Complex Mapping Relationship Research Articles

Research on Network Intrusion Detection Based on Transformer

With the advancement of technology, the development of various industries has become inseparable from informatization. People's lives have become closely related to the network. While using the network to facilitate our lives, massive data is also generated. Traditional firewall technologies are no longer sufficient to meet current needs. Deep learning algorithms can establish complex mapping relationships between network data, and can extract hidden correlation features between data features to achieve data recognition and prediction. Therefore, this paper introduces Transformer and Bidirectional Long Short-Term Memory (BiLSTM) into the field of intrusion detection, and proposes an intrusion detection method based on the combination of Transformer-Encoder and BiLSTM (TBL). Deep Neural Networks (DNN) are used to further extract data features, and the softmax function is used to output classification results. In order to verify the effectiveness of this method, this paper trains and tests the TBL method on the NSL-KDD dataset, and verifies its feasibility and superiority.

Structural Nonlinear Model Updating Based on an Improved Generative Adversarial Network

This study proposes a novel nonlinear model updating approach based on an improved generative adversarial network (GAN). In the improved GAN, a convolutional neural network (CNN) surrogate model is added to the discriminator network to enhance the capability of the GAN to learn the complex mapping relationship between vibration responses and nonlinear model parameters. To avoid the gradient disappearance present in the traditional GAN, a combined objective function is added to the improved GAN model. In the network training process, the instantaneous amplitudes of the decomposed accelerations are extracted as input samples and the nonlinear model parameters are defined as the GAN output. When the improved GAN is trained, the trained network model is capable of estimating the nonlinear model parameters based on measured instantaneous acceleration amplitudes. To confirm the feasibility of the improved GAN for structural nonlinear model updating, a steel-concrete hybrid bridge tower subjected to seismic excitation is numerically simulated and the effects of different numbers of data points and noise levels are studied. Furthermore, the identification accuracy of the improved GAN is compared with the updated results. For experimental applications, the shake table test of a scaled steel-concrete hybrid bridge tower subjected to seismic excitations is employed to confirm the effectiveness of the proposed nonlinear model updating method. Both numerical and experimental results demonstrate that the improved GAN model is reliable and effective for the nonlinear model updating of structures subjected to seismic excitation.

Response modeling and optimization based on sequential experimental design

Aiming at the modeling problem of complex mapping relationship between influencing factors and responses, this paper implements a mixed sequential design method based on Voronoi and LOLA. And according to the sample points collected by sequential design, a response model is established combined with the LSSVR method. Through function examples, the effectiveness of the method in different complexity levels, different levels of values and different dimensional spaces is verified. Experiments show that the algorithm has high accuracy in different shapes and different dimensions. Finally, the sensitivity analysis of the algorithm is carried out to explore the influence of the number of points added each time on the accuracy of the final response model.

A Data Driven Security Correction Method for Power Systems with UPFC

The access of unified power flow controllers (UPFC) has changed the structure and operation mode of power grids all across the world, and it has brought severe challenges to the traditional real-time calculation of security correction based on traditional models. Considering the limitation of computational efficiency regarding complex, physical models, a data-driven power system security correction method with UPFC is, in this paper, proposed. Based on the complex mapping relationship between the operation state data and the security correction strategy, a two-stage deep neural network (DNN) learning framework is proposed, which divides the offline training task of security correction into two stages: in the first stage, the stacked auto-encoder (SAE) classification model is established, and the node correction state (0/1) output based on the fault information; in the second stage, the DNN learning model is established, and the correction amount of each action node is obtained based on the action nodes output in the previous stage. In this paper, the UPFC demonstration project of Nanjing West Ring Network is taken as a case study to validate the proposed method. The results show that the proposed method can fully meet the real-time security correction time requirements of power grids, and avoid the inherent defects of the traditional model method without an iterative solution and can also provide reasonable security correction strategies for N-1 and N-2 faults.

Multi-output soft sensor modeling approach for penicillin fermentation process based on features of big data

The product quality indicators of the penicillin fermentation process have multiple semantics and are interrelated. There is a complex nonlinear mapping relationship between input characteristics and multiple-output objectives, and the time dependence is strong. As a result, the prediction accuracy of existing soft sensor models is poor, and it is difficult to meet the needs of industrial sites. To solve the above problems, this paper proposes a multi-output soft sensor modeling method for the penicillin fermentation process based on big data feature analysis. In this method, the process data is divided into several batches in order, and then the data features of multivariable and time-dependent datasets are extracted according to the deep sparse self-coding neural network method to realize the effective mining of the relationship between multivariable time series factors, based on the multi-output support vector regression method, several soft sensor models for different prediction targets are established. Meanwhile, to improve the Predictive performance of the soft sensor model, the improved black hole algorithm is used to optimize the model parameters. Finally, a simulation experiment is carried out based on the simulation dataset of the penicillin fermentation process to verify the effectiveness of the proposed method.

