strategies in RectGrad for distinguishing feature gradients and introduces Class activation map Guided BackPropagation (CGBP) to tackle the issue. CGBP leverages class activation maps during backpropagation to enhance gradient selection, achieving consistent improvements across four models (VGG16, VGG19, ResNet50, and ResNet101) on ImageNet’s validation set. Notably, on VGG16, CGBP improves SIC, AIC, and IS scores by 10.3%, 11.5%, and 4.5%, respectively, compared to RectGrad while maintaining competitive DS performance. Moreover, CGBP demonstrates greater sensitivity to model parameter changes than RectGrad, as confirmed by a sanity check. The proposed method has broad applicability in scenarios like model debugging, where it identifies causes of misclassification, and medical image diagnosis, where it enhances user trust by aligning visual explanations with clinical insights.

Insulator and defect detection is a critical technology for the automated inspection of transmission and distribution lines within smart grids. However, the development of a lightweight, real-time detection platform suitable for deployment on drones faces significant challenges. These include the high complexity of existing algorithms, limited availability of UAV images, and persistent issues with false positives and missed detections. To address this issue, this paper proposed a lightweight drone-based insulator defect detection method (LDIDD) that integrates data augmentation and attention mechanisms based on YOLOv8. Firstly, to address the limitations of the existing insulator dataset, data augmentation techniques are developed to enhance the diversity and quantity of samples in the dataset. Secondly, to address the issue of the network model’s complexity hindering its application on UAV equipment, depthwise separable convolution is incorporated for lightweight enhancement within the YOLOv8 algorithm framework. Thirdly, a convolutional block attention mechanism is integrated into the feature extraction module to enhance the detection of small insulator targets in aerial images. The experimental results show that the improved network reduces the computational volume by 46.6% and the mAP stably maintains at 98.3% compared to YOLOv8, which enables the implementation of a lightweight insulator defect network suitable for the UAV equipment side without affecting the detection performance.

Based on the three-dimensional (3D) linear elasticity theory of piezoelectric semiconductor (PS) structures, inspired by the variational principle and the Mindlin plate theory, a two-dimensional (2D) higher-order theory and equations for thin-film devices are established for a rectangular coordinate system, in which Newton’s law (i.e., stress equation of motion), Gauss’s law (i.e., charge equation of electrostatics), Continuity equations (i.e., conservation of charge for holes and electrons), drift–diffusion theory for currents in semiconductors, and unavoidable thermo-deformation-polarization-carrier coupling response in external stimulus field environment are all considered. As a typical application of these equations, the static characteristic analysis of electromechanical fields for the extensional deformation of a PS thin-film device with thermal field excitations is carried out by utilizing established zeroth-order equations and the double trigonometric series solution method. It is revealed that the extensional deformations, electric potential, electron and hole concentration perturbations, and their current densities can be controlled actively via artificially tuning thermal fields of external stimuli. Especially, a higher temperature rise can induce a deeper potential well and a higher potential barrier, which can play a vital role in driving effectively motions and redistributions of electrons and holes. Overall, the derived 2D equations as well as the quantitative results provide us some useful guidelines for investigating the thermal regulation behavior of PS thin-film devices.

With the mutual interaction and dependence of several intelligent services, a crowd intelligence service network has been formed, and a service ecosystem has gradually emerged. Such a development produces an ever-increasing effect on our lives and the functioning of the whole society. These facts call for research on these phenomena with a new theory or perspective, including what a smart society looks like, how it functions and evolves, and where its boundaries and challenges are. However, the research on service ecosystems is distributed in many disciplines and fields, including computer science, artificial intelligence, complex theory, social network, biological ecosystem, and network economics, and there is still no unified research framework. The researchers always have a restricted view of the research process. Under this context, this paper summarizes the research status and future developments of service ecosystems, including their conceptual origin, evolutionary logic, research topic and scale, challenges, and opportunities. We hope to provide a roadmap for the research in this field and promote sound development.

