Fault detection and diagnosis (FDD) is crucial for stable, reliable, safe operation of industrial equipment. In recent years, deep learning models have been widely used in data-driven FDD methods because their automatic feature capability. general, these are trained on historical sensor data, therefore, it very difficult to meet the real-time requirement online applications. Since transfer can solve different but similar problems target domain efficiently effectively with knowledge learned...

Recently, research on data-driven bearing fault diagnosis methods has attracted increasing attention due to the availability of massive condition monitoring data. However, most existing still have difficulties in learning representative features from raw In addition, they assume that feature distribution training data source domain is same as testing target domain, which invalid many real-world problems. Since deep automatic extraction ability and ensemble can improve accuracy generalization...

Nowadays, motor imagery (MI) electroencephalogram (EEG) signal classification has become a hotspot in the research field of brain computer interface (BCI). More recently, deep learning emerged as promising technique to automatically extract features raw MI EEG signals and then classify them. However, learning-based methods still face two challenging problems practical applications: (1) Generally, training model successfully needs large amount labeled data. most data is unlabeled it quite...

Frequent epileptic seizures cause damage to the human brain, resulting in memory impairment, mental decline, and so on. Therefore, it is important detect provide medical treatment a timely manner. Currently, experts recognize seizure activity through visual inspection of electroencephalographic (EEG) signal recordings patients based on their experience, which takes much time effort. In view this, this paper proposes one-dimensional convolutional neural network-long short-term (1D CNN-LSTM)...

Despite the new ideas were inspired in medical treatment by rapid advancement of three‐dimensional (3D) printing technology, there is still rare research work reported on 3D coronary arteries being documented literature. In this work, application value technology cardiovascular diseases has been explored via comparison study between printed vascular solid model and computer aided design (CAD) model. paper, a framework proposed to achieve with high simulation. The patient‐specific...

Electroencephalography (EEG) plays an import role in monitoring the brain activities of patients with epilepsy and has been extensively used to diagnose epilepsy. Clinically reading tens or even hundreds hours EEG recordings is very time consuming. Therefore, automatic detection seizure great importance. But huge diversity signals belonging different makes task much challenging, for both human experts automation methods. We propose three deep transfer convolutional neural networks (CNN)...

Anomaly detection is an important issue in trajectory data mining. Various approaches have been proposed to address this issue. However, most previous studies focus only on outlier but rarely pattern mining of anomalous trajectories. Mining patterns trajectories can reveal the underlying mechanisms these outliers. This paper four distinct trajectories, and proposes a method detect classify them. First, we present difference intersection set (DIS) distance metric evaluate similarity between...

At present, the application of Electroencephalogram (EEG) signal classification to human intention-behavior prediction has become a hot topic in brain computer interface (BCI) research field. In recent studies, introduction convolutional neural networks (CNN) contributed substantial improvements EEG performance. However, there is still key challenge with existing CNN-based methods, accuracy them not very satisfying. This because most methods only utilize feature maps last layer CNN for...

As one of the most important artificial intelligence-enabled industrial applications, fault diagnosis is vital in safe, stable, and reliable operation equipment. Many existing deep learning-based methods assume that distribution training data same as testing data, which almost impossible practical applications. In addition, these are generally memory-intensive computationally expensive. A compressed unsupervised domain adaption model-based method proposed to overcome abovementioned two...

Predictive Maintenance (PdM) is a maintenance strategy which predicts equipment failures before they occur and then performs in advance to avoid the occurrence of failures. A PdM system generally consists four main components: data acquisition preprocessing, fault diagnostics, prognostics decision-making. Recently, massive condition monitoring equipment, also known as industrial big data, has shown explosive growth. large number research works, including theoretical studies applications,...

Recently, coronary heart disease has attracted more and attention, where segmentation analysis for vascular lumen contour are helpful treatment. And intravascular optical coherence tomography (IVOCT) images used to display shapes in clinic. Thus, an automatic method IVOCT is necessary reduce the doctors' workload while ensuring diagnostic accuracy. In this paper, we proposed a deep residual network of multi-scale feature fusion based on attention mechanism (RSM-Network, Residual Squeezed...

This letter proposes a novel 1.5-D algorithm for multi-channel electroencephalogram (EEG) compression. The proposed only needs to perform 1-D Discrete Wavelet Transform (DWT) rather than the 2-D version employed by previous works, and thus it results in lower computational complexity power dissipation. In this algorithm, new arranging method that exploits correlations between different sub-bands is developed concentrate energy, which causes more efficient compression using No List Set...

In many complex manufacturing environments, the running equipment must be monitored by Wireless Sensor Networks (WSNs), which not only requires WSNs to have long service lifetimes, but also achieve rapid and high-quality transmission of monitoring data centers. Traditional routing algorithms in WSNs, such as Basic Ant-Based Routing (BABR) require single shortest path, BABR algorithm converges slowly, easily falling into a local optimum leading premature stagnation algorithm. A new WSN...

Recently, telemedicine has been widely applied in remote diagnosis, treatment and counseling, where the Internet of Things (IoT) technology plays an important role. In process telemedicine, data are collected from medical equipment, such as CT machine MRI machine, then transmitted reconstructed locally three-dimensions. Due to large amount be model small storage capacity, need compressed progressively before transmission. On this basis, we proposed a lightweight progressive transmission...

We introduce RFixer, a tool for repairing complex regular expressions using examples and only consider without non-regular operators (e.g., negative lookahead). Given an incorrect expression sets of positive examples, RFixer synthesizes the closest to original one that is consistent with examples. Automatically requires exploring large search space because practical expressions: i) are large, ii) operate over very alphabets---e.g., UTF-16 ASCII---and iii) employ constructs---e.g., character...

In order to make the smart home system have ability of learning user behavior actively and provide services spontaneously, this paper introduced prediction model which combined back propagation neural network (BPNN) with Hadoop parallel computing traditional system, numerous user-generated environmental parameters data are packaged in particular frame format uploaded cloud platform through 4G or WLAN by gateway. According received historical data, repeated training BPNN run on was utilized...

This paper compares three low power schemes for the multi-hierarchy pipeline design of fixed point finite impulse response (FIR) digital filters, and we adopt an optimal CSD encoding method, minimizing number adders/subtractions in design. In addition, a 16-bit, 16 taps low-pass FIR filter is designed to investigate performance different algorithms. To evaluate them, designs are synthesized SMIC 65nm library. The evaluation shows that scheme better than other two low-power methods at same throughput.

Aortic stent has been widely used in restoring vascular stenosis and assisting patients with cardiovascular disease. The effective visualization of aortic is considered to be critical ensure the effectiveness functions clinical practice. Volume rendering ray casting as an approach enable stent. volume relies on transfer function that converts medical images into optical attributes including color transparency. This article proposes a new function, namely, multi-dimensional provide additional...

A configurable Support Vector Machine (SVM) predicting hardware accelerator applied to biological signal processing is presented in this paper. There are different types of kernel functions SVM, and our design can realize three kernels (linear, polynomial, RBF). Choosing spectral energy as the input data, range usually very large. So data should first take logarithm order narrow scope. What's more, exponential function also needed RBF kernel. Therefore, a CORDIC-based generator which...