Recommendation Information Systems(RIS)are pivotal in helping users in swiftly locating desired content from the vast amount of information available on the Internet.Graph Convolution Network(GCN)algorithms have been ...Recommendation Information Systems(RIS)are pivotal in helping users in swiftly locating desired content from the vast amount of information available on the Internet.Graph Convolution Network(GCN)algorithms have been employed to implement the RIS efficiently.However,the GCN algorithm faces limitations in terms of performance enhancement owing to the due to the embedding value-vanishing problem that occurs during the learning process.To address this issue,we propose a Weighted Forwarding method using the GCN(WF-GCN)algorithm.The proposed method involves multiplying the embedding results with different weights for each hop layer during graph learning.By applying the WF-GCN algorithm,which adjusts weights for each hop layer before forwarding to the next,nodes with many neighbors achieve higher embedding values.This approach facilitates the learning of more hop layers within the GCN framework.The efficacy of the WF-GCN was demonstrated through its application to various datasets.In the MovieLens dataset,the implementation of WF-GCN in LightGCN resulted in significant performance improvements,with recall and NDCG increasing by up to+163.64%and+132.04%,respectively.Similarly,in the Last.FM dataset,LightGCN using WF-GCN enhanced with WF-GCN showed substantial improvements,with the recall and NDCG metrics rising by up to+174.40%and+169.95%,respectively.Furthermore,the application of WF-GCN to Self-supervised Graph Learning(SGL)and Simple Graph Contrastive Learning(SimGCL)also demonstrated notable enhancements in both recall and NDCG across these datasets.展开更多
This paper presents a new approach to synthesize admittance function polynomials and coupling matrices for coupled resonator filters. The N + 2 transversal network method is applied to study a coupled resonator f...This paper presents a new approach to synthesize admittance function polynomials and coupling matrices for coupled resonator filters. The N + 2 transversal network method is applied to study a coupled resonator filter. This method allowed us to determine the polynomials of the reflection and transmission coefficients. A study is made for a 4 poles filter with 2 transmission zeros between the N + 2 transversal network method and the one found in the literature. A MATLAB code was designed for the numerical simulation of these coefficients for the 6, 8, and 10 pole filter with 4 transmission zeros.展开更多
It is a significant task to predict the solar activity for space weather and solar physics. All kinds of approaches have been used to forecast solar activities, and they have been applied to many areas such as the sol...It is a significant task to predict the solar activity for space weather and solar physics. All kinds of approaches have been used to forecast solar activities, and they have been applied to many areas such as the solar dynamo of simulation and space mission planning. In this paper, we employ the long-shortterm memory(LSTM) and neural network autoregression(NNAR) deep learning methods to predict the upcoming 25 th solar cycle using the sunspot area(SSA) data during the period of May 1874 to December2020. Our results show that the 25 th solar cycle will be 55% stronger than Solar Cycle 24 with a maximum sunspot area of 3115±401 and the cycle reaching its peak in October 2022 by using the LSTM method. It also shows that deep learning algorithms perform better than the other commonly used methods and have high application value.展开更多
Tsunami ran-up height is a significant parameter for dimensions of coastal structures. In the present study, tsunami run-up heights are estimated by three different Artificial Neural Network (ANN) models, i.e. Feed ...Tsunami ran-up height is a significant parameter for dimensions of coastal structures. In the present study, tsunami run-up heights are estimated by three different Artificial Neural Network (ANN) models, i.e. Feed Forward Back Propagation (FFBP), Radial Basis Functions (RBF) and Generalized Regression Neural Network (GRNN). As the input for the ANN configuration, the wave height (H) values are employed. It is shown that the tsunami ran-up height values are closely approximated with all of the applied ANN methods. The ANN estimations are slightly superior to those of the empirical equation. It can be seen that the ANN applications are especially significant in the absence of adequate number of laboratory experiments. The results also prove that the available experiment data set can be extended with ANN simulations. This may be helpful to decrease the burden of the experimental studies and to supply results for comparisons.展开更多
<div style="text-align:justify;"> Recent days, heart ailments assume a fundamental role in the world. The physician gives different name for heart disease, for example, cardiovascular failure, heart fa...<div style="text-align:justify;"> Recent days, heart ailments assume a fundamental role in the world. The physician gives different name for heart disease, for example, cardiovascular failure, heart failure and so on. Among the automated techniques to discover the coronary illness, this research work uses Named Entity Recognition (NER) algorithm to discover the equivalent words for the coronary illness content to mine the significance in clinical reports and different applications. The Heart sickness text information given by the physician is taken for the preprocessing and changes the text information to the ideal meaning, at that point the resultant text data taken as input for the prediction of heart disease. This experimental work utilizes the NER to discover the equivalent words of the coronary illness text data and currently uses the two strategies namely Optimal Deep Learning and Whale Optimization which are consolidated and proposed another strategy Optimal Deep Neural Network (ODNN) for predicting the illness. For the prediction, weights and ranges of the patient affected information by means of chosen attributes are picked for the experiment. The outcome is then characterized with the Deep Neural Network and Artificial Neural Network to discover the accuracy of the algorithms. The performance of the ODNN is assessed by means for classification methods, for example, precision, recall and f-measure values. </div>展开更多