When studying the damage law of reinforced concrete building structures under explosive loading, the direct experimental cost is too high and numerical simulations take a long time. Based on the theoretical analysis, the dimensional analysis model of reinforced concrete members under explosive loading can be used to study the damage law of reinforced concrete members under explosive loading. It provides guidance, reduces the number of tests, improves the efficiency of the test, and has certain research significance. In this paper, a typical reinforced concrete column is taken as the main research object. Based on the dimensional analysis method, the relationship between the damage to the reinforced concrete column and the explosion equivalent and explosion distance under explosion loading is studied. The finite element simulation software LS-DYNA 18.2 is used to determine the function relationship between the disturbance in the column and the proportional explosion distance. The results show that when the proportional explosion distance Z is greater than 0.0693 m/kg1/3, the center disturbance of the blasting surface of the reinforced concrete column has a linear relationship with the reciprocal of the proportional explosion distance. The linear relationship can be used to predict the column’s center disturbance under partial explosion conditions, which provides guidance for studying the damage criterion of reinforced concrete under explosion loading.

In recent years, neural networks have been widely used in natural language processing, especially in sentence similarity modeling. Most of the previous studies focused on the current sentence, ignoring the commonsense knowledge related to the current sentence in the task of sentence similarity modeling. Commonsense knowledge can be remarkably useful for understanding the semantics of sentences. CK-Encoder, which can effectively acquire commonsense knowledge to improve the performance of sentence similarity modeling, is proposed in this paper. Specifically, the model first generates a commonsense knowledge graph of the input sentence and calculates this graph by using the graph convolution network. In addition, CKER, a framework combining CK-Encoder and sentence encoder, is introduced. Experiments on two sentence similarity tasks have demonstrated that CK-Encoder can effectively acquire commonsense knowledge to improve the capability of a model to understand sentences.

Abstract It is difficult to remove and eliminate bubbles in microchannels in many devices used in various biomedical fields, such as those needed for microfluidic immunoassays, point-of-care testing, and cell biology evaluations. Accumulated bubbles are associated with a number of negative outcomes, including a decrease in device sensitivity, inaccuracy of analysis results, and even functional failure. Xylem conduits of angiosperm have the ability to remove bubbles in obstructed conduits. Inspired by such an embolism repair mechanism, this paper proposes a bioinspired bubble removal method, which exhibits a prominent ability to dissolve bubbles continuously within a large range of flow rates (2 µL/min–850 µL/min) while retaining the stability and continuity of the flow without auxiliary equipment. Such a method also shows significant bubble removal stability in dealing with Newtonian liquids and non-Newtonian fluids, especially with high viscosity (6.76 Pa s) and low velocity (152 nL/min). Such advantages associated with the proposed bioinspired method reveal promising application prospects in macro/microfluidic fields ranging from 3D printing, implantable devices, virus detection, and biomedical fluid processing to microscale reactor operation and beyond.

As a natural environmental medium, soil has a wide range of sources and is often used as a material for building houses. It can also be used to construct simple protective structures in actual battlefield environments. In order to study the protective effect of sand on prefabricated fragments in practical situations, the authors conducted experiments on spherical tungsten alloy prefabricated fragments penetrating into sand using sand as the target medium. Natural sand (naturally generated rock particles with a particle size less than 4.75 mm) was selected as the sand, and the initial velocities of the fragments were 689 m/s~1761 m/s (fragment diameter 6 mm) and 596 m/s~1325 m/s (fragment diameter 11 mm), respectively. Based on the residual velocity of fragments and experimental phenomena in experimental data, the authors used LS-DYNA software to numerically simulate and compare the residual velocity of fragments and obtained a numerical model for sandy soil media and a calculation formula for the residual velocity of spherical fragments. Based on modeling equivalence in practical environments, the authors studied the impact of different attack angles on fragment trajectory deviation and found that fragments have the highest deviation values at 20° and 70° attack angles. They also analyzed the soil boundary effect within a small scale range, and the effect was significant when the vertical distance was less than 40 mm. The penetration resistance of different thicknesses of soil was calculated, and the maximum soil thickness under prefabricated fragment penetration at different speeds was obtained. The effective protective size of the soil under actual conditions was about 550 mm. This experiment provides a reference for the construction parameters of simple soil defense structures in actual environments.

