Search

The computer simulation trust evaluation enhances the performance and recognition of the simulation system, but this technology still faces the problem of low calculation accuracy. This research proposes a data trust analysis method based on the improved average deviation degree algorithm. First, the average deviation degree and data trust degree are analyzed, and then the average deviation degree calculation model based on the second order oscillation and particle swarm optimization with repulsion factor (SecRPSO) is constructed. Finally, the performance verification and application analysis of the model are carried out. The research shows that the prediction model established based on SecRPSO optimization algorithm has better fitting degree with the real load curve, and has better prediction accuracy and is relatively stable. Comparing the mean error (ME) and absolute mean relative error (MAPE) of three different algorithms, the ME value of the model proposed in this study is 1.5673, while the ME value of Grid search and Basic PSO are 35.0983 and 14.4811, respectively. The ME value of the model proposed in this paper is significantly lower than that of other models. Moreover, its MAPE is generally small, the predicted value can track the change of the actual value, and the error fluctuates within the allowable range of engineering application. The data trust analysis method based on the average deviation algorithm can effectively calculate the expected threshold information of the data, and find out the tolerance threshold. Therefore, this method can be extended to various management data such as finance, engineering, and environmental protection.

This article presents a new image denoising algorithm that uses Gaussian Symmetric Markov random fields based on maximum a posteriori estimation. First, an image denoising model based on Gaussian Symmetric Markov random fields is built, and the image denoising problem was converted to a maximum a posteriori estimation problem. The prior probability of an image can be estimated using the Gibbs distribution, which is equivalent to Markov random fields. Second, the maximum a posteriori estimation is calculated using the expectation-maximization algorithm and conjugate gradient method, where the expectation-maximization algorithm is used to estimate Gaussian Symmetric Markov random field hyper-parameters and the conjugate gradient method is used to calculate the criterion function. The experimental results for the synthetic images and the standard Berkeley segmentation datasets demonstrate the success of the proposed Gaussian denoising filter, as compared with the state-of-the-art methods such as BM3D, WNNM, SGWD-HMMs, SSLBD, DnCNN and BUIFD.

As a key component of electromechanical equipment in the intelligent manufacturing process, rolling bearings play an important role to secure a safe, stable, and efficient operation. Deep learning can be used to guide a data-driven fault diagnosis which requires that all data are independently identically distribution (i.i.d). When the equipment is operated with multiple working conditions, the collected samples violates the assumption of i.i.d, which will inevitably make it difficult to extract accurate feature involved in the data. This paper proposes a deep learning based fault diagnosis model to recursively fuse the multiscale feature on cross working conditions, such that data without working condition label can also be referred to train a satisfying deep learning model for fault diagnosis of bearing operated in multiple working conditions. In the case when only a small number size of training samples for a separated working condition are available, the proposes fusion mechanism aims to establish a jointly learning mechanism between different working conditions. To verify the effectiveness of the proposed algorithm, experimental validation was performed using the Case Western Reserve University (CWRU) rolling bearing public data set. The experimental results show that the proposed method can make full use of a small amount of labeled data with working conditions and a large amount of labeled data without working conditions. In ten types of fault diagnosis tasks with different fault sizes, the fault diagnosis accuracy reaches more than 94% for 4 working conditions and more than 86% for 8 working conditions.

Recently, learning-based image inpainting has gained much attention. It widely utilizes an auto-encoder structure and can obtain compact feature representation in the encoder to achieve high-quality image inpainting. Although this approach has achieved encouraging inpainting results, it inevitably reduces the high-resolution representation due to interval downsampling. In order to solve this problem and achieve an excellent image inpainting effect, this paper proposes a brand-new generative network, Interactive Separation Network, which retains the high-resolution information and extracts the semantic features from corrupted images. Furthermore, this paper also discusses network designs with different complexity in different application scenarios. Finally, to improve the effectiveness and robustness of our proposal to large corrupted regions in the inpainting image, we further propose a flexible and highly reusable reconstruction scheme to complete the inpainting in the prediction process gradually. Experiments show that our proposed generation network and reconstruction scheme can significantly improve the quality of repaired images. The proposed method significantly outperforms the state-of-the-art image inpainting approaches in image quality.

A fundamental bottleneck in text-to-image synthesis is that there are rarely subjective quality evaluation metrics for a single generated image. To address this issue, this paper proposed a procedure to evaluate the single generated image, which includes a specific dataset named multiple metrics quality assessment for birds(MMQA Birds) and a learning model named blind generated image evaluator(BGIE). The motivation of our proposal is twofold. On the one hand, subjective image quality evaluation is a human perceptual task; Therefore, it tends to be a process of supervised learning. To the best of our knowledge, there are not any datasets for this study. Thus, we handle this problem via designing a specific dataset. On the other hand, we observed that the spatial content of generated image attracts more attention when humans judge its quality; According to this finding, an efficient machine-learning model that combines both pixel-level features and spatial features is proposed. Extensive experiments manifest our method can solve this problem to some extent. In the generated image dataset, BGIE surpasses the state-of-art NSS-based method by 6.3% in PLCC and SRCC. In practice, we further discuss the rationality of the MMQA Birds dataset and the application of BGIE. It proves that both in subjective and objective aspects, our method achieves convincing results.