Three recent breakthroughs due to AI in arts and science serve as motivation:An award winning digital image,protein folding,fast matrix multiplication.Many recent developments in artificial neural networks,particularl...Three recent breakthroughs due to AI in arts and science serve as motivation:An award winning digital image,protein folding,fast matrix multiplication.Many recent developments in artificial neural networks,particularly deep learning(DL),applied and relevant to computational mechanics(solid,fluids,finite-element technology)are reviewed in detail.Both hybrid and pure machine learning(ML)methods are discussed.Hybrid methods combine traditional PDE discretizations with ML methods either(1)to help model complex nonlinear constitutive relations,(2)to nonlinearly reduce the model order for efficient simulation(turbulence),or(3)to accelerate the simulation by predicting certain components in the traditional integration methods.Here,methods(1)and(2)relied on Long-Short-Term Memory(LSTM)architecture,with method(3)relying on convolutional neural networks.Pure ML methods to solve(nonlinear)PDEs are represented by Physics-Informed Neural network(PINN)methods,which could be combined with attention mechanism to address discontinuous solutions.Both LSTM and attention architectures,together with modern and generalized classic optimizers to include stochasticity for DL networks,are extensively reviewed.Kernel machines,including Gaussian processes,are provided to sufficient depth for more advanced works such as shallow networks with infinite width.Not only addressing experts,readers are assumed familiar with computational mechanics,but not with DL,whose concepts and applications are built up from the basics,aiming at bringing first-time learners quickly to the forefront of research.History and limitations of AI are recounted and discussed,with particular attention at pointing out misstatements or misconceptions of the classics,even in well-known references.Positioning and pointing control of a large-deformable beam is given as an example.展开更多
Based on the first arrival P and S data of 4 625 regional earthquakes recorded at 174 stations dispersed in the Yunnan and Sichuan Provinces, the 3-D velocity structure of crust and upper mantle in the region is deter...Based on the first arrival P and S data of 4 625 regional earthquakes recorded at 174 stations dispersed in the Yunnan and Sichuan Provinces, the 3-D velocity structure of crust and upper mantle in the region is determined, incorporating with previous deep geophysical data. In the upper crust, a positive anomaly velocity zone exists in the Sichuan basin, whereas a negative anomaly velocity zone exists in the western Sichuan plateau. The boundary between the positive and negative anomaly zones is the Longmenshan fault zone. The images of lower crust and upper mantle in the Longmenshan fault, Xianshuihe fault, Honghe fault and others show the characteristic of tectonic boundary, indicating that the faults likely penetrate the Moho discontinuity. The negative velocity anomalies at the depth of 50 km in the Tengchong volcanic area and the Panxi tectonic zone appear to be associated with the temperature and composition variations in the upper mantle. The overall features of the crustal and the upper mantle structures in the SichuanYunnan region are the lower average velocity in both crust and uppermost mantle, the large crustal thickness variations, and the existence of high conductivity layer in the crust or/and upper mantle, and higher geothermal value. All these features are closely related to the collision between the India and the Asia plates. The crustal velocity in the SichuanYunnan rhombic block generally shows normal value or positive anomaly, while the negative anomaly exists in the area along the large strike-slip faults as the block boundary. It is conducive to the crustal block side-pressing out along the faults. In the major seismic zones, the seismicity is relative to the negative anomaly velocity. Most strong earthquakes occurred in the upper-mid crust with positive anomaly or normal velocity, where the negative anomaly zone generally exists below.展开更多
High-speed trains often use temperature sensors to monitor the motion state of bearings.However,the temperature of bearings can be affected by factors such as weather and faults.Therefore,it is necessary to analyze in...High-speed trains often use temperature sensors to monitor the motion state of bearings.However,the temperature of bearings can be affected by factors such as weather and faults.Therefore,it is necessary to analyze in detail the relationship between the bearing temperature and influencing factors.In this study,a dynamics model of the axle box bearing of high-speed trains is established.The model can obtain the contact force between the rollers and raceway and its change law when the bearing contains outer-ring,inner-ring,and rolling-element faults.Based on the model,a thermal network method is introduced to study the temperature field distribution of the axle box bearings of high-speed trains.In this model,the heat generation,conduction,and dispersion of the isothermal nodes can be solved.The results show that the temperature of the contact point between the outer-ring raceway and rolling-elements is the highest.The relationships between the node temperature and the speed,fault type,and fault size are analyzed,finding that the higher the speed,the higher the node temperature.Under different fault types,the node temperature first increases and then decreases as the fault size increases.The effectiveness of the model is demonstrated using the actual temperature data of a high-speed train.This study proposes a thermal network model that can predict the temperature of each component of the bearings on a high-speed train under various speed and fault conditions.展开更多
Based on the characteristics of fractures in naturally fractured reservoir and a discrete-fracture model, a fracture network numerical well test model is developed. Bottom hole pressure response curves and the pressur...Based on the characteristics of fractures in naturally fractured reservoir and a discrete-fracture model, a fracture network numerical well test model is developed. Bottom hole pressure response curves and the pressure field are obtained by solving the model equations with the finite-element method. By analyzing bottom hole pressure curves and the fluid flow in the pressure field, seven flow stages can be recognized on the curves. An upscaling method is developed to compare with the dual-porosity model (DPM). The comparisons results show that the DPM overestimates the inter-porosity coefficient ), and the storage factor w. The analysis results show that fracture conductivity plays a leading role in the fluid flow. Matrix permeability influences the beginning time of flow from the matrix to fractures. Fractures density is another important parameter controlling the flow. The fracture linear flow is hidden under the large fracture density. The pressure propagation is slower in the direction of larger fracture density.展开更多