PurposeThe behavioral decision-making of digital-self is one of the important research contents of the network of crowd intelligence. The factors and mechanisms that affect decision-making have attracted the attention of many researchers. Among the factors that influence decision-making, the mind of digital-self plays an important role. Exploring the influence mechanism of digital-selfs’ mind on decision-making is helpful to understand the behaviors of the crowd intelligence network and improve the transaction efficiency in the network of CrowdIntell.Design/methodology/approachIn this paper, the authors use behavioral pattern perception layer, multi-aspect perception layer and memory network enhancement layer to adaptively explore the mind of a digital-self and generate the mental representation of a digital-self from three aspects including external behavior, multi-aspect factors of the mind and memory units. The authors use the mental representations to assist behavioral decision-making.FindingsThe evaluation in real-world open data sets shows that the proposed method can model the mind and verify the influence of the mind on the behavioral decisions, and its performance is better than the universal baseline methods for modeling user interest.Originality/valueIn general, the authors use the behaviors of the digital-self to mine and explore its mind, which is used to assist the digital-self to make decisions and promote the transaction in the network of CrowdIntell. This work is one of the early attempts, which uses neural networks to model the mental representation of digital-self.

Deep-learning methods provide a promising approach for measuring in-vivo knee joint motion from fast registration of two-dimensional (2D) to three-dimensional (3D) data with a broad range of capture. However, if there are insufficient data for training, the data-driven approach will fail. We propose a feature-based transfer-learning method to extract features from fluoroscopic images. With three subjects and fewer than 100 pairs of real fluoroscopic images, we achieved a mean registration success rate of up to 40%. The proposed method provides a promising solution, using a learning-based registration method when only a limited number of real fluoroscopic images is available.

In a side-pumped laser module, we simulate the pumping power distribution across the cross-section of Nd:YAG rods at varying diameters. The coupled rate equations with the spatial overlap efficiency for a four-level actively Q-switched side-pumped laser are considered complete. After a Q-switched pulse terminates, the expressions of laser output parameters are derived. We assess the single pulse energy, pulse width, and peak power using the analysis function. The 200 Hz Q-switched side-pumped Nd:YAG laser with a peak power of 103 kW and beam quality factors M2 = 1.67 is developed to the validity of theoretical models.

Purpose – The purpose of this paper is to detect abnormal data of complex and sophisticated industrial equipment with sensors quickly and accurately. Due to the rapid development of the Internet of Things, more and more equipment is equipped with sensors, especially more complex and sophisticated industrial equipment is installed with a large number of sensors. A large amount of monitoring data is quickly collected to monitor the operation of the equipment. How to detect abnormal data quickly and accurately has become a challenge. Design/methodology/approach – In this paper, the authors propose an approach called Multiple Group Correlation-based Anomaly Detection (MGCAD), which can detect equipment anomaly quickly and accurately. The single-point anomaly degree of equipment and the correlation of each kind of data sequence are modeled by using multi-group correlation probability model (a probability distribution model which is helpful to the anomaly detection of equipment), and the anomaly detection of equipment is realized. Findings – The simulation data set experiments based on real data show that MGCAD has better performance than existing methods in processing multiple monitoring data sequences. Originality/value – The MGCAD method can detect abnormal data quickly and accurately, promote the intelligent level of smart articles and ultimately help to project the real world into cyber space in CrowdIntell Network.

ABSTRACT: Magneto-mechanical coupling vibration arises in the in-vessel components of Tokamak devices especially during the plasma disruption. Strong electromagnetic forces cause the structures to vibrate while the motion in turn changes the distribution of the electromagnetic field. To ensure the Tokamak devices operating in a designed state, numerical analysis on the coupling vibration is of great importance. This paper introduces two numerical methods for the magneto-mechanical coupling problems. The coupling term of velocity and magnetic flux density is manipulated in both Eulerian and Lagrangian description, which brings much simplification in numerical implementation. Corresponding numerical codes have been developed and applied to the dynamic simulation of a test module in J-TEXT and the vacuum vessel of HL-2M during plasma disruptions. The results reveal the evident influence of the magnetic stiffness and magnetic damping effects on the vibration behavior of the in-vessel structures. Finally, to deal with the halo current injection problem, a numerical scheme is described and validated which can simulate the distribution of the halo current without complicated manipulations. Keywords: Magneto-mechanical coupling analysis, In-vessel component, Plasma disruption, Halo current, Numerical simulation