In the case of characterizing waveguide mode converter, the widely applied back-to-back method can only extract two mode converters' joint S-parameters, the accuracy of the far-field radiation pattern method is limited. This paper experimentally demonstrates a method for accurately characterizing single circular TM01-TE11 mode converter, which consists of extracting the asymmetric two-port TM01 and TE11 mode generator test fixtures' S-parameters and de-embedding them from the “TM01 mode generator + TM01-TE11 mode Converter + TE11 mode generator” cascade measurement results. The peak problem occurring in measuring overmoded circular waveguide devices with VNA is analyzed and settled with an offset waveguide based peak-shifting tactic plus the Loess smoothing method. Fairly good agreement between the simulated and finally extracted complex S-parameter matrix of a TM01-TE11 mode converter throughout the device's working frequency span validates the proposed method's merits over traditional methods.

In the research of computer vision, artificial controllability of image synthesis is a significant and challenging task. At present, there are two available methods. One is to utilize a simple contour to determine the shape of the synthetic object. This method has a promising effect, but it can only control the shape information of the synthetic object, but not the specific content. The other is to employ the text description to synthesize the corresponding image, which effectively controls the specific content of the synthesis, but it cannot do anything for the synthesized shape. In this paper, we propose a highly flexible and human customizable image synthesis model based on simple contour and natural language description, in which the specific content of contour and text description can be determined artificially. The contour determines basic synthetic object shape, and the natural language describes specific object content. Based on these, highly authentic and customizable images can be synthesized. The experiments are executed in the Caltech-UCSD Birds (CUB) and Oxford-102 flower datasets, and the experimental results demonstrate the effectiveness and superiority of our method. The results not only maintain the contour but also conform to the natural language description. Simultaneously, the high-quality image synthesis results, based on artificial hand-drawing contour and text description, are displayed to illustrate the high flexibility and customizability of our model.

Despite the rapid development of natural language processing (NLP) implementation in electronic medical records (EMRs), Chinese EMRs processing remains challenging due to the limited corpus and specific grammatical characteristics, especially for radiology reports. In this study, we designed an NLP pipeline for the direct extraction of clinically relevant features from Chinese radiology reports, which is the first key step in computer-aided radiologic diagnosis. The pipeline was comprised of named entity recognition, synonyms normalization, and relationship extraction to finally derive the radiological features composed of one or more terms. In named entity recognition, we incorporated lexicon into deep learning model bidirectional long short-term memory-conditional random field (BiLSTM-CRF), and the model finally achieved an F1 score of 93.00%. With the extracted radiological features, least absolute shrinkage and selection operator and machine learning methods (support vector machine, random forest, decision tree, and logistic regression) were used to build the classifiers for liver cancer prediction. For liver cancer diagnosis, random forest had the highest predictive performance in liver cancer diagnosis (F1 score 86.97%, precision 87.71%, and recall 86.25%). This work was a comprehensive NLP study focusing on Chinese radiology reports and the application of NLP in cancer risk prediction. The proposed NLP pipeline for the radiological feature extraction could be easily implemented in other kinds of Chinese clinical texts and other disease predictive tasks.

With the development of the Vehicular Ad-Hoc Network (VANET), protecting user privacy while ensuring security has become an immense challenge. A ring signature has the characteristics of no manager and unconditional anonymity, which can substantially protect user privacy. However, two problems exist in practical applications: first, unconditional anonymity cannot track the signer, which has potential security risks; and second, ordinary ring signatures cannot resist quantum attacks. In response to these problems, this paper proposes a Traceable Ring Signature (TRS) scheme on ideal lattice. In the scheme, the traceability of the ring signature is achieved by adding information to the ring signature. Selecting the algorithm on the ideal lattice to generate the master key can effectively resist quantum attacks. The scheme gives a detailed security proof and efficiency analysis. The new scheme has strong unforgeability for the difficult problem of Short Integer Solution (SIS) Question and a smaller signature size.

Multivariate statistical method is one of data-driven fault diagnosis methods, which is widely used in complex industrial systems to realize faults detection. And, the common methods include the principal component analysis, principal component regression, and partial least squares (PLS). Compared with the PLS, improved principal component regression (IPCR) improves the alarm performance in the quality-related and quality-independent parts. Considering the advantage of the mutual information is good for selecting quality-related variables for modeling, the mutual information is employed to improve the principal component regression (MIIPCR). Through the design and analysis of the algorithms, the MIIPCR can give a performance boost for the principal components, especially for the IPCR in fault detection of Tennessee Eastman process (TEP). Compared with the IPCR, the MIIPCR has the strong advantage to improve feedback failures five and seven of the TEP. In addition, the MIIPCR also could apply to detect the other faults and confirm the higher than the IPCR. And for the unrelated faults, the MIIPCR also has the great identification ability. Lots of simulation had been on for the TEP, the simulation results showed that it's the great effectiveness of the MIIPCR.