In this paper, we made a new breakthrough, which proposes a new recursion–transform(RT) method with potential parameters to evaluate the nodal potential in arbitrary resistor networks. For the first time, we found ...In this paper, we made a new breakthrough, which proposes a new recursion–transform(RT) method with potential parameters to evaluate the nodal potential in arbitrary resistor networks. For the first time, we found the exact potential formulae of arbitrary m × n cobweb and fan networks by the RT method, and the potential formulae of infinite and semi-infinite networks are derived. As applications, a series of interesting corollaries of potential formulae are given by using the general formula, the equivalent resistance formula is deduced by using the potential formula, and we find a new trigonometric identity by comparing two equivalence results with different forms.展开更多
Based on Newton’s second law and the thermal network method,a mechanical thermal coupling model of the bearing rotor system of high-speed trains is established to study the interaction between the bearing vibration a...Based on Newton’s second law and the thermal network method,a mechanical thermal coupling model of the bearing rotor system of high-speed trains is established to study the interaction between the bearing vibration and temperature.The influence of lubrication on the vibration and temperature characteristics of the system is considered in the model,and the real-time relationship between them is built up by using the transient temperature field model.After considering the lubrication,the bearing outer ring vibration acceleration and node temperature considering grease are lower,which shows the necessity of adding the lubrication model.The corresponding experiments for characteristics of vibration and temperature of the model are respectively conducted.In the envelope spectrum obtained from the simulation signal and the experimental signal,the frequency values corresponding to the peaks are close to the theoretical calculation results,and the error is very small.In the three stages of the temperature characteristic experiment,the node temperature change of the simulation model is consistent with the experiment.The good agreement between simulation and experiments proves the effectiveness of the model.By studying the influence of the bearing angular and fault size on the system node temperature,as well as the change law of bearing lubrication characteristics and temperature,it is found that the worse the working condition is,the higher the temperature is.When the ambient temperature is low,the viscosity of grease increases,and the oil film becomes thicker,which increases the sliding probability of the rolling element,thus affecting the normal operation of the bearing,which explains the phenomenon of frequent bearing faults of high-speed trains in the low-temperature area of Northeast China.Further analysis shows that faults often occur in the early stage of train operation in the low-temperature environment.展开更多
The reflection characteristics of gird structures are calculated by the spatial network method in the case of normal incidence plane electromagnetic wave. The numerical result shows that the grid panels without electr...The reflection characteristics of gird structures are calculated by the spatial network method in the case of normal incidence plane electromagnetic wave. The numerical result shows that the grid panels without electromagnetic wave absorbing foams are not ideal. However, the absorbing ability can be achieved as low as -25 dBsm from 8 GHz to 12 GHz when the grid cells are filled with foam absorbers. Also it is noted from computation that the foam filled grid structures with larger cell size, higher and thinner ribs will improve the absorbing abilities, which illustrates that they can be used as the effective light-weight stealth structures for aeronautical application.展开更多
An exact and a numerical solutions to the problem of a steady mixed convective MHD flow of an incompressible viscous electrically conducting fluid past an infinite vertical porous plate with combined heat and mass tra...An exact and a numerical solutions to the problem of a steady mixed convective MHD flow of an incompressible viscous electrically conducting fluid past an infinite vertical porous plate with combined heat and mass transfer are presented.A uniform magnetic field is assumed to be applied transversely to the direction of the flow with the consideration of the induced magnetic field with viscous and magnetic dissipations of energy.The porous plate is subjected to a constant suction velocity as well as a uniform mixed stream velocity.The governing equations are solved by the perturbation technique and a numerical method.The analytical expressions for the velocity field,the temperature field,the induced magnetic field,the skin-friction,and the rate of heat transfer at the plate are obtained.The numerical results are demonstrated graphically for various values of the parameters involved in the problem.The effects of the Hartmann number,the chemical reaction parameter,the magnetic Prandtl number,and the other parameters involved in the velocity field,the temperature field,the concentration field,and the induced magnetic field from the plate to the fluid are discussed.An increase in the heat source/sink or the Eckert number is found to strongly enhance the fluid velocity values.The induced magnetic field along the x-direction increases with the increase in the Hartmann number,the magnetic Prandtl number,the heat source/sink,and the viscous dissipation.It is found that the flow velocity,the fluid temperature,and the induced magnetic field decrease with the increase in the destructive chemical reaction.Applications of the study arise in the thermal plasma reactor modelling,the electromagnetic induction,the magnetohydrodynamic transport phenomena in chromatographic systems,and the magnetic field control of materials processing.展开更多
In this paper, based on simulated annealing a new method to rank important nodes in complex networks is presented.First, the concept of an importance sequence(IS) to describe the relative importance of nodes in comp...In this paper, based on simulated annealing a new method to rank important nodes in complex networks is presented.First, the concept of an importance sequence(IS) to describe the relative importance of nodes in complex networks is defined. Then, a measure used to evaluate the reasonability of an IS is designed. By comparing an IS and the measure of its reasonability to a state of complex networks and the energy of the state, respectively, the method finds the ground state of complex networks by simulated annealing. In other words, the method can construct a most reasonable IS. The results of experiments on real and artificial networks show that this ranking method not only is effective but also can be applied to different kinds of complex networks.展开更多