Purpose – The purpose of this paper is to build cyber-physical-psychological ternary fusion crowd intelligence network and realize comprehensive, real, correct and synchronous projection in cyber–physical–psychological ternary fusion system. Since the network of crowd intelligence is the future interconnected network system that takes on the features of large scale, openness and self-organization. The Digital-selfs in the network of crowd intelligence interact and cooperate with each other to finish transactions and achieve co-evolution eventually. Design/methodology/approach – To realize comprehensive, real, correct and synchronous projection between cyber–physical–psychological ternary fusion system, the authors propose the rules and methods of projection from real world to the CrowdIntell Network. They build the mental model of the Digital-self including structure model and behavior model in four aspects: identity, provision, demand and connection, thus forming a theoretical mental model framework of Digital-self. Findings – The mental model is excepted to lay a foundation for the theory of modeling and simulation in the research of crowd science and engineering. Originality/value – This paper is the first one to propose the mental model framework and projection rules and methods of Digital-selfs in network of crowd intelligence, which lays a solid foundation for the theory of modeling, simulation, intelligent transactions, evolution and stability of CrowdIntell Network system, thus promoting the development of crowd science and engineering.

In order to study the influence of the forest environment on the killing power of prefabricated fragments, the relationship between the wood medium in the forest and the residual velocity of prefabricated fragments and a simple calculation model were obtained. The impact test of 300 mm thick pine target plate was carried out with 6 mm and 11 mm diameter tungsten alloy spherical fragments. The LS-DYNA finite element modeling of wood was carried out and compared with the experimental parameters, which verified the reliability of the numerical simulation method on the residual velocity of fragments. Based on this, a simple mathematical calculation method for the forest environment is constructed to calculate the penetration track length of fragments in the forest environment. Combined with the model and the formula for calculating the residual velocity of wood, the impact of forests on fragments within a certain radius can be evaluated. According to the characteristics of the forest environment, the boundary effects of the multi-layer spacing targets and trees were further studied. The research shows that the wood reduces the fragment power mainly by affecting the penetration length. The influence of forest density, tree diameter, and other parameters on the fragment velocity attenuation in the forest environment is analyzed. Using this method, the influence of forest environmental parameters on fragment dispersion parameters can be simply evaluated.

An explosion inside a cabin will converge at the corners to form high-pressure areas, significantly impacting the destruction of a bulkhead structure. This paper investigates shock wave convergence characteristics at the corners when the explosive detonates at the center of the cabin, based on a combination of the wall reflection law for shock waves and a numerical simulation method. The parameter K represents the aspect ratio of the cabin structure. This study shows that when 1 ≤ K ≤ 1.19, the high pressure at the corner is caused by the superposition of Mach waves along both wall surfaces. However, for the initial shock wave, when 1.2 < K ≤ 2, the high pressure is caused by the superposition of Mach waves along the longer wall surface and regular reflected waves on the shorter wall surface; when 2 < K, the cause are Mach waves along the longer wall surface and the corresponding positive reflection on the shorter wall surface. The influence of K on the range for the high-pressure region at the corner is also analyzed, the functional relationship between the range of the high-pressure area and K is given, and the universality is verified.

During the process of emergency landing, the energy of the impact in the vertical direction is dissipated through the deformation of the structure. The landing load is transferred to the spine of the occupant through the landing gear, the fuselage and finally the seat. This can cause serious damage to the human body. Since the seat is in direct contact with the human body, the energy absorption capacity of the seat is the most direct manifestation of the Crashworthiness of the helicopter. The solutions proposed in the paper may reduce the impact of the seat on the spine of the occupant during the collision by optimizing the key energy-absorbing structure. Taking the seat of the H135 helicopter as a case, the mechanical model of the human–seat coupling system, which is based on the theories of energy methods and structural mechanics, is simplified. Additionally, the simulation was considered reasonable by comparing the simulation results with the results of crashworthiness tests. On the basis of the above, the optimization of the key energy-absorbing structure of the seat was completed by using Latin hypercube sampling and the kriging model. Overall, the optimization effectively enhanced the crashworthiness of this helicopter seat and provided a solution for the passive safety design of aviation seats.

Fault detection in the landing gear retraction/extension hydraulic system is difficult due to uncertainties in component parameters and sensor measurement values. This work lies in the introduction of linear fractional transformation technology and uncertainty analysis theory for the construction of the diagnostic bond graph of the landing gear retraction/extension hydraulic system. Thus, interval analytical redundancy relations can be derived as well as fault signature matrices. By using the fault signature matrix, existing faults can be detected and isolated preliminary. Furthermore, interval analytical redundancy relations can be used to detect system faults in detail. The analysis results of the failure cases of the internal and external leakage of the actuator and landing gear selector valve reversing stuck show that compared to the traditional analytical redundancy relations, this method takes into account the negative factors of uncertainty, so it can effectively reduce missed diagnosis and misdiagnosis; compared to the traditional absolute diagnostic threshold, the interval diagnostic threshold is more accurate and sensitive.

