A Deep Learning Approach to Shape Optimization Problems for Flexoelectric Materials Using the Isogeometric Finite Element Method

A new approach for flexoelectricmaterial shape optimization is proposed in this study.In this work,a proxymodel based on artificial neural network(ANN)is used to solve the parameter optimization and shape optimization problems.To improve the fitting ability of the neural network,we use the idea of pre-training to determine the structure of the neural network and combine different optimizers for training.The isogeometric analysis-finite element method(IGA-FEM)is used to discretize the flexural theoretical formulas and obtain samples,which helps ANN to build a proxy model from the model shape to the target value.The effectiveness of the proposed method is verified through two numerical examples of parameter optimization and one numerical example of shape optimization.

Support design method for deep soft-rock tunnels in non-hydrostatic high in-situ stress field

Due to the long-term plate tectonic movements in southwestern China,the in-situ stress field in deep formations is complex.When passing through deep soft-rock mass under non-hydrostatic high in-situ stress field,tunnels will suffer serious asymmetric deformation.There is no available support design method for tunnels under such a situation in existing studies to clarify the support time and support stiffness.This study first analyzed the mechanical behavior of tunnels in non-hydrostatic in-situ stress field and derived the theoretical equations of the ground squeezing curve(GSC)and ground loosening curve(GLC).Then,based on the convergence confinement theory,the support design method of deep soft-rock tunnels under non-hydrostatic high in-situ stress field was established considering both squeezing and loosening pressures.In addition,this method can provide the clear support time and support stiffness of the second layer of initial support.The proposed design method was applied to the Wanhe tunnel of the China-Laos railway in China.Monitoring data indicated that the optimal support scheme had a good effect on controlling the tunnel deformation in non-hydrostatic high in-situ stress field.Field applications showed that the secondary lining could be constructed properly.

Electrical structure identification of deep shale gas reservoir in complex structural area using wide field electromagnetic method

To fully exploit the technical advantages of the large-depth and high-precision artificial source electromagnetic method in the complex structure area of southern Sichuan and compensate for the shortcomings of the conventional electromagnetic method in exploration depth,precision,and accuracy,the large-depth and high-precision wide field electromagnetic method is applied to the complex structure test area of the Luochang syncline and Yuhe nose anticline in the southern Sichuan.The advantages of the wide field electromagnetic method in detecting deep,low-resistivity thin layers are demonstrated.First,on the basis of the analysis of physical property data,a geological–geoelectric model is established in the test area,and the wide field electromagnetic method is numerically simulated to analyze and evaluate the response characteristics of deep thin shale gas layers on wide field electromagnetic curves.Second,a wide field electromagnetic test is conducted in the complex structure area of southern Sichuan.After data processing and inversion imaging,apparent resistivity logging data are used for calibration to develop an apparent resistivity interpretation model suitable for the test area.On the basis of the results,the characteristics of the electrical structure change in the shallow longitudinal formation of 6 km are implemented,and the transverse electrical distribution characteristics of the deep shale gas layer are delineated.In the prediction area near the well,the subsequent data verification shows that the apparent resistivity obtained using the inversion of the wide field electromagnetic method is consistent with the trend of apparent resistivity revealed by logging,which proves that this method can effectively identify the weak response characteristics of deep shale gas formations in complex structural areas.This experiment,it is shown shows that the wide field electromagnetic method with a large depth and high precision can effectively characterize the electrical characteristics of deep,low-resistivity thin layers in complex structural areas,and a new set of low-cost evaluation technologies for shale gas target layers based on the wide field electromagnetic method is explored.