Fatigue Performance Prediction of RC Beams Based on Optimized Machine Learning Technology.

The development of fatigue damage in reinforced concrete (RC) beams is affected by various factors such as repetitive loads and material properties, and there exists a complex nonlinear mapping relationship between their fatigue performance and each factor. To this end, a fatigue performance prediction model for RC beams was proposed based on the deep belief network (DBN) optimized by particle swarm optimization (PSO). The original database of fatigue loading tests was established by conducting fatigue loading tests on RC beams. The mid-span deflection, reinforcement strain, and concrete strain during fatigue loading of RC beams were predicted and evaluated. The fatigue performance prediction results of the RC beam based on the PSO-DBN model were compared with those of the single DBN model and the BP model. The models were evaluated using the R2 coefficient, mean absolute percentage error, mean absolute error, and root mean square error. The results showed that the fatigue performance prediction model of RC beams based on PSO-DBN is more accurate and efficient.

HPE-GCN: Predicting efficacy of tonic formulae via graph convolutional networks integrating traditionally defined herbal properties.

Chinese herbal formulae are the heritage of traditional Chinese medicine (TCM) in treating diseases through thousands of years. The formula function is not just a simple herbal efficacy addition, but produces complex and nonlinear relationships between different herbs and their overall efficacy, which brings challenges to the formula efficacy analysis. In our study, we proposed a model called HPE-GCN that combines graph convolutional networks (GCN) with TCM-defined herbal properties (TCM-HPs) to predict formulae efficacy. In addition, to process the unstructured natural language in the formula text, we proposed a weighting calculation method related to herb frequency and the number of herbs in a formula called Formula-Herb dependence degree (FHDD), to assess the dependency degree of a formula with its herbs. In our research, 214 classic tonic formulae from ancient TCM books such as Synopsis of the Golden Chamber, Jingyue's Complete Works and the Golden Mirror of Medicin were collected as datasets. The performance of HPE-GCN on multi-classification of tonic formulae reached the best result compared with classic machine learning models, such as support vector machine, naive Bayes, logistic regression, gradient boosting decision tree, and K-nearest neighbors. The evaluated index Macro-Precision, Macro-Recall, Macro-F1 of HPE-GCN on the test set were 87.70%, 84.08% and 83.51% respectively, increased by 7.27%, 7.41% and 7.30% respectively from second best compared models. GCN has the advantage of low-dimensional feature expression for herbs and formulae, and is an effective analysis tool for TCM research. HPE-GCN integrates TCM-HPs and fits the complex nonlinear mapping relationship between TCM-HPs and formulae efficacy, which provides new ideas for related research.

Detecting Points of Interest on 3D Model via Alternating Optimization Strategy

A supervised 3D points of interest (POI) detection algorithm is proposed based on alternating optimization. Firstly, the geometric features of a 3D shape are calculated from several hand-crafted feature descriptors and used as the input of the neural network. Secondly, the biharmonic distance field is utilized to assign a label to each vertex, which is regarded as the neural network’s output. Thirdly, the complex mapping relationships between the feature vectors and the labels are learned through the neural network. Fourthly, predictions are made on the training set using the trained neural network. The differences between the predicted points of interest and ground truth are compared to further optimize the vertices’ labels, which are then used as the output to train the neural network. The third and fourth steps for alternating optimization are repeated for several times, and a neural network is finally obtained. The experimental results on the SHREC 2011 dataset show that, due to the alternate optimization strategy, our algorithm is better than the traditional methods in the key evaluation indicators FNE and FPE, and the accuracy of proposed algorithm has achieved an average improvement of more than 11%.

An Efficient Knowledge-Based Artificial Neural Network for the Design of Circularly Polarized 3-D-Printed Lens Antenna

An efficient knowledge-based artificial neural network (KBANN) is proposed, and it is used for the design of circularly polarized (CP) lens antenna in this article. In this KBANN, forward neural network (FNN) and inverse neural network (INN) are included. In this model, INN is the major component to predict the antenna structure parameters. As multiple performance indices are required, INN requires a large number of training samples to deduce complex mapping relationship. To solve this problem, FNN is introduced to provide prior knowledge for INN. FNN generates a huge training dataset for INN training, and then the trained INN can directly output the geometric parameters by feeding the target electromagnetic responses as input. This article solves the problem of multiple performance indices in antenna design, and a CP lens antenna with wideband, good axial ratio, and high gain is designed and fabricated to verify the effectiveness of the KBANN model.