Link prediction is a demanding task in real-world scenarios, such as recommender systems, which targets to predict the unobservable links between different objects by learning network-structured data. In this paper, we propose a novel multi-view graph convolutional neural network (MV-GCN) model to solve this problem based on Matrix Completion method by simultaneously exploiting the interactive relationship and the content information of different objects. Unlike existing approaches directly concatenate the interactive and content information as a single view, the proposed MV-GCN improves the accuracy of the predictions by restricting the consistencies on the graph embedding from multiple views. Experimental results on six primary benchmark datasets, including two homogeneous datasets and four heterogeneous datasets, both show that MV-GCN outperforms the recent state-of-the-art methods.

Unsupervised feature learning, as a promising tool for extracting features automatically, overcomes shortcomings of traditional feature extraction methods which generally take plenty of effort on designing features. Among various unsupervised feature learning methods, sparse filtering is an efficient and popular one owing to its simplicity with fewer hyper-parameters. However, the standard sparse filtering fails to consider the local structure of input samples, which may affect and restrict its feature learning ability. To address this deficiency, a new feature extraction approach named sparse filtering with local structure preserved (SF-Local) is proposed. The objective function of SF-Local is composed by the standard sparse filtering and a local structural regularization (LSR) which is formulated to preserve the local structure of the input samples. To design LSR, we adopt complexity-invariant distance to select neighbors for each sample rather than the widely used Euclidean distance, which is beneficial to establishing a reliable K-nearest graph. Based on SF-Local, an intelligent fault diagnosis method is developed by leveraging a supervised classifier. The developed diagnostic approach is applied to diagnose the bearing faults as well as its performance is evaluated on a bearing data set with a constant operating condition and a data set with variable operating conditions. Experimental results demonstrate the effectiveness and superiority of the proposed SF-Local.

To accurately measure the performance of overmoded devices used in high-power microwave systems, a novel structure that effectively converts the TE10 mode in a rectangular waveguide to a high-purity TM01 mode in a circular waveguide is proposed. The circular TM01 mode is excited by a hybrid rectangular E-T waveguide structure, which uses a circular waveguide as the E element. Convex slots loaded at the rectangular waveguide, axially symmetric about the cylindrical waveguide, are introduced to suppress the competing lower order TE11 mode. A tuning stub loaded at the rectangular short wall is used to increase the frequency tuning range. The scattering matrix technique was used to study the mode converter along with the microwave network analysis method using some generalizations of the input port reflection and conversion efficiency, which were verified by the numerical full-wave simulation. Based on the theoretical analysis, an X-band TM01 mode generator is designed and measured. Low-power tests show that, when the length of the tuner increases from 5.7 to 11.7 mm, a wide frequency tuning range of ~500 MHz is obtained with a conversion efficiency of ~99%. Measured results and the theoretical analysis are in good agreement.

As various faults are inevitable during a satellite's on-orbit operation, a modified intermittent process performance monitoring method is proposed. This performance monitoring of a satellite is performed via attitude information. This method not only overcomes the defects of the traditional intermittent process methods of requiring a priori knowledge and the difficulty in handling points on adjacent edges but also reduces the defects, such as omissions and false positives caused by the process modeling. This method combines the multiphase auto regression principal component analysis monitoring method based on affine propagation clustering (APC) optimized, and at the meanwhile, the population diversity-based particle swarm optimization algorithm is considered in APC. Numerical simulations proved the effectiveness of the proposed approach.

The various faults that inevitably occur represent a primary issue in satellite on-orbit operation. Fault diagnosis is the first step in the fault control process. As a means to perform this step, a semi-supervision fault diagnosis method via attitude information is proposed for a satellite. This method combines static fusion with dynamic updating. The evidence concept is employed to obtain the fault information. It not only allows the detection of the slow change and failure with interference, but also confirms the optimal fusion and updating parameters via historical data. Numerical simulations, including static fusion diagnosis and dynamic updating diagnosis, are all presented with the proposed semi-supervision diagnosis methods to compare and prove the effectiveness of the proposed approach.

A rigid satellite fault diagnosis strategy, subject to faults of external disturbances and thruster faults, is developed. In this design, an equivalent idea is introduced to design a sliding mode observer that can detect and identify the failures indicated previously. Considering the measurability of parameters of the satellite, such as angular velocity and attitude, a sliding mode observer is implemented. Next, the amplitudes of the faults and disturbances can be detected, identified, and estimated via the sliding mode observer; these tasks are accomplished with zero estimation error within a finite period. A sliding mode-based attitude controller is developed using an exponential reaching law that relies on the system states and the corresponding formation of the estimated parameters. The controller not only guarantees the attitude system to be stable but also governs the attitude and angular velocity to converge to zero within a finite period. The good reliability of the proposed controller can be proved via multiple simulation tests.