Deep learning-based battery state of charge estimation:Enhancing estimation performance with unlabelled training samples

The estimation of state of charge(SOC)using deep neural networks(DNN)generally requires a considerable number of labelled samples for training,which refer to the current and voltage pieces with knowing their corresponding SOCs.However,the collection of labelled samples is costly and time-consuming.In contrast,the unlabelled training samples,which consist of the current and voltage data with unknown SOCs,are easy to obtain.In view of this,this paper proposes an improved DNN for SOC estimation by effectively using both a pool of unlabelled samples and a limited number of labelled samples.Besides the traditional supervised network,the proposed method uses an input reconstruction network to reformulate the time dependency features of the voltage and current.In this way,the developed network can extract useful information from the unlabelled samples.The proposed method is validated under different drive cycles and temperature conditions.The results reveal that the SOC estimation accuracy of the DNN trained with both labelled and unlabelled samples outperforms that of only using a limited number of labelled samples.In addition,when the dataset with reduced number of labelled samples to some extent is used to test the developed network,it is found that the proposed method performs well and is robust in producing the model outputs with the required accuracy when the unlabelled samples are involved in the model training.Furthermore,the proposed method is evaluated with different recurrent neural networks(RNNs)applied to the input reconstruction module.The results indicate that the proposed method is feasible for various RNN algorithms,and it could be flexibly applied to other conditions as required.

Deep learning-based magnetic resonance imaging reconstruction for improving the image quality of reduced-field-of-view diffusionweighted imaging of the pancreas

BACKGROUND It has been reported that deep learning-based reconstruction(DLR)can reduce image noise and artifacts,thereby improving the signal-to-noise ratio and image sharpness.However,no previous studies have evaluated the efficacy of DLR in improving image quality in reduced-field-of-view(reduced-FOV)diffusionweighted imaging(DWI)[field-of-view optimized and constrained undistorted single-shot(FOCUS)]of the pancreas.We hypothesized that a combination of these techniques would improve DWI image quality without prolonging the scan time but would influence the apparent diffusion coefficient calculation.AIM To evaluate the efficacy of DLR for image quality improvement of FOCUS of the pancreas.METHODS This was a retrospective study evaluated 37 patients with pancreatic cystic lesions who underwent magnetic resonance imaging between August 2021 and October 2021.We evaluated three types of FOCUS examinations:FOCUS with DLR(FOCUS-DLR+),FOCUS without DLR(FOCUS-DLR−),and conventional FOCUS(FOCUS-conv).The three types of FOCUS and their apparent diffusion coefficient(ADC)maps were compared qualitatively and quantitatively.RESULTS FOCUS-DLR+(3.62,average score of two radiologists)showed significantly better qualitative scores for image noise than FOCUS-DLR−(2.62)and FOCUS-conv(2.88)(P<0.05).Furthermore,FOCUS-DLR+showed the highest contrast ratio and 600 s/mm^(2)(0.72±0.08 and 0.68±0.08)and FOCUS-DLR−showed the highest CR between cystic lesions and the pancreatic parenchyma for the b-values of 0 and 600 s/mm2(0.62±0.21 and 0.62±0.21)(P<0.05),respectively.FOCUS-DLR+provided significantly higher ADCs of the pancreas and lesion(1.44±0.24 and 3.00±0.66)compared to FOCUS-DLR−(1.39±0.22 and 2.86±0.61)and significantly lower ADCs compared to FOCUS-conv(1.84±0.45 and 3.32±0.70)(P<0.05),respectively.CONCLUSION This study evaluated the efficacy of DLR for image quality improvement in reduced-FOV DWI of the pancreas.DLR can significantly denoise images without prolonging the scan time or decreasing the spatial resolution.The denoising level of DWI can be controlled to make the images appear more natural to the human eye.However,this study revealed that DLR did not ameliorate pancreatic distortion.Additionally,physicians should pay attention to the interpretation of ADCs after DLR application because ADCs are significantly changed by DLR.