Contrapose the highly integrated, multiple types of faults and complex working conditions of aircraft electro hydrostatic actuator (EHA), to effectively identify its typical faults, we propose a fault diagnosis method based on fusion convolutional neural networks (FCNN). First, the aircraft EHA fault data is encoded by gram angle difference field (GADF) to obtain the fault feature images. Then we build a FCNN model that integrates the 1DCNN and 2DCNN, where the original 1D fault data is the input of the 1DCNN model, and the feature images obtained by GADF transformation are used as the input of 2DCNN. Multiple convolution and pooling operations are performed on each of these inputs to extract the features. Next these feature vectors are spliced in the convergence layer, and the fully connected layers and the Softmax layers are finally used to attain the classification of aircraft EHA faults. Furthermore, the multi-strategy hybrid particle swarm optimization (MSPSO) algorithm is applied to optimize the FCNN to obtain a better combination of FCNN hyperparameters; MSPSO incorporates various strategies, including an initialization strategy based on homogenization and randomization, and an adaptive inertia weighting strategy, etc. The experimental result indicates that the FCNN model optimized by MSPSO achieves an accuracy of 96.86% for identifying typical faults of the aircraft EHA, respectively, higher than the 1DCNN and the 2DCNN by about 16.5% and 5.7%. By comparing with LeNet-5, GoogleNet, AlexNet, and GRU, the FCNN model presents the highest diagnostic accuracy, less time in training and testing. The comprehensive performance of the proposed model is demonstrated to be much stronger. Additionally, the FCNN model improved by MSPSO has a higher accuracy rate when compared to PSO.

Underground space is vulnerable to large deformation influenced by the abnormal stress induced by the bearing coal pillar. A numerical simulation model was established to determine the redistribution of the abnormal stress induced by the mining activities. The double-yield model, the strain softening model, the interface model, and the Mohr–Coulomb model were determined to simulate the gob compaction effect, the pillar strength reduction effect, the structure plane discontinuity effect, and the rock mechanical behavior, respectively. This numerical simulation model is reliable to predict the abnormal stress under the bearing coal pillar by the comparison of the abutment stress from this model and the existing theoretical model as well as the entry roof surface displacement from this model and the field measuring method. The results from the validated numerical model indicate that the abnormal stress including stress concentration coefficient, stress gradient, and lateral pressure coefficient will redistribute to another state that the stress concentration coefficient and stress gradient increase gradually and then decrease, and the lateral pressure coefficient decreases gradually, then increases, and finally decreases sharply with the approach of the mining working face. Their maximum increasing rates are calculated as 121.05%, 198.56%, and 236.82%, respectively. This predicted mining-induced redistribution of the abnormal stress is available for designing the underground entry layout in the determination of the entry position, determination of the driving operation time, mining disturbing range warning, and the prediction of the strengthening support area.

The health of the landing gear retraction/extension(R/E) hydraulic system may be assessed using fuzzy comprehensive evaluation (FCE), however the traditional FCE method depends solely on human assessment by specialists, which is excessively subjective. To address the issue of excessive human subjective variables in the assessment, an improved FCE model based on enhanced risk coefficient is provided, which includes four consideration indexes: failure probability, failure severity, failure detection difficulty, and failure repair difficulty. To reduce subjective human judgment errors entirely due to expert experience, the improved FCE takes into account the likelihood of failure using a statistical method, the severity of failure using a fault simulation analysis based on the LMS Imagine.Lab AMESim simulation platform, and the difficulty of fault detection and repair using the aircraft manufacturer’s professional maintenance information. As part of the evaluation model, the range of health assessment values and accompanying treatment methods are included, making it easier to implement on a daily basis in aircraft maintenance. As a final step, the simulation is evaluated, and the simulated faults are calculated.

Pavement maintenance prioritization considering both quality and cost is an important decision-making problem. In this paper, the actual pavement condition index of city roads was calculated using municipal patrol data. A linear optimization model that maximized maintenance quality with limited maintenance costs and a multi-objective optimization model that maximized maintenance quality while minimizing maintenance costs were developed based on the pavement condition index. These models were subsequently employed in making decisions for actual pavement maintenance using sequential quadratic programming and a genetic algorithm. The results showed that the proposed decision-making models could effectively address actual pavement maintenance issues. Additionally, the results of the single-objective linear optimization model verified that the multiobjective optimization model was accurate. Thus, they could provide optimal pavement maintenance schemes for roads according to actual pavement conditions. The reliability of the models was investigated by analyzing their assumptions and validating their optimization results. Furthermore, their applicability in pavement operation-related decision making and preventive maintenance for roads of different grades was confirmed.