The prediction of solitary wave run-up has important practical significance in coastal and ocean engineering, but the calculation precision is limited in the existing models. For improving the calculation precision, a...The prediction of solitary wave run-up has important practical significance in coastal and ocean engineering, but the calculation precision is limited in the existing models. For improving the calculation precision, a solitary wave run-up calculation model was established based on artificial neural networks in this study. A back-propagation (BP) network with one hidden layer was adopted and modified with the additional momentum method and the auto-adjusting learning factor. The model was applied to calculation of solitary wave run-up. The correlation coefficients between the neural network model results and the experimental values was 0.996 5. By comparison with the correlation coefficient of 0.963 5, between the Synolakis formula calculation results and the experimental values, it is concluded that the neural network model is an effective method for calculation and analysis of solitary wave ran-up.展开更多
Superconductive properties for oxides were predicted by artificial neural network (ANN) method with structural and chemical parameters as inputs. The predicted properties include superconductivity for oxides, distribu...Superconductive properties for oxides were predicted by artificial neural network (ANN) method with structural and chemical parameters as inputs. The predicted properties include superconductivity for oxides, distributed ranges of the superconductive transition temperature (Tc) for complex oxides, and Tc values for cuprate superconductors. The calculated results indicated that the adjusted ANN can be used to predict superconductive properties for unknown oxides.展开更多
This paper presents an artificial neural network(ANN)-based response surface method that can be used to predict the failure probability of c-φslopes with spatially variable soil.In this method,the Latin hypercube s...This paper presents an artificial neural network(ANN)-based response surface method that can be used to predict the failure probability of c-φslopes with spatially variable soil.In this method,the Latin hypercube sampling technique is adopted to generate input datasets for establishing an ANN model;the random finite element method is then utilized to calculate the corresponding output datasets considering the spatial variability of soil properties;and finally,an ANN model is trained to construct the response surface of failure probability and obtain an approximate function that incorporates the relevant variables.The results of the illustrated example indicate that the proposed method provides credible and accurate estimations of failure probability.As a result,the obtained approximate function can be used as an alternative to the specific analysis process in c-φslope reliability analyses.展开更多
Aiming at obtaining high power density of surface-mounted and interior permanent magnet synchronous motor(SIPMSM),it is important to accurately calculate the temperature field distribution of SIPMSM,and a magnetic-the...Aiming at obtaining high power density of surface-mounted and interior permanent magnet synchronous motor(SIPMSM),it is important to accurately calculate the temperature field distribution of SIPMSM,and a magnetic-thermal coupling method is proposed.The magnetic-thermal coupling mechanism is analyzed.The thermal network model and finite element model are built by this method,respectively.The effects of power frequency on iron losses and temperature fields are analyzed by the magnetic-thermal coupling finite element model under the condition of rated load,and the relationship between the load and temperature field is researched under the condition of the synchronous speed.In addition,the equivalent thermal network model is used to verify the magnetic-thermal coupling method.Then the temperatures of various nodes are obtained.The results show that there are advantages in both computational efficiency and accuracy for the proposed coupling method,which can be applied to other permanent magnet motors with complex structures.展开更多
BP neural networks is used to mid-term earthquake prediction in this paper. Some usual prediction parameters of seismology are used as the import units of neural networks. And the export units of neural networks is ca...BP neural networks is used to mid-term earthquake prediction in this paper. Some usual prediction parameters of seismology are used as the import units of neural networks. And the export units of neural networks is called as the character parameter W_0 describing enhancement of seismicity. We applied this method to space scanning of North China. The result shows that the mid-term anomalous zone of W_0-value usually appeared obviously around the future epicenter 1~3 years before earthquake. It is effective to mid-term prediction.展开更多
Anisotropy of the strength and deformation behaviors of fractured rock masses is a crucial issue for design and stability assessments of rock engineering structures, due mainly to the non-uniform and non- regular geom...Anisotropy of the strength and deformation behaviors of fractured rock masses is a crucial issue for design and stability assessments of rock engineering structures, due mainly to the non-uniform and non- regular geometries of the fracture systems. However, no adequate efforts have been made to study this issue due to the current practical impossibility of laboratory tests with samples of large volumes con- taining many fractures, and the difficulty for controlling reliable initial and boundary conditions for large-scale in situ tests. Therefore, a reliable numerical predicting approach for evaluating anisotropy of fractured rock masses is needed. The objective of this study is to systematically investigate anisotropy of strength and deformability of fractured rocks, which has not been conducted in the past, using a nu- merical modeling method. A series of realistic two-dimensional (2D) discrete fracture network (DFN) models were established based on site investigation data, which were then loaded in different directions, using the code UDEC of discrete element method (DEM), with changing confining pressures. Numerical results show that strength envelopes and elastic deformability parameters of tested numerical models are significantly anisotropic, and vary with changing axial loading and confining pressures. The results indicate that for design and safety assessments of rock engineering projects, the directional variations of strength and deformability of the fractured rock mass concerned must be treated properly with respect to the directions of in situ stresses. Traditional practice for simply positioning axial orientation of tunnels in association with principal stress directions only may not be adequate for safety requirements. Outstanding issues of the present study and su^zestions for future study are also oresented.展开更多
基金This work was supported by the Kyonggi University Research Grant 2022.