Automatic Diagnosis of Polycystic Ovarian Syndrome Using Wrapper Methodology with Deep Learning Techniques

One of the significant health issues affecting women that impacts their fertility and results in serious health concerns is Polycystic ovarian syndrome(PCOS).Consequently,timely screening of polycystic ovarian syndrome can help in the process of recovery.Finding a method to aid doctors in this procedure was crucial due to the difficulties in detecting this condition.This research aimed to determine whether it is possible to optimize the detection of PCOS utilizing Deep Learning algorithms and methodologies.Additionally,feature selection methods that produce the most important subset of features can speed up calculation and enhance the effectiveness of classifiers.In this research,the tri-stage wrapper method is used because it reduces the computation time.The proposed study for the Automatic diagnosis of PCOS contains preprocessing,data normalization,feature selection,and classification.A dataset with 39 characteristics,including metabolism,neuroimaging,hormones,and biochemical information for 541 subjects,was employed in this scenario.To start,this research pre-processed the information.Next for feature selection,a tri-stage wrapper method such as Mutual Information,ReliefF,Chi-Square,and Xvariance is used.Then,various classification methods are tested and trained.Deep learning techniques including convolutional neural network(CNN),multi-layer perceptron(MLP),Recurrent neural network(RNN),and Bi long short-term memory(Bi-LSTM)are utilized for categorization.The experimental finding demonstrates that with effective feature extraction process using tri stage wrapper method+CNN delivers the highest precision(97%),high accuracy(98.67%),and recall(89%)when compared with other machine learning algorithms.

Quantitative characterization of the brittleness of deep shales by integrating mineral content,elastic parameters,in situ stress conditions and logging analysis

Deep shale reservoirs(3500–4500 m)exhibit significantly different stress states than moderately deep shale reservoirs(2000–3500 m).As a result,the brittleness response mechanisms of deep shales are also different.It is urgent to investigate methods to evaluate the brittleness of deep shales to meet the increasingly urgent needs of deep shale gas development.In this paper,the quotient of Young’s modulus divided by Poisson’s ratio based on triaxial compression tests under in situ stress conditions is taken as SSBV(Static Standard Brittleness Value).A new and pragmatic technique is developed to determine the static brittleness index that considers elastic parameters,the mineral content,and the in situ stress conditions(BIEMS).The coefficient of determination between BIEMS and SSBV reaches 0.555 for experimental data and 0.805 for field data.This coefficient is higher than that of other brittleness indices when compared to SSBV.BIEMS can offer detailed insights into shale brittleness under various conditions,including different mineral compositions,depths,and stress states.This technique can provide a solid data-based foundation for the selection of‘sweet spots’for single-well engineering and the comparison of the brittleness of shale gas production layers in different areas.

Toward Improved Accuracy in Quasi-Static Elastography Using Deep Learning

Elastography is a non-invasive medical imaging technique to map the spatial variation of elastic properties of soft tissues.The quality of reconstruction results in elastography is highly sensitive to the noise induced by imaging measurements and processing.To address this issue,we propose a deep learning(DL)model based on conditional Generative Adversarial Networks(cGANs)to improve the quality of nonhomogeneous shear modulus reconstruction.To train this model,we generated a synthetic displacement field with finite element simulation under known nonhomogeneous shear modulus distribution.Both the simulated and experimental displacement fields are used to validate the proposed method.The reconstructed results demonstrate that the DL model with synthetic training data is able to improve the quality of the reconstruction compared with the well-established optimization method.Moreover,we emphasize that our DL model is only trained on synthetic data.This might provide a way to alleviate the challenge of obtaining clinical or experimental data in elastography.Overall,this work addresses several fatal issues in applying the DL technique into elastography,and the proposed method has shown great potential in improving the accuracy of the disease diagnosis in clinical medicine.

Improved Double Deep Q Network Algorithm Based on Average Q-Value Estimation and Reward Redistribution for Robot Path Planning