The high in situ stress can significantly affect the blast-induced rock fragmentation and cause difficulties in deep mining and civil engineering where the drilling and blasting technique is applied. In this study, the rock crack propagation induced by blasting under in situ stress is first analyzed theoretically, and then a numerical model with a decoupled charge in LS-DYNA is developed to reveal how the initiation and propagation of rock cracks are under high in situ stress. Through simulation, the mechanisms of blast-induced crack evolution under various hydrostatic pressures and nonhydrostatic pressures are investigated, and the differences in crack evolution with specific decoupling coefficients are compared. According to the simulation, three damage zones, i.e., the crushed zone, the nonlinear fracture zone, and the radial crack propagation zone, are formed, and the radial crack evolution is greatly suppressed by the high in situ stress which has no much influence on the crack propagation in the crushed and the nonlinear fracture zones. The velocity of crack propagation is slightly reduced, and the process of crack propagation is stopped early when the rock is subjected to high in situ stress. Furthermore, the numerical analysis indicates that the crack grows preferentially in the direction of maximum principal stress, and the radial crack propagation is predominantly controlled by the preloaded pressure, which is vertical to the crack propagation direction. Based on the numerical results, it is suggested that the optimal decoupling coefficients for rock cracking are 2.65, 1.87, 1.37, and 1.22 for 0, 10, 20, and 30 MPa, respectively. This study provides not only an analysis of the rock crack evolution under high in situ stress but also a reference for resolving excavation difficulties in deep mining.

Lipreading aims to recognize sentences being spoken by a talking face. In recent years, the lipreading method has achieved a high level of accuracy on large datasets and made breakthrough progress. However, lipreading is still far from being solved, and existing methods tend to have high error rates on the wild data and have the defects of disappearing training gradient and slow convergence. To overcome these problems, we proposed an efficient end-to-end sentence-level lipreading model, using an encoder based on a 3D convolutional network, ResNet50, Temporal Convolutional Network (TCN), and a CTC objective function as the decoder. More importantly, the proposed architecture incorporates TCN as a feature learner to decode feature. It can partly eliminate the defects of RNN (LSTM, GRU) gradient disappearance and insufficient performance, and this yields notable performance improvement as well as faster convergence. Experiments show that the training and convergence speed are 50% faster than the state-of-the-art method, and improved accuracy by 2.4% on the GRID dataset.

This paper presents the trajectory tracking control of a flexible-joint manipulator driven by permanent magnet synchronous motor (PMSM). Combining the PMSM electrical equation and mechanical equation of robotic manipulators, a novel smooth switching control scheme is proposed. Firstly, the position loop controller of the system is designed, with an improved hierarchical sliding mode control (IHSMC) algorithm proposed to further the response speed of the system, additionally, a robust interconnection and damping assignment passivity-based controller (IDA-PBC) is designed to improve the steady state performance of the system. Then, the IDA-PBC control strategy is leveraged to design the current loop controller of the system, on which basis a hybrid controller with smooth switching is designed. Furthermore, Gaussian function is applied as the smooth switching function of the hybrid controller to promote the switching performance. As a result, the hybrid controller has both good dynamic and steady performance. The simulation results verify the effectiveness of the algorithms.

To enhance the control performance of permanent-magnet synchronous motor (PMSM) drive systems, achieving high-precision motion control, a generalized predictive control (GPC) method based on a novel extended-state observer (ESO) is investigated for the speed control of PMSM. In this paper, the controller design consists of two steps. Firstly, according to the continuous time model of PMSM, using the Taylor series expansion, the predictive value of motor speed in finite time is derived, and the single-loop speed controller by combining the speed loop and q - axis current loop is obtained through the defined cost function. The structure of the controller is simple compared to other forms. Secondly, considering the uncertainty of the load torque and the model uncertainties, a novel extended-state observer is designed to compute the actual torque, and the observed value is introduced to the GPC controller. The simulation and experimental results show that the proposed GPC+ESO control method has superior dynamic performance and strong robustness.