文摘Recommendation Information Systems(RIS)are pivotal in helping users in swiftly locating desired content from the vast amount of information available on the Internet.Graph Convolution Network(GCN)algorithms have been employed to implement the RIS efficiently.However,the GCN algorithm faces limitations in terms of performance enhancement owing to the due to the embedding value-vanishing problem that occurs during the learning process.To address this issue,we propose a Weighted Forwarding method using the GCN(WF-GCN)algorithm.The proposed method involves multiplying the embedding results with different weights for each hop layer during graph learning.By applying the WF-GCN algorithm,which adjusts weights for each hop layer before forwarding to the next,nodes with many neighbors achieve higher embedding values.This approach facilitates the learning of more hop layers within the GCN framework.The efficacy of the WF-GCN was demonstrated through its application to various datasets.In the MovieLens dataset,the implementation of WF-GCN in LightGCN resulted in significant performance improvements,with recall and NDCG increasing by up to+163.64%and+132.04%,respectively.Similarly,in the Last.FM dataset,LightGCN using WF-GCN enhanced with WF-GCN showed substantial improvements,with the recall and NDCG metrics rising by up to+174.40%and+169.95%,respectively.Furthermore,the application of WF-GCN to Self-supervised Graph Learning(SGL)and Simple Graph Contrastive Learning(SimGCL)also demonstrated notable enhancements in both recall and NDCG across these datasets.
文摘This paper presents a new approach to synthesize admittance function polynomials and coupling matrices for coupled resonator filters. The N + 2 transversal network method is applied to study a coupled resonator filter. This method allowed us to determine the polynomials of the reflection and transmission coefficients. A study is made for a 4 poles filter with 2 transmission zeros between the N + 2 transversal network method and the one found in the literature. A MATLAB code was designed for the numerical simulation of these coefficients for the 6, 8, and 10 pole filter with 4 transmission zeros.
基金supported by the National Natural Science Foundation of China under Grant numbers U2031202,U1731124 and U1531247the special foundation work of the Ministry of Science and Technology of the People’s Republic of China under Grant number 2014FY120300the 13th Five-year Informatization Plan of Chinese Academy of Sciences under Grant number XXH13505-04。
文摘It is a significant task to predict the solar activity for space weather and solar physics. All kinds of approaches have been used to forecast solar activities, and they have been applied to many areas such as the solar dynamo of simulation and space mission planning. In this paper, we employ the long-shortterm memory(LSTM) and neural network autoregression(NNAR) deep learning methods to predict the upcoming 25 th solar cycle using the sunspot area(SSA) data during the period of May 1874 to December2020. Our results show that the 25 th solar cycle will be 55% stronger than Solar Cycle 24 with a maximum sunspot area of 3115±401 and the cycle reaching its peak in October 2022 by using the LSTM method. It also shows that deep learning algorithms perform better than the other commonly used methods and have high application value.
文摘Tsunami ran-up height is a significant parameter for dimensions of coastal structures. In the present study, tsunami run-up heights are estimated by three different Artificial Neural Network (ANN) models, i.e. Feed Forward Back Propagation (FFBP), Radial Basis Functions (RBF) and Generalized Regression Neural Network (GRNN). As the input for the ANN configuration, the wave height (H) values are employed. It is shown that the tsunami ran-up height values are closely approximated with all of the applied ANN methods. The ANN estimations are slightly superior to those of the empirical equation. It can be seen that the ANN applications are especially significant in the absence of adequate number of laboratory experiments. The results also prove that the available experiment data set can be extended with ANN simulations. This may be helpful to decrease the burden of the experimental studies and to supply results for comparisons.