Machine Learning Assisted Ultra-Wideband Fiber-Optics Mode Selective Coupler Design

We report a machine learning assisted design strategy for ultra-wideband fiber-optics mode selective coupler (MSC). With the help of deep neural network (DNN), we are able to efficiently obtain the complex mapping relationship between structure parameters of the bridge fiber and effective refractive index (RI) of LP <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">01</sub> mode over the S+C+L band. For high-order modes (HOMs) arising in the arbitrarily chosen few-mode fiber (FMF), the trained DNN can rapidly optimize the bridge fiber parameters. After optimization, the ultra-wideband phase matching between the effective RI of LP <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">01</sub> mode arising in the bridge fiber and the effective RI of specific mode arising in the FMF over the S+C+L band can be realized. As an example, ultra-wideband fiber-optics 4-LP MSC with an average mode conversion efficiency (MCE) of more than 93% and an average mode extinction ratio (MER) of more than 17 dB can be secured. In comparison with the traditional genetic algorithm (GA), the DNN-assisted fiber-optics MSC design can significantly reduce the optimization time from 4 hours to 45 seconds, together with the maintenance of excellent ultra-wideband performance. Meanwhile, the ultra-wideband MSC for orbital angular momentum (OAM) mode conversion is demonstrated. Both OAM <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">+1</sub> and OAM <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">+2</sub> modes with an average mode purity of more than 97% and an average MCE of more than 89% over the S+C+L band can be secured. We envision the DNN-assisted fiber-optics device enables potential applications for future mode division multiplexing and wavelength division multiplexing hybrid optical transmission system.

Cascaded Steering Control Paradigm for the Lateral Automation of Heavy Commercial Vehicles

The nonlinearity and complexity of the lateral dynamics of heavy commercial vehicles (HCVs) and their steering subsystems usually pose considerable challenges in realizing automatic lateral control for HCVs. In this article, a novel cascade control paradigm is presented for the automatic lateral guidance control of HCVs, which is equipped with an electrohydraulic coupling power steering (EHCPS) system. The objective is to propose a control framework for the satisfactory trajectory tracking performance of HCVs. This control framework consists of an outer vehicle chassis dynamics controller and an inner steering controller, which are implemented to mitigate the issues associated with the trajectory tracking process due to the intrinsic parameter-varying nonlinearity of the vehicle lateral dynamics model and the steering tracking process caused by the complex mapping relationship between the steering wheel angle input and front wheel angle output, respectively. Using the linear parameter-varying (LPV) technique as a general framework, an outer loop controller is designed to produce the desired front wheel angle, which features a gain scheduling mechanism to cope with the time-varying longitudinal velocity. Based on the analysis of the simplified physical model of the EHCPS system, a feedforward set-point scheduling inner loop controller with sliding-mode observer (SMO) is proposed to ensure good tracking performance of the front wheel angle for the steering subsystem. Through experimentation, conducted on the hardware-in-the-loop (HIL) test platform for HCV automatic lateral control, the feasibility and effectiveness of the proposed cascade steering control paradigm are demonstrated.

Nonlinear overall annoyance level modeling and interior sound quality prediction for pure electric vehicle with extreme gradient boosting algorithm

A single psychoacoustic metric cannot reflect pure electric vehicle (PEV) interior overall sound quality (SQ) accurately. There exists a complex nonlinear mapping relationship between psychoacoustic indexes and SQ prediction. In this work, a nonlinear overall annoyance level modeling method using the extreme gradient boosting (XGBoost) algorithm is proposed. Firstly, comprehensive noise samples inside the cabins of PEV and conventional fuel vehicle (FV) at high speed are collected. Subjective and objective SQ evaluations are conducted for FV and PEV interior noise, respectively. Their differences are analyzed by comparison. The correlation among various SQ indexes is studied, revealing a significant multicollinearity disruption. Then, the relative importance ranking is determined by the random forests algorithm, which provides a basis for feature index selection. Finally, the XGBoost-based nonlinear overall annoyance level model for SQ evaluation is established. Using the XGBoost-based model, traditional multiple linear regression (MLR) model, and linear least absolute shrinkage and selection operator (LASSO) model, PEV SQ is predicted. The results show that the XGBoost-based model exhibits high prediction accuracy, improving 75.6% relative to the MLR model and 74.9% relative to the LASSO model. Meanwhile, it has excellent stability and goodness-of-fitting. Thus, it provides an alternative method for PEV SQ evaluation and prediction.