By integrating deep neural networks with reinforcement learning,the Double Deep Q Network(DDQN)algorithm overcomes the limitations of Q-learning in handling continuous spaces and is widely applied in the path planning of mobile robots.However,the traditional DDQN algorithm suffers from sparse rewards and inefficient utilization of high-quality data.Targeting those problems,an improved DDQN algorithm based on average Q-value estimation and reward redistribution was proposed.First,to enhance the precision of the target Q-value,the average of multiple previously learned Q-values from the target Q network is used to replace the single Q-value from the current target Q network.Next,a reward redistribution mechanism is designed to overcome the sparse reward problem by adjusting the final reward of each action using the round reward from trajectory information.Additionally,a reward-prioritized experience selection method is introduced,which ranks experience samples according to reward values to ensure frequent utilization of high-quality data.Finally,simulation experiments are conducted to verify the effectiveness of the proposed algorithm in fixed-position scenario and random environments.The experimental results show that compared to the traditional DDQN algorithm,the proposed algorithm achieves shorter average running time,higher average return and fewer average steps.The performance of the proposed algorithm is improved by 11.43%in the fixed scenario and 8.33%in random environments.It not only plans economic and safe paths but also significantly improves efficiency and generalization in path planning,making it suitable for widespread application in autonomous navigation and industrial automation.

An improved algorithm based on equivalent sound velocity profile method at large incident angle

With the development of ultra-wide coverage technology,multibeam echo-sounder(MBES)system has put forward higher requirements for localization accuracy and computational efficiency of ray tracing method.The classical equivalent sound speed profile(ESSP)method replaces the measured sound velocity profile(SVP)with a simple constant gradient SVP,reducing the computational workload of beam positioning.However,in deep-sea environment,the depth measurement error of this method rapidly increases from the central beam to the edge beam.By analyzing the positioning error of the ESSP method at edge beam,it is discovered that the positioning error increases monotonically with the incident angle,and the relationship between them could be expressed by polynomial function.Therefore,an error correction algorithm based on polynomial fitting is obtained.The simulation experiment conducted on an inclined seafloor shows that the proposed algorithm exhibits comparable efficiency to the original ESSP method,while significantly improving bathymetry accuracy by nearly eight times in the edge beam.

Nonparametric Statistical Feature Scaling Based Quadratic Regressive Convolution Deep Neural Network for Software Fault Prediction

The development of defect prediction plays a significant role in improving software quality. Such predictions are used to identify defective modules before the testing and to minimize the time and cost. The software with defects negatively impacts operational costs and finally affects customer satisfaction. Numerous approaches exist to predict software defects. However, the timely and accurate software bugs are the major challenging issues. To improve the timely and accurate software defect prediction, a novel technique called Nonparametric Statistical feature scaled QuAdratic regressive convolution Deep nEural Network (SQADEN) is introduced. The proposed SQADEN technique mainly includes two major processes namely metric or feature selection and classification. First, the SQADEN uses the nonparametric statistical Torgerson–Gower scaling technique for identifying the relevant software metrics by measuring the similarity using the dice coefficient. The feature selection process is used to minimize the time complexity of software fault prediction. With the selected metrics, software fault perdition with the help of the Quadratic Censored regressive convolution deep neural network-based classification. The deep learning classifier analyzes the training and testing samples using the contingency correlation coefficient. The softstep activation function is used to provide the final fault prediction results. To minimize the error, the Nelder–Mead method is applied to solve non-linear least-squares problems. Finally, accurate classification results with a minimum error are obtained at the output layer. Experimental evaluation is carried out with different quantitative metrics such as accuracy, precision, recall, F-measure, and time complexity. The analyzed results demonstrate the superior performance of our proposed SQADEN technique with maximum accuracy, sensitivity and specificity by 3%, 3%, 2% and 3% and minimum time and space by 13% and 15% when compared with the two state-of-the-art methods.

DLBT:Deep Learning-Based Transformer to Generate Pseudo-Code from Source Code

Understanding the content of the source code and its regular expression is very difficult when they are written in an unfamiliar language.Pseudo-code explains and describes the content of the code without using syntax or programming language technologies.However,writing Pseudo-code to each code instruction is laborious.Recently,neural machine translation is used to generate textual descriptions for the source code.In this paper,a novel deep learning-based transformer(DLBT)model is proposed for automatic Pseudo-code generation from the source code.The proposed model uses deep learning which is based on Neural Machine Translation(NMT)to work as a language translator.The DLBT is based on the transformer which is an encoder-decoder structure.There are three major components:tokenizer and embeddings,transformer,and post-processing.Each code line is tokenized to dense vector.Then transformer captures the relatedness between the source code and the matching Pseudo-code without the need of Recurrent Neural Network(RNN).At the post-processing step,the generated Pseudo-code is optimized.The proposed model is assessed using a real Python dataset,which contains more than 18,800 lines of a source code written in Python.The experiments show promising performance results compared with other machine translation methods such as Recurrent Neural Network(RNN).The proposed DLBT records 47.32,68.49 accuracy and BLEU performance measures,respectively.