文摘<div style="text-align:justify;"> Recent days, heart ailments assume a fundamental role in the world. The physician gives different name for heart disease, for example, cardiovascular failure, heart failure and so on. Among the automated techniques to discover the coronary illness, this research work uses Named Entity Recognition (NER) algorithm to discover the equivalent words for the coronary illness content to mine the significance in clinical reports and different applications. The Heart sickness text information given by the physician is taken for the preprocessing and changes the text information to the ideal meaning, at that point the resultant text data taken as input for the prediction of heart disease. This experimental work utilizes the NER to discover the equivalent words of the coronary illness text data and currently uses the two strategies namely Optimal Deep Learning and Whale Optimization which are consolidated and proposed another strategy Optimal Deep Neural Network (ODNN) for predicting the illness. For the prediction, weights and ranges of the patient affected information by means of chosen attributes are picked for the experiment. The outcome is then characterized with the Deep Neural Network and Artificial Neural Network to discover the accuracy of the algorithms. The performance of the ODNN is assessed by means for classification methods, for example, precision, recall and f-measure values. </div>
文摘Three recent breakthroughs due to AI in arts and science serve as motivation:An award winning digital image,protein folding,fast matrix multiplication.Many recent developments in artificial neural networks,particularly deep learning(DL),applied and relevant to computational mechanics(solid,fluids,finite-element technology)are reviewed in detail.Both hybrid and pure machine learning(ML)methods are discussed.Hybrid methods combine traditional PDE discretizations with ML methods either(1)to help model complex nonlinear constitutive relations,(2)to nonlinearly reduce the model order for efficient simulation(turbulence),or(3)to accelerate the simulation by predicting certain components in the traditional integration methods.Here,methods(1)and(2)relied on Long-Short-Term Memory(LSTM)architecture,with method(3)relying on convolutional neural networks.Pure ML methods to solve(nonlinear)PDEs are represented by Physics-Informed Neural network(PINN)methods,which could be combined with attention mechanism to address discontinuous solutions.Both LSTM and attention architectures,together with modern and generalized classic optimizers to include stochasticity for DL networks,are extensively reviewed.Kernel machines,including Gaussian processes,are provided to sufficient depth for more advanced works such as shallow networks with infinite width.Not only addressing experts,readers are assumed familiar with computational mechanics,but not with DL,whose concepts and applications are built up from the basics,aiming at bringing first-time learners quickly to the forefront of research.History and limitations of AI are recounted and discussed,with particular attention at pointing out misstatements or misconceptions of the classics,even in well-known references.Positioning and pointing control of a large-deformable beam is given as an example.
基金Foundation item: National Scientific and Technological Development Program (95-973-02-02) the Climb Program (95-S-05-01) of National Scientific and Technological Ministry of China and the State Natural Sciences Foundation of China (49874021).
文摘Based on the first arrival P and S data of 4 625 regional earthquakes recorded at 174 stations dispersed in the Yunnan and Sichuan Provinces, the 3-D velocity structure of crust and upper mantle in the region is determined, incorporating with previous deep geophysical data. In the upper crust, a positive anomaly velocity zone exists in the Sichuan basin, whereas a negative anomaly velocity zone exists in the western Sichuan plateau. The boundary between the positive and negative anomaly zones is the Longmenshan fault zone. The images of lower crust and upper mantle in the Longmenshan fault, Xianshuihe fault, Honghe fault and others show the characteristic of tectonic boundary, indicating that the faults likely penetrate the Moho discontinuity. The negative velocity anomalies at the depth of 50 km in the Tengchong volcanic area and the Panxi tectonic zone appear to be associated with the temperature and composition variations in the upper mantle. The overall features of the crustal and the upper mantle structures in the SichuanYunnan region are the lower average velocity in both crust and uppermost mantle, the large crustal thickness variations, and the existence of high conductivity layer in the crust or/and upper mantle, and higher geothermal value. All these features are closely related to the collision between the India and the Asia plates. The crustal velocity in the SichuanYunnan rhombic block generally shows normal value or positive anomaly, while the negative anomaly exists in the area along the large strike-slip faults as the block boundary. It is conducive to the crustal block side-pressing out along the faults. In the major seismic zones, the seismicity is relative to the negative anomaly velocity. Most strong earthquakes occurred in the upper-mid crust with positive anomaly or normal velocity, where the negative anomaly zone generally exists below.
基金National Key R&D Program(Grant No.2020YFB2007700),National Natural Science Foundation of China(Grant Nos.11790282,12032017,12002221 and 11872256)S&T Program of Hebei(Grant No.20310803D)+1 种基金Natural Science Foundation of Hebei Province(Grant No.A2020210028)State Foundation for Studying Abroad.
文摘High-speed trains often use temperature sensors to monitor the motion state of bearings.However,the temperature of bearings can be affected by factors such as weather and faults.Therefore,it is necessary to analyze in detail the relationship between the bearing temperature and influencing factors.In this study,a dynamics model of the axle box bearing of high-speed trains is established.The model can obtain the contact force between the rollers and raceway and its change law when the bearing contains outer-ring,inner-ring,and rolling-element faults.Based on the model,a thermal network method is introduced to study the temperature field distribution of the axle box bearings of high-speed trains.In this model,the heat generation,conduction,and dispersion of the isothermal nodes can be solved.The results show that the temperature of the contact point between the outer-ring raceway and rolling-elements is the highest.The relationships between the node temperature and the speed,fault type,and fault size are analyzed,finding that the higher the speed,the higher the node temperature.Under different fault types,the node temperature first increases and then decreases as the fault size increases.The effectiveness of the model is demonstrated using the actual temperature data of a high-speed train.This study proposes a thermal network model that can predict the temperature of each component of the bearings on a high-speed train under various speed and fault conditions.