Prediction of Ozone Hourly Concentrations Based on Machine Learning Technology

To optimize the accuracy of ozone (O3) concentration prediction, this paper proposes a combined prediction model of O3 hourly concentration, FC-LsOA-KELM, which integrates multiple machine learning methods. The model has three parts. The first part is the feature construction (FC), which is based on correlation analysis and incorporates time-delay effect analysis to provide a valuable feature set. The second part is the kernel extreme learning machine (KELM), which can establish a complex mapping relationship between feature set and prediction object. The third part is the lioness optimization algorithm (LsOA), which is purposed to find the optimal parameter combination of KELM. Then, we use air pollution data from 11 cities on Fenwei Plain in China from 2 January 2015 to 30 December 2019 to test the validity of FC-LsOA-KELM and compare it with other prediction methods. The experimental results show that FC-LsOA-KELM can obtain better prediction results and has a better performance.

Digital refocusing based on deep learning in optical coherence tomography.

We present a deep learning-based digital refocusing approach to extend depth of focus for optical coherence tomography (OCT) in this paper. We built pixel-level registered pairs of en face low-resolution (LR) and high-resolution (HR) OCT images based on experimental data and introduced the receptive field block into the generative adversarial networks to learn the complex mapping relationship between LR-HR image pairs. It was demonstrated by results of phantom and biological samples that the lateral resolutions of OCT images were improved in a large imaging depth clearly. We firmly believe deep learning methods have broad prospects in optimizing OCT imaging.

Parallel interactive delayed attention network for pansharpening

ABSTRACT As deep learning has gained wide attention in computer vision tasks, researchers have also started to explore the application of deep learning in panchromatic sharpening. In recent years, various convolutional neural network-based methods for panchromatic sharpening have been proposed, and these methods have achieved good results. However, shallower networks are unable to learn complex mapping relationships. Networks that are too deep are prone to overfitting, which leads to the degradation of the fusion effect. Moreover, since the convolutional operations are concentrated in local regions, it is difficult to obtain the association information with other regions even in deep networks. Therefore, more accurate feature selection and representation with limited network depth is needed. In this paper, a parallel interactive delayed attention network for panchromatic sharpening (PIDAN) is proposed. The network consists of multiple resolution subnetworks with parallel inputs to improve the representation of high-resolution feature maps by multiscale low-resolution features of the same depth and similar level. In addition, we propose a delayed channel attention module. This module obtains correlations between low-frequency and high-frequency information through adaptive learning, making the network more flexible in processing different types of information. The depth feature reweighting restriction combined with the residual unit can further avoid the overfitting representation of features as the network deepens. Experimental results on Gaofen-2 and WorldView-2 datasets show that the proposed method is competitive in both quality assessment and visual perception.

A novel RSG-based intelligent bearing fault diagnosis method for motors in high-noise industrial environment

Bearing fault diagnosis is a critical and challenging task for prognostics and health management of motors. The ability to efficiently and accurately classify the fault categories based on sensor signals is the key to successful bearing fault diagnosis. Although various data-driven methods have been developed for fault diagnosis in recent years, automatic and effective extraction of discriminative fault features from high-noise vibration signals generated in the real-world industrial environment remains a challenging task. To tackle this challenge, this paper proposes a novel deep learning method based on the combination of residual building Unit, soft thresholding and global context, called RSG, to solve the complex mapping relationship between vibration signals and different types of bearing faults. The proposed RSG integrates the working mechanisms of soft threshold and global context to achieve effective noise reduction and feature extraction. A comparative analysis is performed to demonstrate the advantages of the proposed method. Furthermore, the proposed method is tested on a faulty motor dataset collected by our developed intelligent motor test platform based on Industrial Internet of Things. Experimental results show that our method can achieve an average fault diagnosis accuracy of 98%. Thus, the proposed method proves to be an efficient solution for intelligent bearing fault diagnosis for motors in a high-noise industrial environment.