Application of the “Three Threes” Method in Clinical Teaching of Internal Jugular Vein Puncture

Objective: To clarify the role of the “Three Threes” method in clinical teaching of internal jugular vein puncture and explore improvements in teaching methods. Methods: A doctor was assigned to the induction room of the Second Affiliated Hospital of Naval Medical University (Shanghai Changzheng Hospital) for two months. The time required for catheterization, the first puncture success rate, and occurrence of puncture-related complications were compared before and after learning the “Three Threes” method. Results: Using the “Three Threes” method reduced the catheterization time by 43%, increased the first puncture success rate by 17%, and led to fewer puncture-related complications. Conclusion: The application of the “Three Threes” method not only improves the success rate of internal jugular vein puncture but also reduces complications, making it easier for students to master the technique.

Stoichiometry and Stability Constant Values for Copper (II) Chelates with Ethylene Diamine in Deep Eutectic Solvents (DES) (Ethaline) Solutions

In this study we used the deep eutectic solvents (ionic liquids) to investigate the reaction between copper (II) with ethylene diamine (en). Two of the existing methods for analyzing spectrophotometric measurements have been applied for establishing, the stoichiometry and whenever possible, the stability constants of the chelates formed. The method of continuous variations was necessary to determine first whether, the metal ion and the ligand ethylene diamine form one or more than one chelate, when more than one chelate formed, the results obtained depend on the wavelength and for meaningful conclusions the wavelengths were carefully selected. The empirical formulae of the chelates were further substantiated by the molar ratio method. The effect of time and temperature on the formation and stability of these chelates in solution is also studied. The stability constants, K1 and K2 for the copper (II) chelates were calculated, though reliable, and are comparable to literature values.

Fully Distributed Learning for Deep Random Vector Functional-Link Networks

In the contemporary era, the proliferation of information technology has led to an unprecedented surge in data generation, with this data being dispersed across a multitude of mobile devices. Facing these situations and the training of deep learning model that needs great computing power support, the distributed algorithm that can carry out multi-party joint modeling has attracted everyone’s attention. The distributed training mode relieves the huge pressure of centralized model on computer computing power and communication. However, most distributed algorithms currently work in a master-slave mode, often including a central server for coordination, which to some extent will cause communication pressure, data leakage, privacy violations and other issues. To solve these problems, a decentralized fully distributed algorithm based on deep random weight neural network is proposed. The algorithm decomposes the original objective function into several sub-problems under consistency constraints, combines the decentralized average consensus (DAC) and alternating direction method of multipliers (ADMM), and achieves the goal of joint modeling and training through local calculation and communication of each node. Finally, we compare the proposed decentralized algorithm with several centralized deep neural networks with random weights, and experimental results demonstrate the effectiveness of the proposed algorithm.

基于DeepLabV3+模型改进的图像分割方法研究

近年来,计算机视觉快速发展,其中图像分割在计算机视觉中也举足轻重,在城市现代化建设,智能驾驶,地理勘测等方面都得到了充分应用。但是,大多数分割方法只关注图像特征的纵向深层特征与浅层特征的简单融合,而忽略了同一层图像特征的横向远程关系。针对此问题,基于DeepLabV3+框架,加入Swin-Transformer block,利用其自注意力机制特点,进行网络特征提取,以提高图像分割的全局和细节优化。其次,改进DeepLabV3+中上采样方法,利用CARAFE上采样模块取代简单的双线性插值法。实验表明,改进后的模型相较于基线模型MIoU提升2%,ACC提升1%。

RESEARCH ON THE EVALUATION METHOD OF FRICTION AND LUBRICATION IN DEEP DRAWING

The deformation characteristic of bland in deep drawing is discussed. It is pointed out that the friction and lubrication conditions in for drawing are different from that in mechanical motion or machine work or other plastic process. The common test methods in laboratories are analyzed. It shows that though all those test methods can test the friction coefficient, the probe test method is most suitable for the research of friction and lubrication and the process in deep drawing, for this method is identical with the actual work condition either from the test principle or deformation status of the blank. Last the successful application in the deep drawing simulator newly developed the the probe method are intro- duced in detail.