基金Project supported by the National Natural Science Foundation of China(No.5140232)the National Science and Technology Major Project(No.2011ZX05038003)the China Postdoctoral Science Foundation(No.2014M561074)
文摘Based on the characteristics of fractures in naturally fractured reservoir and a discrete-fracture model, a fracture network numerical well test model is developed. Bottom hole pressure response curves and the pressure field are obtained by solving the model equations with the finite-element method. By analyzing bottom hole pressure curves and the fluid flow in the pressure field, seven flow stages can be recognized on the curves. An upscaling method is developed to compare with the dual-porosity model (DPM). The comparisons results show that the DPM overestimates the inter-porosity coefficient ), and the storage factor w. The analysis results show that fracture conductivity plays a leading role in the fluid flow. Matrix permeability influences the beginning time of flow from the matrix to fractures. Fractures density is another important parameter controlling the flow. The fracture linear flow is hidden under the large fracture density. The pressure propagation is slower in the direction of larger fracture density.
基金Project supported by the Natural Science Foundation of Jiangsu Province,China(Grant No.BK20161278)
文摘In this paper, we made a new breakthrough, which proposes a new recursion–transform(RT) method with potential parameters to evaluate the nodal potential in arbitrary resistor networks. For the first time, we found the exact potential formulae of arbitrary m × n cobweb and fan networks by the RT method, and the potential formulae of infinite and semi-infinite networks are derived. As applications, a series of interesting corollaries of potential formulae are given by using the general formula, the equivalent resistance formula is deduced by using the potential formula, and we find a new trigonometric identity by comparing two equivalence results with different forms.
基金supported by the National Key R&D Program of China(No.2020YFB2007700)the National Natural Science Foundation of China(Nos.11790282,12032017,12002221,and 11872256)+1 种基金the S&T Program of Hebei Province of China(No.20310803D)the Natural Science Foundation of Hebei Province of China(No.A2020210028)。
文摘Based on Newton’s second law and the thermal network method,a mechanical thermal coupling model of the bearing rotor system of high-speed trains is established to study the interaction between the bearing vibration and temperature.The influence of lubrication on the vibration and temperature characteristics of the system is considered in the model,and the real-time relationship between them is built up by using the transient temperature field model.After considering the lubrication,the bearing outer ring vibration acceleration and node temperature considering grease are lower,which shows the necessity of adding the lubrication model.The corresponding experiments for characteristics of vibration and temperature of the model are respectively conducted.In the envelope spectrum obtained from the simulation signal and the experimental signal,the frequency values corresponding to the peaks are close to the theoretical calculation results,and the error is very small.In the three stages of the temperature characteristic experiment,the node temperature change of the simulation model is consistent with the experiment.The good agreement between simulation and experiments proves the effectiveness of the model.By studying the influence of the bearing angular and fault size on the system node temperature,as well as the change law of bearing lubrication characteristics and temperature,it is found that the worse the working condition is,the higher the temperature is.When the ambient temperature is low,the viscosity of grease increases,and the oil film becomes thicker,which increases the sliding probability of the rolling element,thus affecting the normal operation of the bearing,which explains the phenomenon of frequent bearing faults of high-speed trains in the low-temperature area of Northeast China.Further analysis shows that faults often occur in the early stage of train operation in the low-temperature environment.
基金Funded by the National Natural Science Foundation of China(No.10572012)
文摘The reflection characteristics of gird structures are calculated by the spatial network method in the case of normal incidence plane electromagnetic wave. The numerical result shows that the grid panels without electromagnetic wave absorbing foams are not ideal. However, the absorbing ability can be achieved as low as -25 dBsm from 8 GHz to 12 GHz when the grid cells are filled with foam absorbers. Also it is noted from computation that the foam filled grid structures with larger cell size, higher and thinner ribs will improve the absorbing abilities, which illustrates that they can be used as the effective light-weight stealth structures for aeronautical application.
文摘An exact and a numerical solutions to the problem of a steady mixed convective MHD flow of an incompressible viscous electrically conducting fluid past an infinite vertical porous plate with combined heat and mass transfer are presented.A uniform magnetic field is assumed to be applied transversely to the direction of the flow with the consideration of the induced magnetic field with viscous and magnetic dissipations of energy.The porous plate is subjected to a constant suction velocity as well as a uniform mixed stream velocity.The governing equations are solved by the perturbation technique and a numerical method.The analytical expressions for the velocity field,the temperature field,the induced magnetic field,the skin-friction,and the rate of heat transfer at the plate are obtained.The numerical results are demonstrated graphically for various values of the parameters involved in the problem.The effects of the Hartmann number,the chemical reaction parameter,the magnetic Prandtl number,and the other parameters involved in the velocity field,the temperature field,the concentration field,and the induced magnetic field from the plate to the fluid are discussed.An increase in the heat source/sink or the Eckert number is found to strongly enhance the fluid velocity values.The induced magnetic field along the x-direction increases with the increase in the Hartmann number,the magnetic Prandtl number,the heat source/sink,and the viscous dissipation.It is found that the flow velocity,the fluid temperature,and the induced magnetic field decrease with the increase in the destructive chemical reaction.Applications of the study arise in the thermal plasma reactor modelling,the electromagnetic induction,the magnetohydrodynamic transport phenomena in chromatographic systems,and the magnetic field control of materials processing.