A straightforward and miniature implementation method of postural synergies to replicate human grasp characteristics accurately and intuitively

Postural synergies have great potential for replicating human grasp characteristics, simplify grasp control and reduce the number of hardware actuators required. However, due to their complex mapping relationship and jagged transmission ratio, the implemented mechanisms are always too bulky and loose, which greatly limits their application. With current solutions, the replication accuracy of motion characteristics or intuitive control is compromised, and hitherto no work in the literature has reported the replication errors. To overcome these limitations, we present a novel design framework to determine the actuation configuration, implementation scheme and physical parameters. In this way, the mechanism is miniaturized and can be compactly embedded in the palm of the hand. A self-contained synergistic robot hand with integrated mechanism, sensors and a suitable electrical system is built. The experiments demonstrate that the robot hand can accurately replicate the motion characteristics of two primary synergies, maintain intuitive control to simplify grasp control, has a good capability for anthropomorphic motion and can grasp different objects with versatile grasp functionality.

Hierarchical cognize framework for the multi-fault diagnosis of the interconnected system based on domain knowledge and data fusion

An interconnected system (ICS) is a complex industry system with multiple sensors, multiple tasks, and massive interaction. It is also of great importance for conducting the fault diagnosis technology research. Multi-fault diagnosis (MFD) is an urgent problem in engineering, while the complex mapping relationships among the system sensors, data patterns in single sensors, and fault modes in ICSs bringing severe challenges. The faults of ICS are similar to human disease in multiple dimensions. Enlightening the understanding of diseases in medicine guides us: hierarchical cognition and knowledge-data-fusion are important systematic ideas. Inspired by these, we propose a hierarchical cognize framework (HCF), which covers the cognition of sensors, data patterns in single sensors, and data climates. Subsequently, we propose a fuzzy neighbourhood three-way decision (FN3WD), experience fused self-adaptation Gaussian-mixture-model (EFSA-GMM), and coding-with-knowledge-discrimination (CWKD) to construct an HCF. To comprehensively verify the HCF, we successfully apply the HCF to the MFD of a satellite power system. Classic models of two-mainstream strategies are introduced as comparisons, specifically, MC-DCNN, MC-SVM, ML-DCNN, and ML-SVM. Compared to the comparative models, the HCF performs an increase of 12.35%, 7.72%, 6.90%, and 8.10% at least in accuracy, precision, recall, and F1-score, respectively, in 10 times cross-validation. Benefitting from the fusion of knowledge, the HCF has cognitive advantages in obtaining a high accuracy and precision diagnosis results. Meanwhile, the time consumption of the HCF is approximately 130 s, which is considerably reduced by as much as 50% compared with the deep learning models.

Transformer Model for Remaining Useful Life Prediction of Aeroengine

Accurate aeroengine remaining useful life (RUL) prediction plays a vital role in ensuring safe operation and reducing maintenance losses. In order to improve the accuracy of aeroengine RUL prediction, an aeroengine RUL prediction method based on the Transformer model is proposed, which gives greater weight to the characteristics of important time steps through self attention mechanism, and solves the memory degradation problem caused by too long sequence in engine RUL prediction, and excavates the complex mapping relationship between input features and aeroengine RUL. Experiments on the C-MAPSS data set show that the Transformer model can better predict the aeroengine RUL based on the aeroengine degradation data. Compared with the long short term memory network model, the root mean square error of the two sub data sets is reduced by 6.57% and 5.63% respectively.

Precise 3-D microscopic profilometry using diffractive image microscopy and artificial neural network in single-exposure manner

A single-exposure microscopic profilometry using artificial neural network (ANN) was developed for 3-D profile reconstruction of precise surface geometry. Optical profilometry is an important technique widely applied in many fields. However, in addition to optical aberration of the microscope itself, tilting of a local testing surface can affect the direction of the reflected light, thus causing undesired measurement errors. To overcome these problems, various methods or optical strategies have been developed, but most of them suffer from undue measurement errors attributed to the local surface gradient formed by the measured point and its neighboring surface geometry. This study proposed a novel method that combined diffractive image microscopy (DIM) with ANN. To train the ANN model, diffractive images of various surface orientations were collected for learning the complex mapping relationships between diffractive images and their corresponding surface orientations. The proposed method could simultaneously measure the height, tilting angle and tilting direction of a local surface. 3-D surface reconstruction by the proposed method required no prior knowledge of neighboring geometric information adjacent to the measured point. Using the developed approach achieved a surface height repeatability of 0.257 μm and a surface inclined angle repeatability of 0.037°, evidencing the realization of precise 3-D microscopic profilometry.