A Deep Collocation Method for the Bending Analysis of Kirchhoff Plate

In this paper,a deep collocation method(DCM)for thin plate bending problems is proposed.This method takes advantage of computational graphs and backpropagation algorithms involved in deep learning.Besides,the proposed DCM is based on a feedforward deep neural network(DNN)and differs from most previous applications of deep learning for mechanical problems.First,batches of randomly distributed collocation points are initially generated inside the domain and along the boundaries.A loss function is built with the aim that the governing partial differential equations(PDEs)of Kirchhoff plate bending problems,and the boundary/initial conditions are minimised at those collocation points.A combination of optimizers is adopted in the backpropagation process to minimize the loss function so as to obtain the optimal hyperparameters.In Kirchhoff plate bending problems,the C^1 continuity requirement poses significant difficulties in traditional mesh-based methods.This can be solved by the proposed DCM,which uses a deep neural network to approximate the continuous transversal deflection,and is proved to be suitable to the bending analysis of Kirchhoff plate of various geometries.

Excavation-induced deep hard rock fracturing:Methodology and applications

To analyze and predict the mechanical behaviors of deep hard rocks,some key issues concerning rock fracturing mechanics for deep hard rock excavations are discussed.First,a series of apparatuses and methods have been developed to test the mechanical properties and fracturing behaviors of hard rocks under high true triaxial stress paths.Evolution mechanisms of stress-induced disasters in deep hard rock excavations,such as spalling,deep cracking,massive roof collapse,large deformation and rockbursts,have been recognized.The analytical theory for the fracturing process of hard rock masses,including the three-dimensional failure criterion,stress-induced mechanical model,fracturing degree index,energy release index and numerical method,has been established.The cracking-restraint method is developed for mitigating or controlling rock spalling,deep cracking and massive collapse of deep hard rocks.An energy-controlled method is also proposed for the prevention of rockbursts.Finally,two typical cases are used to illustrate the application of the proposed methodology in the Baihetan caverns and Bayu tunnels of China.

Assessment of strain bursting in deep tunnelling by using the finite-discrete element method

Rockbursting in deep tunnelling is a complex phenomenon posing significant challenges both at the design and construction stages of an underground excavation within hard rock masses and under high in situ stresses. While local experience, field monitoring, and informed data-rich analysis are some of the tools commonly used to manage the hazards and the associated risks, advanced numerical techniques based on discontinuum modelling have also shown potential in assisting in the assessment of rockbursting. In this study, the hybrid finite-discrete element method(FDEM) is employed to investigate the failure and fracturing processes, and the mechanisms of energy storage and rapid release resulting in bursting, as well as to assess its utility as part of the design process of underground excavations.Following the calibration of the numerical model to simulate a deep excavation in a hard, massive rock mass, discrete fracture network(DFN) geometries are integrated into the model in order to examine the impact of rock structure on rockbursting under high in situ stresses. The obtained analysis results not only highlight the importance of explicitly simulating pre-existing joints within the model, as they affect the mobilised failure mechanisms and the intensity of strain bursting phenomena, but also show how the employed joint network geometry, the field stress conditions, and their interaction influence the extent and depth of the excavation induced damage. Furthermore, a rigorous analysis of the mass and velocity of the ejected rock blocks and comparison of the obtained data with well-established semi-empirical approaches demonstrate the potential of the method to provide realistic estimates of the kinetic energy released during bursting for determining the energy support demand.