基金Project supported by the National Natural Science Foundation of China(Grant No.61573017)the Natural Science Foundation of Shaanxi Province,China(Grant No.2016JQ6062)
文摘In this paper, based on simulated annealing a new method to rank important nodes in complex networks is presented.First, the concept of an importance sequence(IS) to describe the relative importance of nodes in complex networks is defined. Then, a measure used to evaluate the reasonability of an IS is designed. By comparing an IS and the measure of its reasonability to a state of complex networks and the energy of the state, respectively, the method finds the ground state of complex networks by simulated annealing. In other words, the method can construct a most reasonable IS. The results of experiments on real and artificial networks show that this ranking method not only is effective but also can be applied to different kinds of complex networks.
基金supported by State Key Development Program of Basic Research of China (Grant No.2010CB429001)
文摘The prediction of solitary wave run-up has important practical significance in coastal and ocean engineering, but the calculation precision is limited in the existing models. For improving the calculation precision, a solitary wave run-up calculation model was established based on artificial neural networks in this study. A back-propagation (BP) network with one hidden layer was adopted and modified with the additional momentum method and the auto-adjusting learning factor. The model was applied to calculation of solitary wave run-up. The correlation coefficients between the neural network model results and the experimental values was 0.996 5. By comparison with the correlation coefficient of 0.963 5, between the Synolakis formula calculation results and the experimental values, it is concluded that the neural network model is an effective method for calculation and analysis of solitary wave ran-up.
文摘Superconductive properties for oxides were predicted by artificial neural network (ANN) method with structural and chemical parameters as inputs. The predicted properties include superconductivity for oxides, distributed ranges of the superconductive transition temperature (Tc) for complex oxides, and Tc values for cuprate superconductors. The calculated results indicated that the adjusted ANN can be used to predict superconductive properties for unknown oxides.
基金financially supported by the National Natural Science Foundation of China(Grant No.51278217)
文摘This paper presents an artificial neural network(ANN)-based response surface method that can be used to predict the failure probability of c-φslopes with spatially variable soil.In this method,the Latin hypercube sampling technique is adopted to generate input datasets for establishing an ANN model;the random finite element method is then utilized to calculate the corresponding output datasets considering the spatial variability of soil properties;and finally,an ANN model is trained to construct the response surface of failure probability and obtain an approximate function that incorporates the relevant variables.The results of the illustrated example indicate that the proposed method provides credible and accurate estimations of failure probability.As a result,the obtained approximate function can be used as an alternative to the specific analysis process in c-φslope reliability analyses.
基金This work was supported by Natural Science Foundation of China(Item number:51777060,U1361109)Natural Science Foundation of Henan province(Item number:162300410117)the he innovative research team plan of Henan Polytechnic University(Item number:T2015-2).
文摘Aiming at obtaining high power density of surface-mounted and interior permanent magnet synchronous motor(SIPMSM),it is important to accurately calculate the temperature field distribution of SIPMSM,and a magnetic-thermal coupling method is proposed.The magnetic-thermal coupling mechanism is analyzed.The thermal network model and finite element model are built by this method,respectively.The effects of power frequency on iron losses and temperature fields are analyzed by the magnetic-thermal coupling finite element model under the condition of rated load,and the relationship between the load and temperature field is researched under the condition of the synchronous speed.In addition,the equivalent thermal network model is used to verify the magnetic-thermal coupling method.Then the temperatures of various nodes are obtained.The results show that there are advantages in both computational efficiency and accuracy for the proposed coupling method,which can be applied to other permanent magnet motors with complex structures.
文摘BP neural networks is used to mid-term earthquake prediction in this paper. Some usual prediction parameters of seismology are used as the import units of neural networks. And the export units of neural networks is called as the character parameter W_0 describing enhancement of seismicity. We applied this method to space scanning of North China. The result shows that the mid-term anomalous zone of W_0-value usually appeared obviously around the future epicenter 1~3 years before earthquake. It is effective to mid-term prediction.
文摘Anisotropy of the strength and deformation behaviors of fractured rock masses is a crucial issue for design and stability assessments of rock engineering structures, due mainly to the non-uniform and non- regular geometries of the fracture systems. However, no adequate efforts have been made to study this issue due to the current practical impossibility of laboratory tests with samples of large volumes con- taining many fractures, and the difficulty for controlling reliable initial and boundary conditions for large-scale in situ tests. Therefore, a reliable numerical predicting approach for evaluating anisotropy of fractured rock masses is needed. The objective of this study is to systematically investigate anisotropy of strength and deformability of fractured rocks, which has not been conducted in the past, using a nu- merical modeling method. A series of realistic two-dimensional (2D) discrete fracture network (DFN) models were established based on site investigation data, which were then loaded in different directions, using the code UDEC of discrete element method (DEM), with changing confining pressures. Numerical results show that strength envelopes and elastic deformability parameters of tested numerical models are significantly anisotropic, and vary with changing axial loading and confining pressures. The results indicate that for design and safety assessments of rock engineering projects, the directional variations of strength and deformability of the fractured rock mass concerned must be treated properly with respect to the directions of in situ stresses. Traditional practice for simply positioning axial orientation of tunnels in association with principal stress directions only may not be adequate for safety requirements. Outstanding issues of the present study and su^zestions for future study are also oresented.